test.jsonl

{"id": "id1659", "doc_id": "XIN_ENG_20061102.0112.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLocal people were also <e1>called</e1> on to report <e2>deaths</e2> of wild birds to the provincial wildlife protection center.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id684", "doc_id": "AFP_ENG_20061211.0352.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061211.0352\n\nJOHANNESBURG, <t1>Dec 11 , 2006</t1>, 2006\n\nSouth African lifespans cut short by AIDS\n\n\n\n\nSouth Africa, which has the world's second heaviest caseload of HIV/AIDS, has seen average life expectancy fall by 13 years since 1990 to 51, a new study has <e1>said</e1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6232", "doc_id": "XIN_ENG_20051104.0292.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nUS President George W. Bush, who arrived here <t1>Thursday night</t1>, was attending the two-day summit which will be focusing on topics like free trade and <e1>fighting</e1> poverty.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6668", "doc_id": "AFP_ENG_20051204.0479.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe <e1>trial</e1> is <e2>set</e2> to close February 10 after all the suspects have gotten a change to make a final declaration and the judges are expected to hand down a verdict on February 24.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id7132", "doc_id": "AFP_ENG_19970427.0491.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970427.0491\n\nSUVA, April 28 , 1997 (AFP)\n\nFiji family killed in landslide\n\n\n\n\nA 16-year-old girl was recovering in hospital Monday after seeing her sister and grandparents <e1>killed</e1> as their home was buried by a landslide, emergency authorities <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id277", "doc_id": "XIN_ENG_20051102.0223.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe friendly exchange and <e1>cooperation</e1> between China and Japan witnessed sound progress since the two countries normalized diplomatic relations 33 years ago, Jia <e2>said</e2>, noting it is worthy to be cherished and preserved by the two sides.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6573", "doc_id": "XIN_ENG_20051107.0284.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nConservative Leader Stephen Harper, who spoke moments after Layton's <e1>speech</e1>, <e2>said</e2> he believes the NDP leader still has not made up his mind and is still negotiating with the Liberals.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6415", "doc_id": "PRI19980213.2000.0313.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nPRI19980213.2000.0313 \n NEWS STORY \n\n\n\n\n\n Officials in California  are <e1>warning</e1> residents that oncoming <e2>rains</e2>  will cause dangerous and unpredictable landslides.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2183", "doc_id": "XIN_ENG_20061127.0236.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n27 , 2006 (Xinhua)\n\nWritethru: Italy to complete troop withdrawal from Iraq within days\n\n\n\n\nItalian Prime Minister Romano Prodi said on <t1>Monday</t1> that Italy will complete its troop withdrawal from Iraq by <t2>Dec.</t2>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1870", "doc_id": "AFP_ENG_20051206.0284.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We think there is a good chance of reaching an <e1>agreement</e1> at the council next week,\" he <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id818", "doc_id": "AFP_ENG_19970414.0322.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nNaidu's party, which decimated the Congress in his home state <t1>last year</t1>, <e1>enjoys</e1> grass-root support across the coastal state.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id792", "doc_id": "XIN_ENG_20061115.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBut I personally <e1>support</e1> the <e2>election</e2> of a young, capable and vigorous writer as the new president,\" writer Hai Yan was quoted as saying by the China News Service.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2923", "doc_id": "AFP_ENG_20051207.0417.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"My heart was telling me to stay in Likud but my head told me otherwise and so therefore I am today <e1>announcing</e1> my resignation from Likud after representing the party for the last 17 years in the Knesset,\" he <e2>told</e2> a press conference.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id3723", "doc_id": "XIN_ENG_20051103.0169.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn a separate statement, the US embassy <e1>said</e1> the Visiting Forces Agreement (VFA) , signed in 1998 between the Philippine and US governments to allow US troops return to the Philippines after they pulled out in 1993, provides the mechanism for US and Philippine authorities to work together to <e2>determine</e2> the facts of the case and to ensure that accused individuals are available to both Filipino and US investigators.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id7928", "doc_id": "XIN_ENG_20051116.0158.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"I am <e1>looking</e1> forward to facing a great competition on a fast <e2>course</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id2873", "doc_id": "AFP_ENG_19970417.0042.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe \"biggest losers\" were South Korea, Malaysia, where a third of the world's microprocessor production is <e1>tested</e1> and packaged, Singapore, which <e2>makes</e2> 40 percent of the world's disk drives, and Taiwan, which makes eight out of 10 circuit boards and half of all keyboards.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3620", "doc_id": "AFP_ENG_20061214.0704.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe gunmen, mostly youths, <e1>smashed</e1> windows and furniture inside the terminal -- Gaza's only border <e2>crossing</e2> that bypasses Israel -- while shots were fired into the air and at the building itself.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6551", "doc_id": "XIN_ENG_20051107.0284.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLayton has <e1>made</e1> general comments, but not once has he said what specific change he'd like to see to this package that will strengthen public health care,\" the minister <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id5931", "doc_id": "AFP_ENG_19970403.0524.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970403.0524\n\nLONDON, <t1>April 3 , 1997</t1> (AFP)\n\nClimber killed abseiling on Britain's highest peak\n\n\n\n\nA climber who fell while abseiling on Britain's highest mountain has <e1>died</e1> from his injuries, police said Thursday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id5954", "doc_id": "AFP_ENG_19970420.0013.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA similar confrontation <e1>happened</e1> in <t1>1991</t1>, but no-one was injured.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id851", "doc_id": "XIN_ENG_20051112.0145.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We are <e1>predicting</e1> a peaceful election,\" <e2>said</e2> Silaigwana.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2430", "doc_id": "XIN_ENG_20061127.0437.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe UN mission, known by its French acronym MONUC, <e1>said</e1> that its blue helmets (UN forces), fighting alongside the National Army Force in the DRC (FARDC), <e2>launched</e2> air strikes on the rebel 81st and the 83rd brigades loyal to General Laurent Nkunda.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id908", "doc_id": "XIN_ENG_20051117.0406.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nWhen local police <e1>approached</e1> <t1>May</t1>'s house in Ky Son district on the morning of Nov.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id716", "doc_id": "AFP_ENG_20061211.0352.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBased on a scenario <e1>developed</e1> by researchers that by 2015 50 percent of those needing treatment will be receiving it, the impact of AIDS deaths could slow down in <t1>the coming years</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7084", "doc_id": "AFP_ENG_20061230.0186.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061230.0186\n\nROME, Dec 30 , 2006, 2006\n\nATTENTION - ADDS further reax ///\n\n\n\n\nItalian Prime Minister Romano Prodi <e1>said</e1> Saturday he was \"concerned\" over tensions fanned in Iraq by the <e2>execution</e2> of former Iraqi leader Saddam Hussein.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2112", "doc_id": "XIN_ENG_20061121.0337.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe prime minister <e1>called</e1> for unity in the deeply divided country and pledged that his government will take up all its responsibilities to <e2>protect</e2> the interests of the people.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1351", "doc_id": "AFP_ENG_20061219.0124.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"The BoJ may need to confirm first that consumer spending has <e1>accelerated</e1> before hiking rates,\" said NLI Research Institute senior economist Taro Saito, who is looking at a rate <e2>increase</e2> either in January or February.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1342", "doc_id": "AFP_ENG_20061219.0124.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn an outlook approved by the cabinet, the government <e1>expected</e1> the economy to continue its gradual <e2>recovery</e2> in the next fiscal year, led by domestic private demand and strong corporate earnings.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id4894", "doc_id": "AFP_ENG_19970413.0013.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"I <e1>think</e1> definitely the home market will be affected but to what extent nobody can tell,\" <e2>said</e2> a property agent, adding that property stocks were expected to stabilise Monday given the week-end selling bout.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id3538", "doc_id": "AFP_ENG_19970418.0499.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nFriday's <e1>talks</e1> were specially choreographed to minimize tensions with China, with an official meeting between Gore and Lee which Clinton <e2>dropped</e2> in on while it was \"in progress.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5875", "doc_id": "XIN_ENG_20051117.0274.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Small fish have to <e1>fight</e1> hard for survival,\" he added, <e2>referring</e2> to traditional and small opposition parties.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6429", "doc_id": "APW19980418.0210.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nOn May 22, 1995, Farkas  was made a brigadier general, and <t1>the\nfollowing year</t1> he  was <e1>appointed</e1> military attache at the Hungarian\nembassy in Washington.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id7786", "doc_id": "XIN_ENG_20061109.0149.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe police have normalized the situation and calm seemed to have returned <t1>last night</t1>,\" Kibunja <e1>said</e1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4099", "doc_id": "AFP_ENG_19970425.0196.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLtd., two of Japan's Big Four stockbrokers, <e1>announced</e1> <t1>Friday</t1> net losses for the year to March, joining market leader Nomura Securities Co.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6741", "doc_id": "XIN_ENG_20051102.0256.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA national committee on fighting the bird flu <e1>issued</e1> a decree <e2>urging</e2> all fever hospitals across Egypt to heighten preventive measures against the disease.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id5310", "doc_id": "XIN_ENG_20061123.0035.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nGroup leader Ron Gould <e1>said</e1> here on Wednesday that both the government and ECZ would abide by the recommendations that would <e2>reflect</e2> in their report to be handed to ECZ in two weeks time.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6302", "doc_id": "AFP_ENG_19970421.0445.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAccording to sources close to <t1>Monday</t1>'s talks, Wu Yi accepted the need for greater flexibility, but insisted that the WTO members take into account China's <t2>current</t2> level of economic development.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id5583", "doc_id": "XIN_ENG_20061119.0263.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe project, jointly <e1>initiated</e1> by the State Environmental Protection Administration and Nature Conservation China, will gradually be <e2>implemented</e2> across the country.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7814", "doc_id": "XIN_ENG_20061129.0098.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe men's program has <e1>varied</e1>, but has become relatively standardized at all major <e2>games</e2> since 1928.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id4256", "doc_id": "XIN_ENG_20061121.0116.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBello said \"issues about availability, accessibility and affordability of food would <e1>feature</e1> prominently at the <e2>summit</e2>.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6364", "doc_id": "AFP_ENG_20061219.0267.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMourinho was enraged by an <e1>incident</e1> in which Johnson jumped over Chelsea goalkeeper Hilario as they both went for the ball and <e2>tumbled</e2> after landing.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2764", "doc_id": "XIN_ENG_20061111.0053.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nZimbabwe and Botswana's <e1>co-operation</e1> in wildlife management including the establishment of the Shashe/Limpopo and Kavango/ Zambezi Transfrontier Conservation areas was greatly <e2>appreciated</e2> by the commission which hailed the move as an indication of close cooperation.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id7765", "doc_id": "XIN_ENG_20061109.0149.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n9 , 2006 (Xinhua)\n\nThousands flee homes in Kenya slum fighting\n\n\n\n\nThousands of Kenyans are fleeing their homes in a <t1>fourth day</t1> of slum gang <e1>violence</e1> that has left at least eight people dead, witnesses and police said on Thursday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id4318", "doc_id": "AFP_ENG_20061221.0605.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We have to <e1>yield</e1> to the new generations ... or continue to <e2>lead</e2> the way for the new generations, gradually of course.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6652", "doc_id": "XIN_ENG_20051125.0186.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n25 , 2005 (Xinhua)\n\nFactbox of women's all-around world champion Chellsie Memmel\n\n\n\n\nFollowing is the factbox of American Chellsie Memmel, who won women's all-around at the 38th gyamnastics world championships here on Friday:\n\nChellsie Memmel, the United States\n\nPlace of birth: West Allis, WI, the United States\n\nDate of birth: Junuary 23, 1988\n\nHeight: 154 cm\n\nWeight: 50 kg\n\nClub: M Gymnastics in the United States\n\nCoach: Andrew Memmel, her father, and Laurie Glazer\n\nIdol: her parents\n\nA steady performer, she helped the American team <e1>win</e1> its first world team title in <t1>2003</t1> and was a co-champion on uneven bars that year.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id2875", "doc_id": "AFP_ENG_19970417.0042.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970417.0042\n\nMANILA, April 17 , 1997 (AFP)\n\nADB forecasts moderate recovery for battered Asian electronics exports\n\n\n\n\nGlobal demand for high-technology electronics products should make a moderate recovery in 1997 but <t1>the boom years</t1> of the early 1990s are unlikely to return, the Philippines-based Asian Development Bank (ADB) said <t2>Thursday</t2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3850", "doc_id": "AFP_ENG_19970413.0153.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA BJP ideologue, Jay Dubashi, <e1>said</e1>: \"The panel will meet next week to thrash out an agenda to suggest <e2>changes</e2> in the finance bill.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id63", "doc_id": "XIN_ENG_20051102.0074.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nWu, chairman of the Standing Committee of the National People's Congress, voiced the hope that the SEM and its advisory council would <e1>increase</e1> innovation capacity and <e2>train</e2> more talented personnel on science and economic management.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id2412", "doc_id": "AFP_ENG_20051204.0086.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n'Monty' <e1>birdied</e1> four, 10 and 14 but bogeyed 11 to be one <e2>shot</e2> back from Kingston, who was 10 under par and leading by a shot until the final hole.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2633", "doc_id": "AFP_ENG_19970406.0166.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHe <e1>said</e1> traffic has been banned during the day because of the warming temperatures that <e2>increase</e2> the risk of avalanches.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3403", "doc_id": "AFP_ENG_20061203.0391.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLast month, Pinochet was charged with alleged crimes at Villa Grimaldi, a secret jail where Michelle Bachelet, now president of Chile, was once <e1>tortured</e1>, along with her mother, for a month in <t1>1975</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id4181", "doc_id": "AFP_ENG_20061213.0712.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe party <e1>conference</e1> last year was <e2>held</e2> against the backdrop of its sweep in elections to a new parliamentary Senate.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id3506", "doc_id": "AFP_ENG_19970415.0517.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAccording to EBRD figures, new projects approved for Hungary plunged from 406 million ECUs (485 million dollars) in <t1>1995</t1> to 133 million ECUs in <t2>1996</t2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6076", "doc_id": "WSJ910225-0066.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Fight them,\" he <e1>urged</e1> Iraqis in a radio\n<e2>address</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id2377", "doc_id": "XIN_ENG_20051128.0108.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe relics, most are still in <e1>use</e1>, have <e2>witnessed</e2> the earliest pervasion of Western religious culture in the oriental continent as well as its ensuing coexistence with the local culture.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id211", "doc_id": "AFP_ENG_20061218.0111.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBut these were <e1>cancelled</e1> at the last minute by Suffolk Police, prompting media speculation that the investigating team needed to <e2>carry</e2> out more tests.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id4530", "doc_id": "AFP_ENG_20061213.0531.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061213.0531\n\nFRANKFURT, <t1>Dec 13 , 2006</t1>, 2006\n\nBundesbank sees German economy growing by 2.75 pct in 2006\n\n\n\n\nThe Bundesbank <e1>expects</e1> the German economy to grow by around 2.75 percent this year, thereby providing \"substantial impulses\" for the 12-country eurozone as a whole, the German central bank's president Axel Weber said on Wednesday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5915", "doc_id": "AFP_ENG_20061202.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLeoncio Amadore, one of the Philippines' foremost meteorologists, <e1>believes</e1> that climate <e2>change</e2> is also contributing to the severity of the typhoons now hitting the Philippines.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6068", "doc_id": "WSJ910225-0066.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHe went on to add\nthat \"in so <e1>doing</e1> you not only <e2>return</e2> a nation to its people,\nbut you will destroy the war machine of a ruthless dictator.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2175", "doc_id": "XIN_ENG_20061127.0236.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn an <e1>interview</e1> with local television Telelombardia, Prodi <e2>said</e2> that Italy currently has only 60 to 70 troops left in Nassiriyah and they will be back home between Dec.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id1000", "doc_id": "AFP_ENG_20051206.0280.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe head of the British delegation, John Sawers, <e1>said</e1> the European Union proposed a <e2>compromise</e2> over the final declaration that was initially accepted by Washington while the Russians were still considering it.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3803", "doc_id": "PRI19980115.2000.0186.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA federal appeals court  has reinstated his state <e1>convictions</e1>\n for securities <e2>fraud</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id5712", "doc_id": "XIN_ENG_20051110.0134.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe bureau's <e1>move</e1> confirmed long-circulating speculation in the market that it would sell copper reserves to <e2>bring</e2> down the runaway copper price.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2320", "doc_id": "AFP_ENG_19970425.0525.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nOn April 9, Iraq also <e1>flouted</e1> the UN sanctions in force since its August 1990 invasion of Kuwait and flew more than 100 Iraqis to Saudi Arabia for the start of the <e2>pilgrimage</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id4669", "doc_id": "AFP_ENG_20061204.0442.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIt was unclear when Bush would <e1>announce</e1> his choice for a successor to Bolton, who will leave his post at <t1>the end of the year</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1758", "doc_id": "XIN_ENG_20061114.0441.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"They could be <e1>fined</e1> up to three times the revenue they got from selling the goods,\" an official with the administration <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2181", "doc_id": "XIN_ENG_20061127.0236.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nProdi added that he <e1>talked</e1> about the <e2>pullout</e2> with U.S.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7280", "doc_id": "XIN_ENG_20051123.0113.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLi Changchun, member of the Standing Committee of the Political Bureau of the CPC Central Committee, <e1>made</e1> the remark when <e2>holding</e2> talks with Philip Mangula, general secretary of the ruling Revolutionary Party of Tanzania.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6182", "doc_id": "XIN_ENG_20061103.0235.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHis body was temporarily <e1>kept</e1> in the then Central Park, which was <e2>renamed</e2> Zhongshan (Sun Yat-sen) Park in 1928.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3467", "doc_id": "AFP_ENG_19970415.0517.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe EBRD was <e1>set</e1> up in 1991 to <e2>assist</e2> the transition to a market economy in eastern and central Europe and the former Soviet Union.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3673", "doc_id": "AFP_ENG_20061229.0165.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We could <e1>increase</e1> transit with the volume needed to ensure stable functioning of the economies of our European neighbours,\" Ukraine's Energy Minister Yury Boyko <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6434", "doc_id": "APW19980418.0210.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHowever , cited by District of Columbia traffic police in\n<t1>December</t1> for driving under the influence of alcohol, Farkas  was ordered home and <e1>retired</e1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1911", "doc_id": "AFP_ENG_20061229.0486.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHistory will judge the deeply flawed Dujail trial and this <e1>execution</e1> harshly,\" Dicker <e2>said</e2> in a statement.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1450", "doc_id": "XIN_ENG_20051118.0318.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Chelsea lost the last game they played and for me when this happens to any team it <e1>makes</e1> them a dangerous animal,\" Souness told reporters on <t1>Friday</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3839", "doc_id": "AFP_ENG_19970413.0153.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIndia's financial year <e1>begins</e1> <t1>April 1</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5067", "doc_id": "XIN_ENG_20061110.0141.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"<t1>Today</t1> we see the <e1>use</e1> of home-made improvised explosive devices.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4560", "doc_id": "XIN_ENG_20051115.0131.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA nine year old boy and his sister <e1>developed</e1> symptoms of pneumonia in xianggang county, central-south Hunan province, last month after <e2>eating</e2> a chicken that had died of bird flu.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2895", "doc_id": "AFP_ENG_20051207.0417.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nPolice <e1>opened</e1> an investigation in August last year after a report by state comptroller Eliezer Goldberg found Hanegbi had \"trampled on the law\" by <e2>appointing</e2> dozens of political supporters to posts while environment minister.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2635", "doc_id": "AFP_ENG_19970406.0166.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970406.0166\n\nMOSCOW, April 6 , 1997 (AFP)\n\nTwo killed by avalanche on road in Caucasus\n\n\n\n\nTwo people were killed and two others injured when their car was hit by an <e1>avalanche</e1> on a main road running through the Russian republic of Northern Ossetia, ITAR-TASS news agency <e2>reported</e2> Sunday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7674", "doc_id": "AFP_ENG_19970422.0733.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn the last election in 1992, polls <e1>predicted</e1> a Labour <e2>victory</e2> but the Conservatives eventually won.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5895", "doc_id": "AFP_ENG_20061202.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"The risk to human life from natural disasters in the Philippines has <e1>increased</e1> dramatically over the past generation,\" he <e2>said</e2> in a recent report.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6278", "doc_id": "AFP_ENG_19970421.0445.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe breakthroughs in the March talks were China's <e1>undertaking</e1> to grant individual overseas firms import-export rights three years after <e2>joining</e2> the WTO, and its pledge to abide by all WTO rules on intellectual property rights from the moment of its accession.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6533", "doc_id": "XIN_ENG_20051107.0284.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nLast week, Federal Health Minister Ujjal Dosanjh <e1>provided</e1> Layton with a <e2>proposal</e2> on promoting public health care, which was criticized by Layton as \"disappointing\" Friday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1902", "doc_id": "AFP_ENG_20061229.0486.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061229.0486\n\nBAGHDAD, Dec 30 , 2006, 2006\n\nRights group condemns Saddam execution\n\n\n\n\nUS-based rights group Human Rights Watch (HRW) Saturday <e1>condemned</e1> the hanging of deposed Iraqi dictator Saddam Hussein, saying history would <e2>judge</e2> his trial and execution \"harshly\".\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id3647", "doc_id": "AFP_ENG_20061229.0165.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nGazprom vice-president Alexander Medvedev <e1>told</e1> France's Le Figaro daily on Friday that this amounted to \"grotesque blackmail\" and <e2>warned</e2> that European clients could face shortfalls.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2927", "doc_id": "AFP_ENG_20051207.0417.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHanegbi, a minister without portfolio, <e1>became</e1> the sixth minister to follow Sharon out of Likud as part of the continuing turmoil in Israeli politics in the run-up to a March general <e2>election</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6108", "doc_id": "WSJ910225-0066.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\ncommanders said 5,500 Iraqi prisoners were <e1>taken</e1> in the first\nhours of the ground <e2>war</e2>, though some military officials later said the total may have climbed above 8,000.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5286", "doc_id": "XIN_ENG_20061111.0084.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n11 , 2006 (Xinhua)\n\ndecease\n\n\n\n\nThe Palestinians in the West Bank and Gaza Strip <e1>marked</e1> on Saturday two-year anniversary of Palestinian leader Yasser Arafat's <e2>decease</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1936", "doc_id": "AFP_ENG_19970409.0548.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nSpeaking to journalists after <e1>talks</e1> with UN Secretary General Kofi Annan, Denktash <e2>said</e2> he was optimistic about UN plans to try and get the leaders of the feuding communities on the divided island together for a June meeting.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1746", "doc_id": "AFP_ENG_19970421.0123.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe British strength is now around 1,800 which will <e1>fall</e1> to around 1,600 by the end of May, leaving only a few to be withdrawn in the final few weeks before the <e2>handover</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1393", "doc_id": "XIN_ENG_20061123.0116.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe mourning period, <e1>starting</e1> on Thursday, <e2>came</e2> after all 23 miners trapped in the Halemba coal mine in the southern Polish city of Ruda Slaskain were confirmed dead, the report said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7302", "doc_id": "AFP_ENG_20061213.0676.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nUnder the new law, the federal government <e1>banned</e1> semi-automatics and pump-action rifles and shotguns and offered to buy these weapons from gunowners, <e2>using</e2> a levy on income tax.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1969", "doc_id": "XIN_ENG_20061116.0230.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe Commission should meet regularly to <e1>review</e1> the ongoing projects and issues and explore possibility for further avenues of cooperation, the Russian foreign minister <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6088", "doc_id": "WSJ910225-0066.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nColin Powell, chairman of the\nJoint Chiefs of Staff, <e1>returned</e1> from a <e2>visit</e2> with military\ncommanders in Saudi Arabia, administration officials say.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2279", "doc_id": "AFP_ENG_19970425.0525.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA second group of <e1>returning</e1> pilgrims was <e2>expected</e2> at Jodaidat Arar, three kilometres (two miles) from the border, on Saturday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id4138", "doc_id": "AFP_ENG_19970425.0196.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nYuki said earnings were pressured by <e1>declining</e1> profits at subsidiaries in Europe, the United States and Asia in <t1>the year</t1> to March although banking units in the United States and Luxembourg booked firm profits.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id3715", "doc_id": "AFP_ENG_20061213.0683.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMicrosoft chief operating officer Kevin Turner said the partnership was aimed at <e1>winning</e1> a major share of the business network market, which is <e2>predicted</e2> to exceed 50 billion dollars in 2007.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7411", "doc_id": "XIN_ENG_20051115.0235.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n15 , 2005 (Xinhua)\n\n53 birds die of avian flu in Britain\n\n\n\n\nFifty-three birds have died at a British quarantine center <e1>hit</e1> by an outbreak of avian flu, according to a report <e2>released</e2> on Tuesday by the British Department for the Environment, Food and Rural Affairs.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id2529", "doc_id": "XIN_ENG_20051107.0285.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nSeparately, the US Federal Reserve said Monday that the average yield for one-year Treasury bills, a popular index for making changes in adjustable rate mortgages, <e1>rose</e1> to 4.32 percent last week from 4.26 percent <t1>the previous week.</t1>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5686", "doc_id": "AFP_ENG_20051205.0111.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA second installment of the program, <e1>making</e1> more explosive claims concerning the validity of Hwang's <e2>research</e2>, was scheduled to air Tuesday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6866", "doc_id": "AFP_ENG_20061227.0080.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nRegional peacemaker Kenya, which helped form the Somali government in 2004, <e1>asked</e1> Ethiopia to cease military operations, as it announced <e2>plans</e2> to organise a regional meeting to discuss the escalation of the conflict in Somalia.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3442", "doc_id": "AFP_ENG_20061203.0391.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nDozens of Pinochet's men are behind bars in Chile for killing and <e1>torturing</e1> some 3,000 persons, by official count, under his rule\n\nBut Pinochet has yet to <e2>face</e2> trial for any crime.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id38", "doc_id": "XIN_ENG_20061101.0308.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nRelations between Moscow and Tbilisi have been strained by tensions over Georgia's breakaway regions of South Ossetia and Abkhazia and the country's <e1>bid</e1> to <e2>join</e2> NATO since Saakashvili came to power in 2003.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id2410", "doc_id": "AFP_ENG_20051204.0086.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20051204.0086\n\nHONG KONG, Dec 4 , 2005\n\n= (PICTURES) =\n\n\n\n\nUBS Hong Kong Open champion Colin Montgomerie did his best to <e1>console</e1> a tearful James Kingston on <t1>Sunday</t1> after the South African threw away a one-shot lead on the last hole.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id703", "doc_id": "AFP_ENG_20061211.0352.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAccording to government figures for September, 213,000 infected people now <e1>benefit</e1> from a government-funded antiretroviral plan, and 11,000 more <e2>join</e2> each month.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2666", "doc_id": "XIN_ENG_20061106.0199.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Albania is home to the internationally renowned archeological sites, such as the Butrint National Park, the country's Stone City of Gjirokastra and its 2500-year-old city of Berati,\" he <e1>said</e1>, <e2>adding</e2> that Berati is expected to be granted World Heritage site status in 2007.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id7115", "doc_id": "AFP_ENG_19970412.0213.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970412.0213\n\nSARAJEVO, April 12 , 1997 (AFP)\n\nPolice foil apparent attempt to kill pope\n\n\n\n\nPolice in Sarajevo on <t1>Saturday</t1> foiled an apparent attempt to <e1>kill</e1> Pope John Paul II, discovering 23 landmines wired to explode under a bridge the pope was set to cross later in the day.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7363", "doc_id": "AFP_ENG_20051214.0488.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"It is important we <e1>counter</e1> that, and we have to counter it truthfully, to some significant degree to counter it with transparency, and we are looking at all kinds of ways to do that,\" he <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2636", "doc_id": "AFP_ENG_19970406.0166.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970406.0166\n\nMOSCOW, <t1>April 6 , 1997</t1> (AFP)\n\nTwo killed by avalanche on road in Caucasus\n\n\n\n\nTwo people were killed and two others injured when their car was hit by an <e1>avalanche</e1> on a main road running through the Russian republic of Northern Ossetia, ITAR-TASS news agency reported Sunday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3901", "doc_id": "AFP_ENG_20051216.0176.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe Democrats have also struggled to find a common voice on Iraq, torn between the public disenchantment with the US military campaign and the need to offer a viable alternative to the administration's \"stay the course\" approach -- particularly after Democratic Representative John Murtha <t1>last month</t1> <e1>called</e1> for an immediate troop pullout.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id2580", "doc_id": "XIN_ENG_20061119.0216.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe main covers came on and off several times during <t1>the afternoon</t1> because of the <e1>rain</e1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6300", "doc_id": "AFP_ENG_19970421.0445.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Much of the important <e1>work</e1> -- and the toughest issues -- have been left to the end,\" he said, while <e2>warning</e2> Beijing that WTO membership was crucial to the country's economic future.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4151", "doc_id": "AFP_ENG_20061213.0712.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMugabe opens the conference on Friday in the obscure rural centre of Goromonzi although preparatory meetings were <e1>beginning</e1> on <t1>Thursday</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id1724", "doc_id": "AFP_ENG_19970421.0123.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nDutton gave a veiled warning about deep-rooted fears among the many Hongkongers who <e1>associate</e1> the PLA with bloody images of tanks crushing student demonstrators in Tiananmen Square in <t1>June 1989</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id2029", "doc_id": "XIN_ENG_20061117.0389.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"They <e1>need</e1> to be told that the fees they have paid for treating the wastewater have indeed been used for that purpose and the money to build facilities for the more efficient use of water resources has really been used in that area,\" he <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1231", "doc_id": "AFP_ENG_20061220.0510.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe revalidation of the 2002/2003 voters cards is \"to <e1>ensure</e1> that all eligible voters are able to vote during the forthcoming general elections,\" the Independent National Electoral Commission (INEC) <e2>said</e2> in an official statement.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2378", "doc_id": "XIN_ENG_20051128.0108.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n<e1>Joining</e1> UNESCO's Convention Concerning the Protection of the World Cultural and Natural Heritage in 1985, China now has 31 sites on the World Heritage List, the third most next to Italy and Spain, and it is now playing an active role in <e2>protecting</e2> the world heritages.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id962", "doc_id": "AFP_ENG_20061228.0348.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061228.0348\n\nAGEN, France, Dec 28 , 2006, 2006\n\nLaporte looks to Irish strategy for World Cup success\n\n\n\n\nFrance coach Bernard Laporte <e1>revealed</e1> Thursday that he would be <e2>looking</e2> to emulate Ireland's strategy as the hosts prepare for next year's rugby World Cup.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1201", "doc_id": "AFP_ENG_19970417.0441.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_19970417.0441\n\nFARGO, North Dakota, <t1>April 17 , 1997</t1> (AFP)\n\nRed River floods set new record high in North Dakota town\n\n\n\n\nFloodwaters from the swollen Red River soared to record high levels in this upper Midwest town <t2>Thursday</t2> and kept on rising as residents raced to shore up weakening sandbag dikes.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1766", "doc_id": "XIN_ENG_20061114.0441.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nTwenty-eight brands of clothes failed inspections by the Beijing Industry and Commerce Administration (BICA), and four brands were <e1>found</e1> to <e2>contain</e2> excessive levels of toxic aromatic ammine or formaldehyde.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3779", "doc_id": "XIN_ENG_20051103.0169.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nUnder the VFA, US troops who take part in joint exercises in the Philippines are <e1>provided</e1> with special legal protection for actions taken as part of the war games but are <e2>given</e2> less protection for crimes committed while on leave.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1673", "doc_id": "XIN_ENG_20061123.0200.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe index, initiated in <t1>August 1994</t1> and readjusted on <t2>Sept.</t2>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id911", "doc_id": "XIN_ENG_20051117.0406.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nWhen local police approached <t1>May</t1>'s house in Ky Son district on <t2>the morning of Nov.</t2>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3621", "doc_id": "AFP_ENG_20061214.0704.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nFive people were injured in the shoot-out, one of them seriously, which occurred as the Palestinian premier left the Gaza border terminal where he had been <e1>blocked</e1> for several hours, a government source <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id263", "doc_id": "XIN_ENG_20061115.0456.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMartin collected 19 points and 20 rebounds in <t1>the first two games of the season</t1>, but swelling in the right knee forced him to miss last <t2>Wednesday</t2>'s game against the New York Knicks.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id5106", "doc_id": "XIN_ENG_20061113.0148.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nChina should follow a slightly more stringent monetary policy in 2007, <e1>restricting</e1> newly increased loans to three trillion yuan, <e2>said</e2> the report.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6431", "doc_id": "APW19980418.0210.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHowever , <e1>cited</e1> by District of Columbia traffic police in\nDecember for <e2>driving</e2> under the influence of alcohol, Farkas  was ordered home and retired.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7487", "doc_id": "bbc_20130322_1353.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe Israeli government <e1>admitted</e1> <e2>mistakes</e2> were made in intelligence-gathering and planning, but insisted its commandos used lethal force because activists had attacked them.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id5376", "doc_id": "AFP_ENG_20061214.0086.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe company hopes to <e1>sell</e1> 600 of the made-to-measure cars in 2006, <e2>setting</e2> a 16-year production record, and will launch its long-awaited, yet-to-be-named convertible version in January 2007.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3266", "doc_id": "wsj_1025.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nOperating revenue <e1>surged</e1> 54% in <t1>the latest period</t1> to $42.9 million from $27.7 million.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id771", "doc_id": "XIN_ENG_20061115.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"It ends the <e1>era</e1> of a literary world dominated by great masters and ushers in the age of common people,\" fellow writer Sun Yunxiao <e2>wrote</e2> in his blog, commenting on Tie's election.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7943", "doc_id": "XIN_ENG_20051116.0158.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe former five-time world cross country champion said he is looking forward to future challenges including the next London Marathon on April 23, 2006 which will feature a rematch with Hendrick Ramaala, and another encounter with Haile Gebrselassie who ran 2:06:20 in Amsterdam on <t1>October 16</t1>, the fastest time in <t2>the year</t2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id7043", "doc_id": "XIN_ENG_20051113.0236.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe <e1>attack</e1> <e2>took</e2> place at a police barricade outside Tres Vidas, where conflict between a private hotel and farming interest has sparked a series of conflict.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5750", "doc_id": "XIN_ENG_20061129.0450.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe minister <e1>said</e1> the government hopes to achieve economic growth of six percent <t1>a year</t1> and to bring down inflation to no more than five percent from the current level of eight percent.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id8019", "doc_id": "AP900815-0044.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nPresident Bush on Tuesday said\nthe United States may <e1>extend</e1> its naval quarantine to Jordan's Red\nSea port of Aqaba to <e2>shut</e2> off Iraq's last unhindered trade route.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1912", "doc_id": "AFP_ENG_20061229.0486.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Saddam Hussein was responsible for horrific, widespread human rights <e1>violations</e1>, but those acts, however brutal, cannot <e2>justify</e2> his execution, a cruel and inhuman punishment,\" said HRW International Justice Programme director Richard Dicker.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6932", "doc_id": "AFP_ENG_20061212.0392.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"Governor Mohammad Daud was about to leave and as his motorcade and bodyguards were <e1>prepared</e1> the attacker penetrated into the governor's house and tried to approach the governor,\" Mullahkil <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id4875", "doc_id": "AFP_ENG_19970413.0013.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMarket speculation <t1>Friday</t1> <e1>sent</e1> property stocks reeling, with the Stock Exchange of Singapore Property Index ending down five percent -- the largest one-day fall since the May curbs.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id1027", "doc_id": "AFP_ENG_20051206.0280.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nWhile a compromise was reached over OSCE reforms due to start next year, Russia could not <e1>agree</e1> with the US and the European Union on a joint final declaration, which must be unanimously approved by participants before the <e2>end</e2> of the conference.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id7521", "doc_id": "XIN_ENG_20051124.0170.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n24 , 2005 (Xinhua)\n\njoint statement\n\n\n\n\nA four-day conference for young political leaders in Asia and Europe concluded here <t1>Thursday</t1> with the <e1>adoption</e1> of a joint statement featuring agreements on youth issues in the two continents.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id769", "doc_id": "XIN_ENG_20061115.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAlthough Tie has <e1>turned</e1> down all media requests for interviews since the appointment this week, her position in both the literary and political arena, as well as her earlier words and deeds, might still <e2>belie</e2> her would-be agenda for the organization with more than 7,600 member writers.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7189", "doc_id": "XIN_ENG_20061120.0268.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"And I think the most important thing for the Syrian government is to look to themselves and <e1>take</e1> the actions that are necessary to help <e2>improve</e2> the situation in Iraq,\" Casey said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id6263", "doc_id": "AFP_ENG_19970421.0445.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nDespite the difficulties <e1>presented</e1> by these issues, Rockwell <e2>said</e2> the current atmosphere was overwhelmingly optimistic.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1795", "doc_id": "XIN_ENG_20061110.0046.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n10 , 2006 (Xinhua)\n\nIndonesia denies pursuing security pact with Australia\n\n\n\n\nIndonesia <t1>Friday</t1> dismissed a report that it would sign a security pact with Australia during the bilateral meeting scheduled <t2>Monday</t2> on the resort island of Lombok.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4209", "doc_id": "AFP_ENG_20061213.0712.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHe said ZANU-PF was considering bringing forward the end of the current parliamentary term from 2010 to 2008 to hold both presidential and parliamentary elections that year or stage the presidential <e1>elections</e1> in <t1>2008</t1> but allow the winner two years and hold presidential polls again two years later.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6660", "doc_id": "XIN_ENG_20051125.0186.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n25 , 2005 (Xinhua)\n\nFactbox of women's all-around world champion Chellsie Memmel\n\n\n\n\nFollowing is the factbox of American Chellsie Memmel, who <e1>won</e1> women's all-around at the 38th gyamnastics world championships here on Friday:\n\nChellsie Memmel, the United States\n\nPlace of birth: West Allis, WI, the United States\n\nDate of birth: Junuary 23, 1988\n\nHeight: 154 cm\n\nWeight: 50 kg\n\nClub: M Gymnastics in the United States\n\nCoach: Andrew Memmel, her father, and Laurie Glazer\n\nIdol: her parents\n\nA steady performer, she helped the American team win its first world team title in <t1>2003</t1> and was a co-champion on uneven bars that year.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4793", "doc_id": "AFP_ENG_19970407.0354.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nEuropean Union foreign ministers <e1>agreed</e1> Monday to <e2>go</e2> ahead with UN resolutions criticising the human rights record of Burma, Zaire, Iraq and Iran despite failing to reach accord on a similar condemnation of China.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id4639", "doc_id": "AFP_ENG_19970402.0390.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe proposed international force's role would be strictly confined to <e1>protecting</e1> the humanitarian convoys, Bonino <e2>noted</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2243", "doc_id": "XIN_ENG_20051113.0163.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nPolice officials <e1>said</e1> Dar was taken to the blast sites to identify the targets and to confirm his <e2>involvement</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5065", "doc_id": "XIN_ENG_20061110.0141.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"<t1>Today</t1>, my officers and the police are <e1>working</e1> to contend with some 200 groupings or networks, totaling over 1,600 identified individuals who are actively engaged in plotting or facilitating terrorist acts here and overseas,\" she said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id576", "doc_id": "AFP_ENG_19970411.0363.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nDutch diplomat Rienko Wilton told AFP that <t1>Friday</t1>'s <e1>demonstration</e1> was \"very peaceful.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id7065", "doc_id": "AFP_ENG_20061230.0186.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nWithin hours of Saddam's dawn hanging in Baghdad, a car bomb <e1>exploded</e1> in a fish market in the central Iraqi Shiite city of Kufa, killing at least 31 people, but it was not immediately clear whether the <e2>attack</e2> represented the first reprisal from his supporters.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6139", "doc_id": "XIN_ENG_20051105.0037.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nChinese Taipei, who beat Japan 89-78 in the preliminary round, will <e1>play</e1> the final on <t1>Sunday</t1>, while South Korea plays against China for the bronze.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id3294", "doc_id": "AFP_ENG_19970426.0106.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nSome 1,000 police were called in to Dusseldorf to <e1>monitor</e1> the <e2>demonstration</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id285", "doc_id": "XIN_ENG_20051102.0223.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe friendly exchange and cooperation between China and Japan <e1>witnessed</e1> sound progress since the two countries normalized diplomatic relations 33 years ago, Jia <e2>said</e2>, noting it is worthy to be cherished and preserved by the two sides.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id2872", "doc_id": "AFP_ENG_19970417.0042.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe ADB's annual Asian Development Outlook report <e1>said</e1> the <e2>upturn</e2> would be good news for Asia, whose export success in the sector helped establish \"from virtually nothing\" the world's newest economic miracles.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id3405", "doc_id": "AFP_ENG_20061203.0391.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe latest arrest warrant by judge Victor Montiglio is linked to the disappearance of two security guards of socialist president Salvador Allende, who committed <e1>suicide</e1> during the <t1>September 1973</t1> military coup.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id4415", "doc_id": "AFP_ENG_19970423.0174.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMore than 10,000 teachers launched their strike at the start of <t1>the month</t1> to press demands for a 200 percent pay hike and for the reinstatement of 19 teachers dismissed over a warning strike held <t2>the previous month</t2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id782", "doc_id": "XIN_ENG_20061115.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nChina's literary world has <e1>become</e1> increasingly <e2>diversified</e2> since the 1980s, leaving behind the era when the literary landscape was dominated by several great masters who were admired by an awestruck public.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id2186", "doc_id": "AFP_ENG_19970411.0067.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHe is the first indigenous bishop since Vanuatu's catholic mission <e1>arrived</e1> here <t1>more than 110 years ago</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5293", "doc_id": "XIN_ENG_20061123.0035.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nGroup leader Ron Gould said here on Wednesday that both the government and ECZ would abide by the recommendations that would reflect in their report to be <e1>handed</e1> to ECZ in <t1>two weeks time</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5034", "doc_id": "XIN_ENG_20061110.0141.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"More and more people are <e1>moving</e1> from passive sympathy towards active terrorism through being radicalized or indoctrinated by friends, families or organized training events here and overseas,\" she <e2>added</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id6008", "doc_id": "WSJ910225-0066.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We looked to the United States, we\n<e1>expected</e1> you to <e2>have</e2> the moral edge,\" says Nasser Tahboub, a\nJerusalem-born Jordanian who has an American wife and a\ndoctorate in political science from Duke University.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5970", "doc_id": "AFP_ENG_19970430.0572.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nZairean President Mobutu Sese Seko will <e1>meet</e1> with rebel leader Laurent Kabila Friday for their first face-to-face <e2>talks</e2> on board a South African ship in a deal brokered by the US mediator, UN Ambassador William Richardson.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6528", "doc_id": "XIN_ENG_20051107.0284.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMartin has promised to call an election within 30 days after the final sponsorship report is <e1>delivered</e1> on <t1>Feb.</t1>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id675", "doc_id": "wsj_0713.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn addition, the bank has an option to <e1>buy</e1> a 30.84% stake in BIP from Societe Generale after <t1>Jan.</t1>\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1216", "doc_id": "AFP_ENG_20061220.0510.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe lawmakers and most Nigerians <e1>criticised</e1> the use of hi-tech polling e-machines which store voters' thumbprints and photographs, but <e2>proved</e2> ineffective in a demonstration in October.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id1390", "doc_id": "XIN_ENG_20061123.0116.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n23 , 2006 (Xinhua)\n\nPoland announces national mourning for mine blast victims\n\n\n\n\nPolish President Lech Kaczynski <e1>announced</e1> <t1>three days</t1> of national mourning for the victims in a mine blast in southern Poland, the official PAP news agency reported on Thursday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id5718", "doc_id": "XIN_ENG_20051110.0134.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n10 , 2005 (Xinhua)\n\nSome 20,000 tons of copper to be auctioned to ease domestic supply\n\n\n\n\nThe State Reserves Bureau under the National Development and Reform Commission (NDRC) is going to <e1>auction</e1> 20,000 tons of copper reserves in a bid to ease tight supply in domestic market, China Daily reported on <t1>Thursday</t1>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1646", "doc_id": "XIN_ENG_20061102.0112.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n2 , 2006 (Xinhua)\n\nEast China province reinstates bird flu monitoring\n\n\n\n\nShandong Province, east China, has <e1>reinstated</e1> round-the-clock bird flu monitoring to <e2>prevent</e2> possible outbreaks as more and more migratory birds fly over the province on their way further south.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6383", "doc_id": "XIN_ENG_20051117.0202.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n17 , 2005 (Xinhua)\n\ninvestigation\n\n\n\n\nThe chemical plant <e1>blasts</e1> which left five workers dead and one missing on Sunday in Jilin Province were caused by the incorrect operation by a worker, <e2>said</e2> investigators on Thursday.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id324", "doc_id": "AFP_ENG_19970421.0360.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nA Western Australian state member of parliament, John Halden, has also been <e1>charged</e1> with one count of perjury arising from the royal commission, the report <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id1504", "doc_id": "AFP_ENG_20061228.0078.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nErshad told reporters late <t1>Wednesday</t1> that his party and its allies, including the main opposition Awami League, may again decide to boycott <t2>next month</t2>'s polls if he is barred from contesting elections.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id5040", "doc_id": "XIN_ENG_20061110.0141.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"It is the youth who are being actively targeted, <e1>groomed</e1>, radicalized and set on a path that frighteningly quickly could end in their involvement in mass murder of their fellow UK citizens,\" she <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7217", "doc_id": "XIN_ENG_20061121.0286.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nIn a statement <e1>issued</e1> in Nairobi, the organization said the registrations which will continue till next February, will <e2>allow</e2> the first rounds of organized returns to communities in southern Sudan, most likely next year.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id1036", "doc_id": "AFP_ENG_20051206.0280.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"The Russians may object to it, in which case the language (declaration) may well be <e1>used</e1> by the chairman in office,\" Sawers said, meaning the conference chairman, Slovenia's Foreign Minister Dimitrij Rupel, would <e2>issue</e2> a final statement if a consensus cannot be reached over the document.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id301", "doc_id": "AFP_ENG_20061221.0099.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061221.0099\n\nLONDON, <t1>Dec 21 , 2006</t1>, 2006\n\nEuropean shares fall at open: London FTSE down 0.16 percent\n\n\n\n\nEuropean stock markets <e1>declined</e1> in early trading on Thursday, with London's FTSE 100 index of leading shares losing 0.16 percent to reach 6,188.70 points.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id383", "doc_id": "AFP_ENG_20051215.0694.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe legislation <e1>gives</e1> police powers to \"<e2>lock</e2> down\" suburbs, close pubs, enforce alcohol bans and confiscate cars.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id5574", "doc_id": "AFP_ENG_20061225.0221.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAFP_ENG_20061225.0221\n\nCAIRO, <t1>Dec 25 , 2006</t1>, 2006\n\nEgypt Brotherhood members accused of money laundering\n\n\n\n\nSix members of the Muslim Brotherhood in Egypt were charged on Monday with money laundering and remanded for <t2>two weeks days</t2>, a security source said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id4624", "doc_id": "AFP_ENG_19970402.0390.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBonino <e1>said</e1> EU food and medical aid could be sent to Albania quickly <e2>using</e2> logistical structures established in the former Yugoslavia.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id8079", "doc_id": "AP900815-0044.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nBush <e1>told</e1> a news conference on <t1>Tuesday</t1> that the naval barricade\nnow in force might be extended to Jordan's Aqaba ``if it is a hole\nthrough which commerce flows'' in and out of Iraq.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id1705", "doc_id": "AFP_ENG_19970411.0145.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n<t1>April 11 , 1997</t1> (AFP)\n\nSeoul shares close 0.9 percent higher\n\n\n\n\nShare prices rebounded 0.9 percent on the Korea Stock Exchange on <t2>Friday</t2>, with buying interest emerging on undervalued stocks after a short correction, dealers said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}
{"id": "id3574", "doc_id": "AFP_ENG_19970418.0499.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"I have no doubt at all after this meeting ... that the US government will <e1>defend</e1> Hong Kong's freedoms,\" Lee <e2>said</e2> following his 25-minute talks with Clinton and Vice President Al Gore.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3619", "doc_id": "AFP_ENG_20061214.0704.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nHamas gunmen went on the rampage waiting for Haniya to cross back into the strip after being <e1>detained</e1> by an Israeli closure, an AFP correspondent <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id7304", "doc_id": "AFP_ENG_20061213.0676.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nUnder the new law, the federal government <e1>banned</e1> semi-automatics and pump-action rifles and shotguns and <e2>offered</e2> to buy these weapons from gunowners, using a levy on income tax.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id5492", "doc_id": "AFP_ENG_19970408.0377.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nThe <e1>violence</e1> began when two Israeli settlers killed a Palestinian with submachinegun fire in Hebron's central market after they said the man attacked them with tear gas spray, police <e2>said</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id3609", "doc_id": "AFP_ENG_20061214.0704.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n\"We <e1>deplore</e1> the events that are taking place in Rafah,\" he <e2>added</e2>.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "BEFORE"}
{"id": "id3282", "doc_id": "AFP_ENG_19970426.0106.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nAt least 10,000 had <e1>arrived</e1> <t1>early Saturday morning</t1>, police said.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id749", "doc_id": "XIN_ENG_20061115.0007.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\n<e1>Born</e1> in <t1>1957</t1>, Tie was elected president of the association on Sunday at the seventh congress of the association in Beijing, becoming China's top literary official.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id1077", "doc_id": "XIN_ENG_20061124.0262.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nEarlier in the day, Doha-based al-Jazeera English channel reported that armed clashes <e1>erupted</e1> on <t1>Friday evening</t1> between U.S.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "OTHER"}
{"id": "id6360", "doc_id": "AFP_ENG_20061219.0267.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nEverton had called on Mourinho to retract his remarks but the lack of a response from Stamford Bridge <e1>prompted</e1> them to make things official with spokesman Ian Ross confirming: \"We will now make a formal complaint to the FA over Jose Mourinho's <e2>remarks</e2> regarding Andrew Johnson.\"\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "AFTER"}
{"id": "id6367", "doc_id": "AFP_ENG_20061219.0267.tml", "text": "You are a system for temporal relation classification.\nInput:\nA sentence with two marked spans:\n<e1>/<e2> = events, <t1>/<t2> = times.\nRules:\n- FIRST = <e1> or <t1>\n- SECOND = <e2> or <t2>\n- Relation is from FIRST with respect to SECOND\n- Do NOT flip direction\n\nLabels:\nBEFORE = FIRST earlier than SECOND\nAFTER = FIRST later than SECOND\nNONE = no clear relation\nOTHER = overlap or non-order relation\n\nSentence:\nMourinho was enraged by an incident in which Johnson <e1>jumped</e1> over Chelsea goalkeeper Hilario as they both <e2>went</e2> for the ball and tumbled after landing.\n\nOutput EXACTLY one label: BEFORE, AFTER, NONE, or OTHER.\nDO NOT output anything else.", "label": "NONE"}