| { |
| "paper_id": "J13-3001", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T02:17:59.800460Z" |
| }, |
| "title": "What Is a Paraphrase?", |
| "authors": [ |
| { |
| "first": "Rahul", |
| "middle": [], |
| "last": "Bhagat", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "Eduard", |
| "middle": [], |
| "last": "Hovy", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "hovy@isi.edu" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "Paraphrases are sentences or phrases that convey the same meaning using different wording. Although the logical definition of paraphrases requires strict semantic equivalence, linguistics accepts a broader, approximate, equivalence-thereby allowing far more examples of \"quasiparaphrase.\" But approximate equivalence is hard to define. Thus, the phenomenon of paraphrases, as understood in linguistics, is difficult to characterize. In this article, we list a set of 25 operations that generate quasi-paraphrases. We then empirically validate the scope and accuracy of this list by manually analyzing random samples of two publicly available paraphrase corpora. We provide the distribution of naturally occurring quasi-paraphrases in English text.", |
| "pdf_parse": { |
| "paper_id": "J13-3001", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "Paraphrases are sentences or phrases that convey the same meaning using different wording. Although the logical definition of paraphrases requires strict semantic equivalence, linguistics accepts a broader, approximate, equivalence-thereby allowing far more examples of \"quasiparaphrase.\" But approximate equivalence is hard to define. Thus, the phenomenon of paraphrases, as understood in linguistics, is difficult to characterize. In this article, we list a set of 25 operations that generate quasi-paraphrases. We then empirically validate the scope and accuracy of this list by manually analyzing random samples of two publicly available paraphrase corpora. We provide the distribution of naturally occurring quasi-paraphrases in English text.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "Sentences or phrases that convey the same meaning using different wording are called paraphrases. For example, consider sentences (1) and (2):", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "(1)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "The school said that their buses seat 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "(2) The school said that their buses accommodate 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "Paraphrases are of interest for many current NLP tasks, including textual entailment, machine reading, question answering, information extraction, and machine translation.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "Whenever the text contains multiple ways of saying \"the same thing,\" but the application requires the same treatment of those various alternatives, an automated paraphrase recognition mechanism would be useful. One reason why paraphrase recognition systems have been difficult to build is because paraphrases are hard to define. Although the strict interpretation of the term \"paraphrase\" is quite narrow because it requires exactly identical meaning, in linguistics literature paraphrases are most often characterized by an approximate equivalence of meaning across sentences or phrases. De Beaugrande and Dressler (1981, page 50) define paraphrases as \"approximate conceptual equivalence among outwardly different material.\" Hirst (2003, slide 9) defines paraphrases as \"talk(ing) about the same situation in a different way.\" He argues that paraphrases aren't fully synonymous: There are pragmatic differences in paraphrases, namely, difference of evaluation, connotation, viewpoint, and so forth. According to Mel'cuk (2012, page 7) \"An approximate synonymy of sentences is considered as sufficient for them to be produced from the same SemS.\" He further adds that approximate paraphrases include implications (not in the logical sense, but in the everyday sense). Taking an extreme view, Clark (1992, page 172) rejects the idea of absolute synonymy by saying \"Every two forms (in language) contrast in meaning.\" Overall, there is a large body of work in the linguistics literature that argues that paraphrases are not restricted to strict synonymy.", |
| "cite_spans": [ |
| { |
| "start": 592, |
| "end": 631, |
| "text": "Beaugrande and Dressler (1981, page 50)", |
| "ref_id": null |
| }, |
| { |
| "start": 727, |
| "end": 748, |
| "text": "Hirst (2003, slide 9)", |
| "ref_id": null |
| }, |
| { |
| "start": 1293, |
| "end": 1315, |
| "text": "Clark (1992, page 172)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "In this article, we take a broad view of paraphrases. To avoid the conflict between the notion of strict paraphrases as understood in logic and the broad notion in linguistics, we use the term quasi-paraphrases to refer to the paraphrases that we deal with here. In the context of this article, the term \"paraphrases\" (even without the prefix \"quasi\") means \"quasi-paraphrases.\" We define quasi-paraphrases as 'sentences or phrases that convey approximately the same meaning using different words.' We ignore the fine grained distinctions of meaning between sentences and phrases, introduced due to the speaker's evaluation of the situation, connotation of the terms used, change of modality, and so on. For example, consider sentences (3) and (4).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "( 3)The school said that their buses seat 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "The school said that their buses cram in 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "Here, seat and cram in are not synonymous: They carry different evaluations by the speaker about the same situation. We, however, consider sentences (3) and (4) to be (quasi) paraphrases. Similarly, consider sentences (5) and (6).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "The school said that their buses seat 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(5)", |
| "sec_num": null |
| }, |
| { |
| "text": "The school is saying that their buses might accommodate 40 students each.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(5)", |
| "sec_num": null |
| }, |
| { |
| "text": "Here, said and is saying have different tenses. Also, might accommodate and seat are not synonymous, due to the modal verb might. We consider sentences (5) and (6) to be quasi-paraphrases, however. Note that this article focuses on defining quasi-paraphrases. It does not provide direct implementation/application results of using them. We believe, however, that this work will allow computation-oriented researchers to focus their future work more effectively on a subset of paraphrase types without concern for missing important material, and it will provide linguistics-oriented researchers with a blueprint of the overall distribution of the types of paraphrase.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(5)", |
| "sec_num": null |
| }, |
| { |
| "text": "Although approximate equivalence is hard to characterize, it is not a completely unstructured phenomenon. By studying various existing paraphrase theories-Mel'cuk (2012), Harris (1981) , Honeck (1971)-and through an analysis of paraphrases obtained from two different corpora, we have discovered that one can identify a set of 25 classes of quasi-paraphrases, with each class having its own specific way of relaxing the requirement of strict semantic equivalence. In this section, we define and describe these classes.", |
| "cite_spans": [ |
| { |
| "start": 171, |
| "end": 184, |
| "text": "Harris (1981)", |
| "ref_id": "BIBREF5" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "The classes described here categorize quasi-paraphrases from the lexical perspective. The lexical perspective defines paraphrases in terms of the kinds of lexical changes that can take place in a sentence/phrase resulting in the generation of its paraphrases.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "1. Synonym substitution: Replacing a word/phrase by a synonymous word/phrase, in the appropriate context, results in a paraphrase of the original sentence/phrase. This category covers the special case of genitives, where the clitic 's is replaced by other genitive indicators like of, of the, and so forth. This category also covers near-synonymy, that is, it allows for changes in evaluation, connotation, and so on, of words or phrases between paraphrases. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "(a)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "Google bought YouTube. \u21d4 Google acquired YouTube.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "(b) Chris is slim. \u21d4 Chris is slender. \u21d4 Chris is skinny.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "2. Antonym substitution: Replacing a word/phrase by its antonym accompanied by a negation or by negating some other word, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "(a) Pat ate. \u21d4 Pat did not starve.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Paraphrasing Phenomena Classified", |
| "sec_num": "2." |
| }, |
| { |
| "text": "Replacing a word/phrase with its converse and inverting the relationship between the constituents of a sentence/phrase, in the appropriate context, results in a paraphrase of the original sentence/phrase, presenting the situation from the converse perspective. This substitution may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Converse substitution:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "(a) Google bought YouTube. \u21d4 YouTube was sold to Google.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Converse substitution:", |
| "sec_num": "3." |
| }, |
| { |
| "text": "Changing a verb from its active to passive form and vice versa results in a paraphrase of the original sentence/phrase. This change may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. This often generates the most strictly meaning-preserving paraphrase. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of voice:", |
| "sec_num": "4." |
| }, |
| { |
| "text": "(a) Pat loves Chris. \u21d4 Chris is loved by Pat.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of voice:", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Changing the grammatical person of a referenced object results in a paraphrase of the original sentence/phrase. This change may be accompanied by the addition/deletion of appropriate function words. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "(a) Pat said, \"I like football.\" \u21d4 Pat said that he liked football.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "6. Pronoun/Co-referent substitution: Replacing a pronoun by the noun phrase it co-refers with results in a paraphrase of the original sentence/phrase. This also often generates the most strictly meaning-preserving paraphrase. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "(a) Pat likes Chris, because she is smart. \u21d4 Pat likes Chris, because Chris is smart.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "7. Repetition/Ellipsis: Ellipsis or elliptical construction results in a paraphrase of the original sentence/phrase. Similarly, this often generates the most strictly meaningpreserving paraphrase. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "(a) Pat can run fast and Chris can run fast, too. \u21d4 Pat can run fast and Chris can, too.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of person:", |
| "sec_num": "5." |
| }, |
| { |
| "text": "Changing the function words in a sentence/phrase without affecting its semantics, in the appropriate context, results in a paraphrase of the original sentence/phrase. This can involve replacing a light verb by another light verb, replacing a light verb by copula, replacing certain prepositions with other prepositions, replacing a determiner by another determiner, replacing a determiner by a preposition and vice versa, and addition/removal of a preposition and/or a determiner. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Function word variations:", |
| "sec_num": "8." |
| }, |
| { |
| "text": "(a)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Function word variations:", |
| "sec_num": "8." |
| }, |
| { |
| "text": "Results of the competition have been declared. \u21d4 Results for the competition have been declared.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Function word variations:", |
| "sec_num": "8." |
| }, |
| { |
| "text": "Pat showed a nice demo. \u21d4 Pat's demo was nice.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "9. Actor/Action substitution: Replacing the name of an action by a word/phrase denoting the person doing the action (actor) and vice versa, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "(a) I dislike rash drivers. \u21d4 I dislike rash driving.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "10. Verb/\"Semantic-role noun\" substitution: Replacing a verb by a noun corresponding to the agent of the action or the patient of the action or the instrument used for the action or the medium used for the action, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. 11. Manipulator/Device substitution: Replacing the name of a device by a word/ phrase denoting the person using the device (manipulator) and vice versa, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "(a)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "The pilot took off despite the stormy weather. \u21d4 The plane took off despite the stormy weather.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "12. General/Specific substitution: Replacing a word/phrase by a more general or more specific word/phrase, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words. Hypernym/hyponym substitution is a part of this category. This often generates a quasi-paraphrase. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "(a) I dislike rash drivers. \u21d4 I dislike rash motorists.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "Pat is flying in this weekend. \u21d4 Pat is flying in this Saturday.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "13. Metaphor substitution: Replacing a noun by its standard metaphorical use and vice versa, in the appropriate context, results in a paraphrase of the original sentence/ phrase. This substitution may be accompanied by the addition/deletion of appropriate function words. Example: 18. Noun/Adjective conversion: Replacing a verb by its corresponding adjective form and vice versa, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "(a) I'll fly by the end of June. \u21d4 I'll fly late June.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "19. Verb-preposition/Noun substitution: Replacing a verb and a preposition denoting location by a noun denoting the location and vice versa, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "(a) The finalists will play in Giants stadium. \u21d4 Giants stadium will be the playground for the finalists.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "Changing the tense of a verb, in the appropriate context, results in a paraphrase of the original sentence/phrase. This change may be accompanied by the addition/deletion of appropriate function words. This often generates a quasiparaphrase, although it might be semantically less accurate than many other quasiparaphrases. Example: 22. Change of modality: Addition/deletion of a modal or substitution of one modal by another, in the appropriate context, results in a paraphrase of the original sentence/phrase. This change may be accompanied by the addition/deletion of appropriate function words. This often generates a quasi-paraphrase, although it might be semantically less accurate than many other quasi-paraphrases. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of tense:", |
| "sec_num": "20." |
| }, |
| { |
| "text": "(a) Google must buy YouTube. \u21d4 Google bought YouTube.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Change of tense:", |
| "sec_num": "20." |
| }, |
| { |
| "text": "The government wants to boost the economy. \u21d4 The government hopes to boost the economy.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "Replacing a word/phrase denoting an action, event, and so forth, by a word/phrase denoting its possible future effect, in the appropriate context, results in a paraphrase of the original sentence/phrase. This may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. This often generates a quasi-paraphrase. Example: 24. Approximate numerical equivalences: Replacing a numerical expression (a word/ phrase denoting a number, often with a unit) by an approximately equivalent numerical expression (even perhaps with change of unit), in the appropriate context, results in a paraphrase of the original sentence/phrase. This often generates a quasiparaphrase. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Semantic implication:", |
| "sec_num": "23." |
| }, |
| { |
| "text": "(a) At least 23 U.S. soldiers were killed in Iraq last month. \u21d4 About 25 U.S. soldiers were killed in Iraq last month.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Semantic implication:", |
| "sec_num": "23." |
| }, |
| { |
| "text": "Disneyland is 32 miles from here. \u21d4 Disneyland is around 30 minutes from here.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "Replacing a word/phrase by another word/phrase based on extra-linguistic (world) knowledge, in the appropriate context, results in a paraphrase of the original sentence/phrase. This may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. This often generates a quasiparaphrase, although in some cases preserves meaning exactly. Example:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "External knowledge:", |
| "sec_num": "25." |
| }, |
| { |
| "text": "(a)", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "External knowledge:", |
| "sec_num": "25." |
| }, |
| { |
| "text": "We must work hard to win this election. \u21d4 The Democrats must work hard to win this election.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "External knowledge:", |
| "sec_num": "25." |
| }, |
| { |
| "text": "The government declared victory in Iraq. \u21d4 Bush declared victory in Iraq.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "(b)", |
| "sec_num": null |
| }, |
| { |
| "text": "In Section 2, we presented a list of lexical changes that define quasi-paraphrases. In this section, we seek to validate the scope and accuracy of this list. Our analysis uses two criteria:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Analysis of Paraphrases", |
| "sec_num": "3." |
| }, |
| { |
| "text": "1. Distribution: What is the distribution of each of these lexical changes in a paraphrase corpus?", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Analysis of Paraphrases", |
| "sec_num": "3." |
| }, |
| { |
| "text": "2. Human judgment: If one uses each of the lexical changes, on applicable sentences, how often do each of these changes generate acceptable quasi-paraphrases?", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Analysis of Paraphrases", |
| "sec_num": "3." |
| }, |
| { |
| "text": "We used the following procedure to measure the distribution of the lexical changes:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Distribution", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "1. We downloaded paraphrases from two publicly available data sets containing sentence-level paraphrases: the Multiple-Translations Corpus (MTC) (Huang, Graff, and Doddington 2002) and the Microsoft Research (MSR) paraphrase corpus (Dolan, Quirk, and Brockett 2004) . The paraphrase pairs come with their equivalent parts manually aligned (Cohn, Callison-Burch, and Lapata 2008) .", |
| "cite_spans": [ |
| { |
| "start": 145, |
| "end": 180, |
| "text": "(Huang, Graff, and Doddington 2002)", |
| "ref_id": "BIBREF7" |
| }, |
| { |
| "start": 232, |
| "end": 265, |
| "text": "(Dolan, Quirk, and Brockett 2004)", |
| "ref_id": "BIBREF3" |
| }, |
| { |
| "start": 339, |
| "end": 378, |
| "text": "(Cohn, Callison-Burch, and Lapata 2008)", |
| "ref_id": "BIBREF1" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Distribution", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "2. We selected 30 sentence-level paraphrase pairs from each of these corpora at random and extracted the corresponding aligned and unaligned phrases. 1 This resulted in 210 phrase pairs for the MTC corpus and 145 phrase pairs for the MSR corpus.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Distribution", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "3. We labeled each of the phrase pairs with the appropriate lexical changes defined in Section 2. If any phrase pair could not be labeled by a lexical change from Section 2, we labeled it as unknown.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Distribution", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "4. We finally calculated the distribution of each label (lexical change), over all the labels, for each corpus. Table 1 shows the percentage distribution of the lexical changes in the MTC (column 3) and MSR corpora (column 4).", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 112, |
| "end": 119, |
| "text": "Table 1", |
| "ref_id": "TABREF1" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Distribution", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "In this section, we explain the procedure we used to obtain the human judgments of the changes that define paraphrases from the lexical perspective:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "1. We randomly selected two words or phrases from publicly available resources (depending on the lexical change) for each of the lexical operations from Section 2 (except external knowledge). For example, to obtain words for synonym substitution, we used WordNet (Fellbaum 1998 ) (and selected a word, say buy); to obtain implication rules for semantic implication, we used the DIRT resource (Lin and Pantel 2001) ; and so on. 2. For each selected word or phrase, we obtained five random sentences from the Gigaword corpus. These sentences were manually checked to make sure that they contained the intended sense of the word or phrase. This gave us a total of 10 sentences for each phenomenon. For example, for the word buy, one of the selected sentences might be:", |
| "cite_spans": [ |
| { |
| "start": 263, |
| "end": 277, |
| "text": "(Fellbaum 1998", |
| "ref_id": "BIBREF4" |
| }, |
| { |
| "start": 392, |
| "end": 413, |
| "text": "(Lin and Pantel 2001)", |
| "ref_id": "BIBREF8" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "(a) They want to buy a house.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "3. For each sentence selected in step 2, we applied the corresponding lexical changes to the word or phrase selected in step 1 to generate a potential paraphrase. 2 For example, we might apply synonym substitution to sentence (a) and replace the word buy with its WordNet synonym purchase. This will result in the following sentence:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "(b) They want to purchase a house.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "4. For the phenomenon of external knowledge, we randomly sampled a total of 10 sentence pairs from the MTC and MSR corpora, such that the pairs were paraphrases based on external knowledge.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "5. We gave the sentence pairs to two annotators and asked them to annotate them as either paraphrases or non-paraphrases. For example, the annotator might be given the sentence pair (a) and (b) and she/he might annotate this pair as paraphrases.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "6. We used the annotations from each of the annotators to calculate the precision percentage for each lexical change. The final precision score was calculated as the average of the precision scores obtained from the two annotations. Table 1 shows the percentage precision (column 5) of lexical changes in this test corpus.", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 233, |
| "end": 240, |
| "text": "Table 1", |
| "ref_id": "TABREF1" |
| } |
| ], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "7. We finally calculated the kappa statistic (Siegal and Castellan Jr. 1988) to measure the inter-annotator agreement. A kappa score of \u03ba = 0.66 was obtained on the annotation task.", |
| "cite_spans": [ |
| { |
| "start": 45, |
| "end": 76, |
| "text": "(Siegal and Castellan Jr. 1988)", |
| "ref_id": "BIBREF11" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Human Judgment", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "A definition of what phenomena constitute paraphrases and what do not has been a problem in the past. Whereas some people have used a very narrow interpretation of paraphrases-paraphrases must be exactly logically equivalent-others have taken broader perspectives that consider even semantic implications to be acceptable paraphrases. To the best of our knowledge, outside of specific language interpretation frameworks (like Meaning Text Theory [Mel'cuk 1996]) , no one has tried to create a general, exhaustive list of the transformations that define paraphrases. In this article we provide such a list. We have also tried to empirically quantify the distribution and accuracy of the list. It is notable that certain types of quasi-paraphrases dominate whereas others are very rare. We also observed, however, that the dominating transformations vary based on the type of paraphrase corpus used, thus indicating the variety of behavior exhibited by the paraphrases. Based on the large variety of possible transformations that can generate paraphrases, its seems likely that the kinds of paraphrases that are deemed useful would depend on the application at hand. This might motivate the creation of application-specific lists of the kinds of allowable paraphrases and the development of automatic methods to distinguish the different kinds of paraphrases.", |
| "cite_spans": [ |
| { |
| "start": 446, |
| "end": 461, |
| "text": "[Mel'cuk 1996])", |
| "ref_id": "BIBREF9" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "We assume that any unaligned phrase is paired with a null phrase and we discard it prior to the analysis.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "The words in the new sentence were allowed to be reordered (permuted) if needed and only function words (and no content words) were allowed to be added to the new sentence.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "The authors wish to thank Jerry Hobbs and anonymous reviewers for valuable comments and feedback.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Acknowledgments", |
| "sec_num": null |
| } |
| ], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "Conventionality and contrasts: Pragmatic principles with lexical consequences", |
| "authors": [ |
| { |
| "first": "E", |
| "middle": [ |
| "V" |
| ], |
| "last": "Clark", |
| "suffix": "" |
| } |
| ], |
| "year": 1992, |
| "venue": "Frame, Fields, and Contrasts: New Essays in Semantic Lexical Organization", |
| "volume": "", |
| "issue": "", |
| "pages": "171--188", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Clark, E. V. 1992. Conventionality and contrasts: Pragmatic principles with lexical consequences. In Andrienne Lehrer and Eva Feder Kittay, editors, Frame, Fields, and Contrasts: New Essays in Semantic Lexical Organization. Lawrence Erlbaum Associates, Hillsdale, NJ, pages 171-188.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Constructing corpora for the development and evaluation of paraphrase systems", |
| "authors": [ |
| { |
| "first": "T", |
| "middle": [], |
| "last": "Cohn", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Callison-Burch", |
| "suffix": "" |
| }, |
| { |
| "first": "M", |
| "middle": [], |
| "last": "Lapata", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "Computational Linguistics", |
| "volume": "34", |
| "issue": "4", |
| "pages": "597--614", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Cohn, T., C. Callison-Burch, and M. Lapata. 2008. Constructing corpora for the development and evaluation of paraphrase systems. Computational Linguistics, 34(4):597-614.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Introduction to Text Linguistics", |
| "authors": [ |
| { |
| "first": "R", |
| "middle": [], |
| "last": "De Beaugrande", |
| "suffix": "" |
| }, |
| { |
| "first": "W", |
| "middle": [ |
| "V" |
| ], |
| "last": "Dressler", |
| "suffix": "" |
| } |
| ], |
| "year": 1981, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "De Beaugrande, R. and W. V. Dressler. 1981. Introduction to Text Linguistics. Longman, New York, NY.", |
| "links": null |
| }, |
| "BIBREF3": { |
| "ref_id": "b3", |
| "title": "Unsupervised construction of large paraphrase corpora: Exploiting massively parallel news sources", |
| "authors": [ |
| { |
| "first": "B", |
| "middle": [], |
| "last": "Dolan", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Quirk", |
| "suffix": "" |
| }, |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Brockett", |
| "suffix": "" |
| } |
| ], |
| "year": 2004, |
| "venue": "Proceedings of the Conference on Computational Linguistics (COLING)", |
| "volume": "", |
| "issue": "", |
| "pages": "350--357", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Dolan, B., C. Quirk, and C. Brockett. 2004. Unsupervised construction of large paraphrase corpora: Exploiting massively parallel news sources. In Proceedings of the Conference on Computational Linguistics (COLING), pages 350-357, Geneva.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "An Electronic Lexical Database", |
| "authors": [ |
| { |
| "first": "C", |
| "middle": [], |
| "last": "Fellbaum", |
| "suffix": "" |
| } |
| ], |
| "year": 1998, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Fellbaum, C. 1998. An Electronic Lexical Database. MIT Press, Cambridge, MA.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Co-occurence and transformation in linguistic structure", |
| "authors": [ |
| { |
| "first": "Z", |
| "middle": [], |
| "last": "Harris", |
| "suffix": "" |
| } |
| ], |
| "year": 1981, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Harris, Z. 1981. Co-occurence and transformation in linguistic structure. In Henry Hiz, editor, Papers on Syntax. D. Reidel Publishing Co., Dordrecht, pages 143-210. First published in 1957.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "Paraphrasing paraphrased. Invited talk at the ACL International Workshop on Paraphrasing, Sapporo. Honeck, Richard P. 1971. A study of paraphrases", |
| "authors": [ |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Hirst", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "Journal of Verbal Learning and Verbal Behavior", |
| "volume": "10", |
| "issue": "4", |
| "pages": "367--381", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Hirst, G. 2003. Paraphrasing paraphrased. Invited talk at the ACL International Workshop on Paraphrasing, Sapporo. Honeck, Richard P. 1971. A study of paraphrases. Journal of Verbal Learning and Verbal Behavior, 10(4):367-381.", |
| "links": null |
| }, |
| "BIBREF7": { |
| "ref_id": "b7", |
| "title": "Multiple-translation Chinese corpus", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Huang", |
| "suffix": "" |
| }, |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Graff", |
| "suffix": "" |
| }, |
| { |
| "first": "G", |
| "middle": [], |
| "last": "Doddington", |
| "suffix": "" |
| } |
| ], |
| "year": 2002, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Huang, S., D. Graff, and G. Doddington. 2002. Multiple-translation Chinese corpus. Linguistic Data Consortium, Philadelphia, PA.", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "Dirt: Discovery of inference rules from text", |
| "authors": [ |
| { |
| "first": "D", |
| "middle": [], |
| "last": "Lin", |
| "suffix": "" |
| }, |
| { |
| "first": "P", |
| "middle": [], |
| "last": "Pantel", |
| "suffix": "" |
| } |
| ], |
| "year": 2001, |
| "venue": "ACM SIGKDD International Conference on Knowledge Discovery and Data Mining", |
| "volume": "", |
| "issue": "", |
| "pages": "323--328", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lin, D. and P. Pantel. 2001. Dirt: Discovery of inference rules from text. In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 323-328, San Francisco, CA.", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "Lexical functions: A tool for description of lexical relations in a lexicon", |
| "authors": [ |
| { |
| "first": "I", |
| "middle": [], |
| "last": "Mel'cuk", |
| "suffix": "" |
| } |
| ], |
| "year": 1996, |
| "venue": "Lexical Functions in Lexicography and Natural Language Processing", |
| "volume": "", |
| "issue": "", |
| "pages": "37--102", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mel'cuk, I. 1996. Lexical functions: A tool for description of lexical relations in a lexicon. In Leo Wanner, editor, Lexical Functions in Lexicography and Natural Language Processing. John Benjamins Publishing Co., Philadelphia, PA, pages 37-102.", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Semantics: From Meaning to Text", |
| "authors": [ |
| { |
| "first": "I", |
| "middle": [], |
| "last": "Mel'cuk", |
| "suffix": "" |
| } |
| ], |
| "year": 2012, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mel'cuk, I., 2012. Semantics: From Meaning to Text. John Benjamins Publishing Co., Philadelphia, PA.", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Nonparametric Statistics for the Behavioral Sciences", |
| "authors": [ |
| { |
| "first": "S", |
| "middle": [], |
| "last": "Siegal", |
| "suffix": "" |
| }, |
| { |
| "first": "N", |
| "middle": [ |
| "J" |
| ], |
| "last": "Castellan", |
| "suffix": "" |
| }, |
| { |
| "first": "Jr", |
| "middle": [], |
| "last": "", |
| "suffix": "" |
| } |
| ], |
| "year": 1988, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Siegal, S. and N. J. Castellan, Jr. 1988. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, Columbus, OH.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "num": null, |
| "text": "Pat teaches Chris. \u21d4 Pat is Chris's teacher. (b) Pat teaches Chris. \u21d4 Chris is Pat's student. (c) Pat tiled his bathroom floor. \u21d4 Pat installed tiles on his bathroom floor.", |
| "uris": null, |
| "type_str": "figure" |
| }, |
| "FIGREF1": { |
| "num": null, |
| "text": "loved Chris. \u21d4 Pat loves Chris.21. Change of aspect:Changing the aspect of a verb, in the appropriate context, results in a paraphrase of the original sentence/phrase. This change may be accompanied by the addition/deletion of appropriate function words. Example:(a)Pat is flying in today. \u21d4 Pat flies in today.", |
| "uris": null, |
| "type_str": "figure" |
| }, |
| "FIGREF2": { |
| "num": null, |
| "text": "is in talks to buy YouTube. \u21d4 Google bought YouTube.(b)The Marines are fighting the terrorists. \u21d4 The Marines are eliminating the terrorists.", |
| "uris": null, |
| "type_str": "figure" |
| }, |
| "TABREF0": { |
| "content": "<table/>", |
| "num": null, |
| "html": null, |
| "text": "I had to drive through fog today. \u21d4 I had to drive through a wall of fog today. Replacing a verb by the corresponding adjective form and vice versa, in the appropriate context, results in a paraphrase of the original sentence/phrase. This substitution may be accompanied by the addition/deletion of appropriate function words and sentence restructuring. Example:(a)Pat loves Chris. \u21d4 Chris is lovable to Pat.", |
| "type_str": "table" |
| }, |
| "TABREF1": { |
| "content": "<table><tr><td>#</td><td>Category</td><td colspan=\"3\">% Distribution MTC % Distribution MSR % Precision</td></tr><tr><td colspan=\"2\">1. Synonym substitution</td><td>37</td><td>19</td><td>95</td></tr><tr><td colspan=\"2\">2. Antonym substitution</td><td>0</td><td>0</td><td>65</td></tr><tr><td colspan=\"2\">3. Converse substitution</td><td>1</td><td>0</td><td>75</td></tr><tr><td colspan=\"2\">4. Change of voice</td><td>1</td><td>1</td><td>85</td></tr><tr><td colspan=\"2\">5. Change of person</td><td>0</td><td>1</td><td>80</td></tr><tr><td colspan=\"2\">6. Pronoun/Co-referent</td><td>1</td><td>1</td><td>7 0</td></tr><tr><td colspan=\"2\">substitution</td><td/><td/></tr><tr><td colspan=\"2\">7. Repetition/Ellipsis</td><td>4</td><td>4</td><td>100</td></tr><tr><td colspan=\"2\">8. Function word variations</td><td>37</td><td>30</td><td>85</td></tr><tr><td colspan=\"2\">9. Actor/Action substitution</td><td>0</td><td>0</td><td>75</td></tr><tr><td colspan=\"2\">10. Verb/\"Semantic-role noun\"</td><td>1</td><td>0</td><td>6 0</td></tr><tr><td colspan=\"2\">substitution</td><td/><td/></tr><tr><td colspan=\"2\">11. Manipulator/Device substitution</td><td>0</td><td>0</td><td>30</td></tr><tr><td colspan=\"2\">12. General/Specific substitution</td><td>4</td><td>3</td><td>80</td></tr><tr><td colspan=\"2\">13. Metaphor substitution</td><td>0</td><td>1</td><td>60</td></tr><tr><td colspan=\"2\">14. Part/Whole substitution</td><td>0</td><td>0</td><td>65</td></tr><tr><td colspan=\"2\">15. Verb/Noun conversion</td><td>2</td><td>3</td><td>100</td></tr><tr><td colspan=\"2\">16. Verb/Adjective conversion</td><td>1</td><td>0</td><td>55</td></tr><tr><td colspan=\"2\">17. Verb/Adverb conversion</td><td>0</td><td>0</td><td>65</td></tr><tr><td colspan=\"2\">18. Noun/Adjective conversion</td><td>0</td><td>0</td><td>80</td></tr><tr><td colspan=\"2\">19. Verb-preposition/</td><td>0</td><td>0</td><td>6 5</td></tr><tr><td colspan=\"2\">Noun substitution</td><td/><td/></tr><tr><td colspan=\"2\">20. Change of tense</td><td>4</td><td>1</td><td>70</td></tr><tr><td colspan=\"2\">21. Change of aspect</td><td>1</td><td>0</td><td>95</td></tr><tr><td colspan=\"2\">22. Change of modality</td><td>1</td><td>0</td><td>80</td></tr><tr><td colspan=\"2\">23. Semantic implication</td><td>1</td><td>4</td><td>70</td></tr><tr><td colspan=\"2\">24. Approximate numerical</td><td>0</td><td>2</td><td>9 5</td></tr><tr><td colspan=\"2\">equivalences</td><td/><td/></tr><tr><td colspan=\"2\">25. External knowledge</td><td>6</td><td>32</td><td>95</td></tr><tr><td>26. Unknown</td><td/><td>0</td><td>0</td><td>NA</td></tr></table>", |
| "num": null, |
| "html": null, |
| "text": "Distribution and Precision of paraphrases. Distribution may not sum to 100% due to rounding.", |
| "type_str": "table" |
| } |
| } |
| } |
| } |
