Upload holdout_metrics.json with huggingface_hub

holdout_metrics.json

@@ -1,675 +1,675 @@
 {
   "real": {
-    "n_eval": 9540,
+    "n_eval": 14776,
     "overall": {
-      "tp": 6956,
-      "fp": 2047,
-      "fn": 2183,
-      "precision": 0.7726313451071865,
-      "recall": 0.7611336032388664,
-      "f05": 0.7703040907178135,
-      "f1": 0.7668393782383419
+      "tp": 8778,
+      "fp": 2928,
+      "fn": 2962,
+      "precision": 0.7498718605843158,
+      "recall": 0.7477001703577513,
+      "f05": 0.7494365138993239,
+      "f1": 0.7487844408427877
     },
     "per_key": {
       "producer": {
-        "tp": 1474,
-        "fp": 568,
-        "fn": 752,
-        "precision": 0.7218413320274241,
-        "recall": 0.6621743036837376,
-        "f05": 0.7090629209159129,
-        "f1": 0.6907216494845361
+        "tp": 1909,
+        "fp": 968,
+        "fn": 914,
+        "precision": 0.6635384080639555,
+        "recall": 0.6762309599716614,
+        "f05": 0.6660386574558649,
+        "f1": 0.6698245614035088
       },
       "year": {
-        "tp": 1737,
-        "fp": 370,
-        "fn": 488,
-        "precision": 0.8243948742287612,
-        "recall": 0.7806741573033708,
-        "f05": 0.8152633061109545,
-        "f1": 0.8019390581717452
+        "tp": 2264,
+        "fp": 520,
+        "fn": 631,
+        "precision": 0.8132183908045977,
+        "recall": 0.7820379965457686,
+        "f05": 0.8067849761242963,
+        "f1": 0.7973234724423313
       },
       "geography": {
-        "tp": 2159,
-        "fp": 751,
-        "fn": 652,
-        "precision": 0.7419243986254296,
-        "recall": 0.768054073283529,
-        "f05": 0.7470071275344268,
-        "f1": 0.7547631532948785
+        "tp": 2632,
+        "fp": 984,
+        "fn": 968,
+        "precision": 0.7278761061946902,
+        "recall": 0.7311111111111112,
+        "f05": 0.7285208148804252,
+        "f1": 0.7294900221729491
       },
       "acronym": {
-        "tp": 1586,
-        "fp": 358,
-        "fn": 291,
-        "precision": 0.815843621399177,
-        "recall": 0.8449653702717101,
-        "f05": 0.821506267481612,
-        "f1": 0.8301491756084796
+        "tp": 1973,
+        "fp": 456,
+        "fn": 449,
+        "precision": 0.8122684232194318,
+        "recall": 0.814616019818332,
+        "f05": 0.8127368594496622,
+        "f1": 0.8134405277262421
       }
     },
     "verbatim": {
-      "emitted": 9003,
-      "verbatim": 8964,
-      "rate": 0.9956681106297901,
+      "emitted": 11706,
+      "verbatim": 11659,
+      "rate": 0.9959849649752264,
       "per_key": {
         "producer": {
-          "emitted": 2042,
-          "verbatim": 2024
+          "emitted": 2877,
+          "verbatim": 2858
         },
         "year": {
-          "emitted": 2107,
-          "verbatim": 2107
+          "emitted": 2784,
+          "verbatim": 2784
         },
         "geography": {
-          "emitted": 2910,
-          "verbatim": 2892
+          "emitted": 3616,
+          "verbatim": 3599
         },
         "acronym": {
-          "emitted": 1944,
-          "verbatim": 1941
+          "emitted": 2429,
+          "verbatim": 2418
         }
       }
     },
     "usage_impact": {
-      "data_type_macro_f1": 0.6985499627733744,
-      "usage_action_macro_f1": 0.6339285026503059,
-      "impact_label_macro_f1": 0.5977288525673021,
+      "data_type_macro_f1": 0.7005048753198976,
+      "usage_action_macro_f1": 0.6554475216890102,
+      "impact_label_macro_f1": 0.5842432103101727,
       "per_label": {
         "data_type": {
           "administrative": {
-            "tp": 753,
-            "fp": 461,
-            "fn": 153,
-            "precision": 0.6202635914332785,
-            "recall": 0.8311258278145696,
-            "f1": 0.710377358490566
+            "tp": 1595,
+            "fp": 754,
+            "fn": 381,
+            "precision": 0.6790123456790124,
+            "recall": 0.8071862348178138,
+            "f1": 0.7375722543352602
           },
           "database": {
-            "tp": 573,
-            "fp": 162,
-            "fn": 272,
-            "precision": 0.7795918367346939,
-            "recall": 0.6781065088757396,
-            "f1": 0.7253164556962025
+            "tp": 855,
+            "fp": 248,
+            "fn": 616,
+            "precision": 0.7751586582048957,
+            "recall": 0.5812372535690007,
+            "f1": 0.6643356643356644
           },
           "geospatial": {
-            "tp": 152,
-            "fp": 38,
-            "fn": 65,
-            "precision": 0.8,
-            "recall": 0.7004608294930875,
-            "f1": 0.7469287469287469
+            "tp": 264,
+            "fp": 87,
+            "fn": 84,
+            "precision": 0.7521367521367521,
+            "recall": 0.7586206896551724,
+            "f1": 0.7553648068669527
           },
           "indicator": {
-            "tp": 174,
-            "fp": 77,
-            "fn": 137,
-            "precision": 0.6932270916334662,
-            "recall": 0.5594855305466238,
-            "f1": 0.6192170818505338
+            "tp": 383,
+            "fp": 328,
+            "fn": 136,
+            "precision": 0.5386779184247539,
+            "recall": 0.7379576107899807,
+            "f1": 0.6227642276422765
           },
           "other": {
-            "tp": 237,
-            "fp": 142,
-            "fn": 193,
-            "precision": 0.6253298153034301,
-            "recall": 0.5511627906976744,
-            "f1": 0.5859085290482077
+            "tp": 341,
+            "fp": 263,
+            "fn": 360,
+            "precision": 0.5645695364238411,
+            "recall": 0.4864479315263909,
+            "f1": 0.5226053639846744
           },
           "report": {
-            "tp": 211,
-            "fp": 111,
-            "fn": 168,
-            "precision": 0.65527950310559,
-            "recall": 0.5567282321899736,
-            "f1": 0.6019971469329529
+            "tp": 538,
+            "fp": 182,
+            "fn": 222,
+            "precision": 0.7472222222222222,
+            "recall": 0.7078947368421052,
+            "f1": 0.727027027027027
           },
           "survey": {
-            "tp": 1293,
-            "fp": 142,
-            "fn": 145,
-            "precision": 0.9010452961672474,
-            "recall": 0.8991655076495132,
-            "f1": 0.9001044204664115
+            "tp": 2165,
+            "fp": 281,
+            "fn": 344,
+            "precision": 0.8851185609157809,
+            "recall": 0.862893583100837,
+            "f1": 0.8738647830474269
           }
         },
         "usage_action": {
           "analyze": {
-            "tp": 979,
-            "fp": 349,
-            "fn": 629,
-            "precision": 0.7371987951807228,
-            "recall": 0.6088308457711443,
-            "f1": 0.6668937329700272
+            "tp": 1646,
+            "fp": 803,
+            "fn": 732,
+            "precision": 0.6721110657411188,
+            "recall": 0.6921783010933558,
+            "f1": 0.6819970996478143
           },
           "curate": {
-            "tp": 553,
-            "fp": 234,
-            "fn": 290,
-            "precision": 0.7026683608640406,
-            "recall": 0.6559905100830368,
-            "f1": 0.6785276073619633
+            "tp": 1634,
+            "fp": 647,
+            "fn": 414,
+            "precision": 0.716352476983779,
+            "recall": 0.7978515625,
+            "f1": 0.7549087549087549
           },
           "inform": {
-            "tp": 546,
-            "fp": 473,
-            "fn": 229,
-            "precision": 0.535819430814524,
-            "recall": 0.704516129032258,
-            "f1": 0.608695652173913
+            "tp": 922,
+            "fp": 566,
+            "fn": 669,
+            "precision": 0.6196236559139785,
+            "recall": 0.57950974230044,
+            "f1": 0.598895745371874
           },
           "source": {
-            "tp": 730,
-            "fp": 377,
-            "fn": 224,
-            "precision": 0.6594399277326106,
-            "recall": 0.7651991614255765,
-            "f1": 0.7083939835031537
+            "tp": 1232,
+            "fp": 469,
+            "fn": 435,
+            "precision": 0.7242798353909465,
+            "recall": 0.7390521895620876,
+            "f1": 0.7315914489311164
           },
           "validate": {
-            "tp": 160,
-            "fp": 125,
-            "fn": 186,
-            "precision": 0.5614035087719298,
-            "recall": 0.4624277456647399,
-            "f1": 0.5071315372424722
+            "tp": 246,
+            "fp": 119,
+            "fn": 354,
+            "precision": 0.673972602739726,
+            "recall": 0.41,
+            "f1": 0.5098445595854921
           }
         },
         "impact_label": {
           "evidence-base": {
-            "tp": 2050,
-            "fp": 350,
-            "fn": 647,
-            "precision": 0.8541666666666666,
-            "recall": 0.7601038190582128,
-            "f1": 0.8043947420051011
+            "tp": 3123,
+            "fp": 963,
+            "fn": 785,
+            "precision": 0.76431718061674,
+            "recall": 0.7991299897645855,
+            "f1": 0.7813360020015012
           },
           "none": {
-            "tp": 730,
-            "fp": 426,
-            "fn": 270,
-            "precision": 0.6314878892733564,
-            "recall": 0.73,
-            "f1": 0.6771799628942485
+            "tp": 1408,
+            "fp": 564,
+            "fn": 597,
+            "precision": 0.7139959432048681,
+            "recall": 0.7022443890274315,
+            "f1": 0.7080714106110133
           },
           "policy": {
-            "tp": 586,
-            "fp": 366,
-            "fn": 192,
-            "precision": 0.615546218487395,
-            "recall": 0.7532133676092545,
-            "f1": 0.677456647398844
+            "tp": 1453,
+            "fp": 712,
+            "fn": 674,
+            "precision": 0.6711316397228637,
+            "recall": 0.6831217677480019,
+            "f1": 0.6770736253494875
           },
           "uptake": {
-            "tp": 8,
-            "fp": 10,
-            "fn": 43,
-            "precision": 0.4444444444444444,
-            "recall": 0.1568627450980392,
-            "f1": 0.23188405797101447
+            "tp": 26,
+            "fp": 35,
+            "fn": 218,
+            "precision": 0.4262295081967213,
+            "recall": 0.10655737704918032,
+            "f1": 0.17049180327868851
           }
         }
       },
       "usage_summary": {
-        "n": 4526,
-        "mean_sim": 0.5934299230575562,
-        "grounded_rate": 0.7551922227132125
+        "n": 8284,
+        "mean_sim": 0.5522229075431824,
+        "grounded_rate": 0.6512554321583776
       }
     },
     "examples": [
       {
-        "mention": "SARA",
-        "gold": "{\"acronym\": \"SARA\", \"producer\": \"World Health Organization\"}",
-        "pred": "{\"acronym\": \"SARA\", \"producer\": \"USAID\"}",
-        "correct": "acronym",
-        "verbatim": "producer,acronym"
+        "mention": "ILO ' s Yearbook of Labor Statistics",
+        "gold": "{\"geography\": \"Kenya\", \"producer\": \"ILO\", \"year\": \"1995\"}",
+        "pred": "{\"geography\": \"Kenya\", \"producer\": \"ILO\", \"year\": \"1995\"}",
+        "correct": "producer,year,geography",
+        "verbatim": "producer,year,geography"
       },
       {
-        "mention": "Encuesta Nacional Industrial Annual",
-        "gold": "{\"acronym\": \"ENIA\", \"geography\": \"Chilean\", \"producer\": \"National Statistical Institute\"}",
-        "pred": "{\"acronym\": \"ENIA\", \"geography\": \"Chilean\"}",
-        "correct": "geography,acronym",
-        "verbatim": "geography,acronym"
+        "mention": "BEEPS survey",
+        "gold": "{\"acronym\": \"BEEPS\", \"year\": \"2008\"}",
+        "pred": "{\"acronym\": \"BEEPS\", \"year\": \"2008\"}",
+        "correct": "year,acronym",
+        "verbatim": "year,acronym"
       },
       {
-        "mention": "child protection services data",
-        "gold": "{\"year\": \"2018\", \"geography\": \"Jordan\"}",
-        "pred": "{\"year\": \"2018\", \"geography\": \"Jordan\"}",
-        "correct": "year,geography",
-        "verbatim": "year,geography"
+        "mention": "ENCOVI data",
+        "gold": "{\"geography\": \"Venezuela\"}",
+        "pred": "{\"acronym\": \"ENCOVI\", \"geography\": \"Colombia\"}",
+        "correct": "-",
+        "verbatim": "geography,acronym"
       },
       {
-        "mention": "GBD data",
-        "gold": "{\"geography\": \"Georgia\", \"year\": \"2017\"}",
-        "pred": "{\"acronym\": \"GBD\", \"geography\": \"Georgia\", \"year\": \"2017\"}",
-        "correct": "year,geography",
-        "verbatim": "year,geography,acronym"
+        "mention": "panel data",
+        "gold": "{\"geography\": \"Egypt\"}",
+        "pred": "{\"geography\": \"Egypt\"}",
+        "correct": "geography",
+        "verbatim": "geography"
       },
       {
-        "mention": "Human Capital Index",
-        "gold": "{\"producer\": \"World Bank\", \"geography\": \"Djibouti\"}",
-        "pred": "{\"producer\": \"World Bank\", \"year\": \"2019\", \"geography\": \"Djibouti\", \"acronym\": \"HDI\"}",
-        "correct": "producer,geography",
-        "verbatim": "producer,year,geography,acronym"
+        "mention": "WHO Coronavirus Disease (COVID-19) Dashboard",
+        "gold": "{\"producer\": \"WHO\", \"geography\": \"worldwide\", \"acronym\": \"COVID-19\"}",
+        "pred": "{\"producer\": \"WHO\", \"geography\": \"Uganda\"}",
+        "correct": "producer",
+        "verbatim": "producer,geography"
       },
       {
-        "mention": "night-time lights data",
-        "gold": "{\"geography\": \"district level\"}",
-        "pred": "{\"geography\": \"rural India\"}",
-        "correct": "-",
-        "verbatim": "geography"
+        "mention": "Integrated Household Panel Surveys",
+        "gold": "{\"acronym\": \"IHPS\", \"geography\": \"Malawi\", \"year\": \"2013\"}",
+        "pred": "{\"acronym\": \"IHPS\", \"geography\": \"Malawi\"}",
+        "correct": "geography,acronym",
+        "verbatim": "geography,acronym"
       },
       {
-        "mention": "annual production of four major crops",
-        "gold": "{\"geography\": \"global\", \"producer\": \"Iizumi and Sakai ( 2020 )\", \"year\": \"2020\"}",
-        "pred": "{\"geography\": \"global\", \"year\": \"2020\"}",
-        "correct": "year,geography",
-        "verbatim": "year,geography"
+        "mention": "socio-economic survey",
+        "gold": "{\"year\": \"2010\"}",
+        "pred": "{\"year\": \"2010\"}",
+        "correct": "year",
+        "verbatim": "year"
       },
       {
-        "mention": "registry data",
-        "gold": "{\"producer\": \"UNHCR\"}",
-        "pred": "{\"producer\": \"UNHCR\"}",
-        "correct": "producer",
-        "verbatim": "producer"
+        "mention": "World Bank Enterprise Survey Data for Uganda",
+        "gold": "{\"producer\": \"World Bank\", \"year\": \"2013\", \"geography\": \"Uganda\"}",
+        "pred": "{\"producer\": \"World Bank\", \"year\": \"2013\", \"geography\": \"Uganda\"}",
+        "correct": "producer,year,geography",
+        "verbatim": "producer,year,geography"
       },
       {
-        "mention": "DTM Mali – Septembre 2019",
-        "gold": "{\"year\": \"2019\", \"geography\": \"Mali\", \"acronym\": \"DTM\"}",
-        "pred": "{\"year\": \"2019\", \"geography\": \"Mali\", \"acronym\": \"DTM\"}",
-        "correct": "year,geography,acronym",
-        "verbatim": "year,geography,acronym"
+        "mention": "internally displaced person (IDP) data",
+        "gold": "{\"producer\": \"Internal Displacement Monitoring Centre\", \"geography\": \"Philippines\", \"acronym\": \"IDMC\"}",
+        "pred": "{\"producer\": \"Internal Displacement Monitoring Centre\", \"year\": \"2020\", \"geography\": \"Philippines\", \"acronym\": \"IDMC\"}",
+        "correct": "producer,geography,acronym",
+        "verbatim": "producer,year,geography,acronym"
       },
       {
-        "mention": "natural disaster information",
-        "gold": "{\"geography\": \"LAC\"}",
-        "pred": "{\"geography\": \"LAC\"}",
-        "correct": "geography",
-        "verbatim": "geography"
+        "mention": "Displacement data",
+        "gold": "{\"producer\": \"Internal Displacement Monitoring Centre\"}",
+        "pred": "{\"producer\": \"Internal Displacement Monitoring Centre\"}",
+        "correct": "producer",
+        "verbatim": "producer"
       },
       {
-        "mention": "Enquˆete Nationale sur les Conditions de Vie des",
-        "gold": "{\"acronym\": \"ECVMA\", \"geography\": \"Niger\"}",
-        "pred": "{\"acronym\": \"ECVMA\", \"geography\": \"Niger\"}",
-        "correct": "geography,acronym",
-        "verbatim": "geography,acronym"
+        "mention": "regulatory surveys",
+        "gold": "{\"geography\": \"East Africa\", \"producer\": \"World Bank\"}",
+        "pred": "{\"geography\": \"Kenya\", \"producer\": \"local consultants\"}",
+        "correct": "-",
+        "verbatim": "producer,geography"
       },
       {
-        "mention": "Covid-19 Trade Watch",
-        "gold": "{\"producer\": \"World Bank\", \"year\": \"2020\"}",
-        "pred": "{\"producer\": \"World Bank\", \"year\": \"2020\"}",
-        "correct": "producer,year",
-        "verbatim": "producer,year"
+        "mention": "UNHCR archives",
+        "gold": "{\"producer\": \"UNHCR\"}",
+        "pred": "{\"producer\": \"UNHCR\"}",
+        "correct": "producer",
+        "verbatim": "producer"
       }
     ]
   },
   "synthetic": {
     "n_eval": 2305,
     "overall": {
-      "tp": 1005,
-      "fp": 155,
-      "fn": 165,
-      "precision": 0.8663793103448276,
-      "recall": 0.8589743589743589,
-      "f05": 0.8648881239242684,
-      "f1": 0.8626609442060087
+      "tp": 985,
+      "fp": 236,
+      "fn": 185,
+      "precision": 0.8067158067158067,
+      "recall": 0.8418803418803419,
+      "f05": 0.8135117277832838,
+      "f1": 0.8239230447511501
     },
     "per_key": {
       "producer": {
-        "tp": 252,
-        "fp": 23,
-        "fn": 35,
-        "precision": 0.9163636363636364,
-        "recall": 0.8780487804878049,
-        "f05": 0.9084354722422495,
-        "f1": 0.8967971530249111
+        "tp": 231,
+        "fp": 39,
+        "fn": 56,
+        "precision": 0.8555555555555555,
+        "recall": 0.8048780487804879,
+        "f05": 0.8449158741770301,
+        "f1": 0.829443447037702
       },
       "year": {
-        "tp": 259,
-        "fp": 30,
-        "fn": 46,
-        "precision": 0.8961937716262975,
-        "recall": 0.8491803278688524,
-        "f05": 0.8863791923340176,
-        "f1": 0.8720538720538721
+        "tp": 265,
+        "fp": 50,
+        "fn": 40,
+        "precision": 0.8412698412698413,
+        "recall": 0.8688524590163934,
+        "f05": 0.8466453674121406,
+        "f1": 0.8548387096774194
       },
       "geography": {
-        "tp": 355,
-        "fp": 62,
-        "fn": 42,
-        "precision": 0.8513189448441247,
-        "recall": 0.8942065491183879,
-        "f05": 0.8595641646489104,
-        "f1": 0.8722358722358723
+        "tp": 344,
+        "fp": 63,
+        "fn": 53,
+        "precision": 0.8452088452088452,
+        "recall": 0.8664987405541562,
+        "f05": 0.8493827160493826,
+        "f1": 0.855721393034826
       },
       "acronym": {
-        "tp": 139,
-        "fp": 40,
-        "fn": 42,
-        "precision": 0.776536312849162,
-        "recall": 0.7679558011049724,
-        "f05": 0.7748049052396878,
-        "f1": 0.7722222222222223
+        "tp": 145,
+        "fp": 84,
+        "fn": 36,
+        "precision": 0.6331877729257642,
+        "recall": 0.8011049723756906,
+        "f05": 0.6608933454876937,
+        "f1": 0.7073170731707317
       }
     },
     "verbatim": {
-      "emitted": 1160,
-      "verbatim": 1156,
-      "rate": 0.996551724137931,
+      "emitted": 1221,
+      "verbatim": 1216,
+      "rate": 0.9959049959049959,
       "per_key": {
         "producer": {
-          "emitted": 275,
-          "verbatim": 275
+          "emitted": 270,
+          "verbatim": 270
         },
         "year": {
-          "emitted": 289,
-          "verbatim": 289
+          "emitted": 315,
+          "verbatim": 315
         },
         "geography": {
-          "emitted": 417,
-          "verbatim": 413
+          "emitted": 407,
+          "verbatim": 402
         },
         "acronym": {
-          "emitted": 179,
-          "verbatim": 179
+          "emitted": 229,
+          "verbatim": 229
         }
       }
     },
     "usage_impact": {
-      "data_type_macro_f1": 0.7857483174178359,
-      "usage_action_macro_f1": 0.6624851717052487,
-      "impact_label_macro_f1": 0.5738134926381705,
+      "data_type_macro_f1": 0.7818281325560671,
+      "usage_action_macro_f1": 0.6488552648711132,
+      "impact_label_macro_f1": 0.5147452129823803,
       "per_label": {
         "data_type": {
           "administrative": {
-            "tp": 247,
-            "fp": 82,
-            "fn": 31,
-            "precision": 0.7507598784194529,
-            "recall": 0.8884892086330936,
-            "f1": 0.8138385502471169
+            "tp": 232,
+            "fp": 48,
+            "fn": 46,
+            "precision": 0.8285714285714286,
+            "recall": 0.8345323741007195,
+            "f1": 0.8315412186379929
           },
           "database": {
-            "tp": 190,
-            "fp": 50,
-            "fn": 69,
-            "precision": 0.7916666666666666,
-            "recall": 0.7335907335907336,
-            "f1": 0.7615230460921844
+            "tp": 170,
+            "fp": 33,
+            "fn": 89,
+            "precision": 0.8374384236453202,
+            "recall": 0.6563706563706564,
+            "f1": 0.7359307359307359
           },
           "geospatial": {
-            "tp": 93,
-            "fp": 18,
-            "fn": 24,
-            "precision": 0.8378378378378378,
-            "recall": 0.7948717948717948,
-            "f1": 0.8157894736842105
+            "tp": 106,
+            "fp": 33,
+            "fn": 11,
+            "precision": 0.762589928057554,
+            "recall": 0.905982905982906,
+            "f1": 0.828125
           },
           "indicator": {
-            "tp": 117,
-            "fp": 18,
-            "fn": 37,
-            "precision": 0.8666666666666667,
-            "recall": 0.7597402597402597,
-            "f1": 0.8096885813148789
+            "tp": 138,
+            "fp": 59,
+            "fn": 16,
+            "precision": 0.700507614213198,
+            "recall": 0.8961038961038961,
+            "f1": 0.7863247863247862
           },
           "other": {
-            "tp": 114,
-            "fp": 60,
-            "fn": 56,
-            "precision": 0.6551724137931034,
-            "recall": 0.6705882352941176,
-            "f1": 0.6627906976744184
+            "tp": 115,
+            "fp": 61,
+            "fn": 55,
+            "precision": 0.6534090909090909,
+            "recall": 0.6764705882352942,
+            "f1": 0.6647398843930635
           },
           "report": {
-            "tp": 30,
-            "fp": 9,
-            "fn": 15,
-            "precision": 0.7692307692307693,
-            "recall": 0.6666666666666666,
-            "f1": 0.7142857142857142
+            "tp": 27,
+            "fp": 5,
+            "fn": 18,
+            "precision": 0.84375,
+            "recall": 0.6,
+            "f1": 0.7012987012987012
           },
           "survey": {
-            "tp": 564,
-            "fp": 45,
-            "fn": 50,
-            "precision": 0.9261083743842364,
-            "recall": 0.9185667752442996,
-            "f1": 0.9223221586263287
+            "tp": 566,
+            "fp": 44,
+            "fn": 48,
+            "precision": 0.9278688524590164,
+            "recall": 0.9218241042345277,
+            "f1": 0.9248366013071895
           }
         },
         "usage_action": {
           "analyze": {
-            "tp": 76,
-            "fp": 4,
-            "fn": 8,
-            "precision": 0.95,
-            "recall": 0.9047619047619048,
-            "f1": 0.9268292682926829
+            "tp": 77,
+            "fp": 11,
+            "fn": 7,
+            "precision": 0.875,
+            "recall": 0.9166666666666666,
+            "f1": 0.8953488372093024
           },
           "causal": {
-            "tp": 56,
-            "fp": 19,
-            "fn": 52,
-            "precision": 0.7466666666666667,
-            "recall": 0.5185185185185185,
-            "f1": 0.6120218579234973
+            "tp": 59,
+            "fp": 22,
+            "fn": 49,
+            "precision": 0.7283950617283951,
+            "recall": 0.5462962962962963,
+            "f1": 0.6243386243386243
           },
           "curate": {
-            "tp": 32,
-            "fp": 2,
-            "fn": 8,
-            "precision": 0.9411764705882353,
-            "recall": 0.8,
-            "f1": 0.8648648648648648
+            "tp": 34,
+            "fp": 6,
+            "fn": 6,
+            "precision": 0.85,
+            "recall": 0.85,
+            "f1": 0.85
           },
           "inform": {
-            "tp": 138,
-            "fp": 48,
-            "fn": 51,
-            "precision": 0.7419354838709677,
-            "recall": 0.7301587301587301,
-            "f1": 0.736
+            "tp": 140,
+            "fp": 77,
+            "fn": 49,
+            "precision": 0.6451612903225806,
+            "recall": 0.7407407407407407,
+            "f1": 0.689655172413793
           },
           "measure": {
-            "tp": 188,
-            "fp": 146,
-            "fn": 82,
-            "precision": 0.562874251497006,
-            "recall": 0.6962962962962963,
-            "f1": 0.6225165562913907
+            "tp": 169,
+            "fp": 119,
+            "fn": 101,
+            "precision": 0.5868055555555556,
+            "recall": 0.6259259259259259,
+            "f1": 0.6057347670250895
           },
           "model": {
-            "tp": 90,
-            "fp": 49,
-            "fn": 68,
-            "precision": 0.6474820143884892,
-            "recall": 0.569620253164557,
-            "f1": 0.6060606060606062
+            "tp": 79,
+            "fp": 36,
+            "fn": 79,
+            "precision": 0.6869565217391305,
+            "recall": 0.5,
+            "f1": 0.5787545787545788
           },
           "other": {
-            "tp": 25,
-            "fp": 21,
-            "fn": 36,
-            "precision": 0.5434782608695652,
-            "recall": 0.4098360655737705,
-            "f1": 0.4672897196261682
+            "tp": 28,
+            "fp": 24,
+            "fn": 33,
+            "precision": 0.5384615384615384,
+            "recall": 0.45901639344262296,
+            "f1": 0.4955752212389381
           },
           "produce": {
-            "tp": 57,
-            "fp": 29,
-            "fn": 61,
-            "precision": 0.6627906976744186,
-            "recall": 0.4830508474576271,
-            "f1": 0.5588235294117647
+            "tp": 75,
+            "fp": 74,
+            "fn": 43,
+            "precision": 0.5033557046979866,
+            "recall": 0.635593220338983,
+            "f1": 0.5617977528089888
           },
           "source": {
-            "tp": 276,
-            "fp": 165,
-            "fn": 121,
-            "precision": 0.6258503401360545,
-            "recall": 0.6952141057934509,
-            "f1": 0.6587112171837709
+            "tp": 242,
+            "fp": 126,
+            "fn": 155,
+            "precision": 0.657608695652174,
+            "recall": 0.6095717884130982,
+            "f1": 0.6326797385620916
           },
           "validate": {
-            "tp": 114,
-            "fp": 70,
-            "fn": 46,
-            "precision": 0.6195652173913043,
-            "recall": 0.7125,
-            "f1": 0.6627906976744186
+            "tp": 103,
+            "fp": 47,
+            "fn": 57,
+            "precision": 0.6866666666666666,
+            "recall": 0.64375,
+            "f1": 0.664516129032258
           },
           "wrangle": {
-            "tp": 24,
-            "fp": 8,
-            "fn": 28,
-            "precision": 0.75,
-            "recall": 0.46153846153846156,
-            "f1": 0.5714285714285714
+            "tp": 38,
+            "fp": 51,
+            "fn": 14,
+            "precision": 0.42696629213483145,
+            "recall": 0.7307692307692307,
+            "f1": 0.5390070921985815
           }
         },
         "impact_label": {
           "evidence-base": {
-            "tp": 942,
-            "fp": 197,
-            "fn": 86,
-            "precision": 0.8270412642669008,
-            "recall": 0.9163424124513618,
-            "f1": 0.8694047069681587
+            "tp": 1000,
+            "fp": 287,
+            "fn": 28,
+            "precision": 0.777000777000777,
+            "recall": 0.9727626459143969,
+            "f1": 0.8639308855291578
           },
           "market-use": {
             "tp": 0,
-            "fp": 1,
+            "fp": 0,
             "fn": 8,
             "precision": 0.0,
             "recall": 0.0,
             "f1": 0.0
           },
           "none": {
-            "tp": 54,
-            "fp": 19,
-            "fn": 42,
-            "precision": 0.7397260273972602,
-            "recall": 0.5625,
-            "f1": 0.6390532544378699
+            "tp": 43,
+            "fp": 21,
+            "fn": 53,
+            "precision": 0.671875,
+            "recall": 0.4479166666666667,
+            "f1": 0.5375
           },
           "operational-use": {
-            "tp": 9,
-            "fp": 6,
-            "fn": 57,
-            "precision": 0.6,
-            "recall": 0.13636363636363635,
-            "f1": 0.22222222222222218
+            "tp": 8,
+            "fp": 11,
+            "fn": 58,
+            "precision": 0.42105263157894735,
+            "recall": 0.12121212121212122,
+            "f1": 0.18823529411764706
           },
           "policy": {
-            "tp": 254,
-            "fp": 82,
-            "fn": 110,
-            "precision": 0.7559523809523809,
-            "recall": 0.6978021978021978,
-            "f1": 0.7257142857142858
+            "tp": 163,
+            "fp": 11,
+            "fn": 201,
+            "precision": 0.9367816091954023,
+            "recall": 0.4478021978021978,
+            "f1": 0.6059479553903345
           },
           "uptake": {
-            "tp": 73,
-            "fp": 0,
-            "fn": 2,
-            "precision": 1.0,
-            "recall": 0.9733333333333334,
-            "f1": 0.9864864864864865
+            "tp": 75,
+            "fp": 18,
+            "fn": 0,
+            "precision": 0.8064516129032258,
+            "recall": 1.0,
+            "f1": 0.8928571428571428
           }
         }
       },
       "usage_summary": {
         "n": 1637,
-        "mean_sim": 0.5707868337631226,
-        "grounded_rate": 0.6823457544288333
+        "mean_sim": 0.5752291679382324,
+        "grounded_rate": 0.6768478924862553
       }
     },
     "examples": [
       {
-        "mention": "Women Business and Law",
-        "gold": "{\"acronym\": \"WBL\", \"producer\": \"World Bank\"}",
-        "pred": "{\"acronym\": \"WBL\", \"producer\": \"World Bank\"}",
-        "correct": "producer,acronym",
-        "verbatim": "producer,acronym"
+        "mention": "C8te d ' Ivoire data set",
+        "gold": "{\"geography\": \"C8te d ' Ivoire\"}",
+        "pred": "{\"geography\": \"C8te d ' Ivoire\"}",
+        "correct": "geography",
+        "verbatim": "geography"
       },
       {
-        "mention": "Quarterly Public Finance Database",
-        "gold": "{\"producer\": \"national sources\"}",
-        "pred": "{\"producer\": \"national sources\"}",
-        "correct": "producer",
-        "verbatim": "producer"
+        "mention": "Access to Finance Survey",
+        "gold": "{\"acronym\": \"Access to Finance Survey\", \"geography\": \"Nigeria\", \"year\": \"2008\"}",
+        "pred": "{\"acronym\": \"Access to Finance\", \"geography\": \"Nigeria\", \"year\": \"2008\"}",
+        "correct": "year,geography",
+        "verbatim": "year,geography,acronym"
       },
       {
-        "mention": "IMF data",
-        "gold": "{\"geography\": \"40 low-income countries\", \"producer\": \"IMF\"}",
-        "pred": "{\"geography\": \"40 low-income countries\", \"producer\": \"IMF\"}",
-        "correct": "producer,geography",
-        "verbatim": "producer,geography"
+        "mention": "Data from Mongolia and Myanmar",
+        "gold": "{\"geography\": \"Mongolia and Myanmar\"}",
+        "pred": "{\"geography\": \"Mongolia and Myanmar\"}",
+        "correct": "geography",
+        "verbatim": "geography"
       },
       {
-        "mention": "yearly data on firms ’ export transactions",
-        "gold": "{\"acronym\": \"ECI\", \"producer\": \"INE\"}",
-        "pred": "{\"acronym\": \"ECI\", \"producer\": \"INE\"}",
-        "correct": "producer,acronym",
-        "verbatim": "producer,acronym"
+        "mention": "HIES 2016 / 17",
+        "gold": "{\"geography\": \"Bangladesh\", \"year\": \"2016\"}",
+        "pred": "{\"acronym\": \"HIES\", \"geography\": \"Bangladesh\", \"year\": \"2016\"}",
+        "correct": "year,geography",
+        "verbatim": "year,geography,acronym"
       },
       {
-        "mention": "Inspectorate 2011 data",
-        "gold": "{\"geography\": \"Tanzania\", \"year\": \"2011\"}",
-        "pred": "{\"geography\": \"Tanzania\", \"year\": \"2011\"}",
-        "correct": "year,geography",
+        "mention": "survey of households",
+        "gold": "{\"geography\": \"Nicaraguan\"}",
+        "pred": "{\"geography\": \"Nicaraguan\", \"year\": \"1998\"}",
+        "correct": "geography",
         "verbatim": "year,geography"
       },
       {
-        "mention": "2018 / 19 NLSS data",
-        "gold": "{\"acronym\": \"NLSS\", \"geography\": \"Nigeria\", \"producer\": \"National Bureau of Statistics\"}",
-        "pred": "{\"acronym\": \"NLSS\", \"geography\": \"Nigeria\", \"producer\": \"Nigeria ’ s National Bureau of Statistics ( NBS )\"}",
-        "correct": "geography,acronym",
-        "verbatim": "producer,geography,acronym"
+        "mention": "National Annual Industrial Survey",
+        "gold": "{\"geography\": \"Chile\", \"year\": \"2001\"}",
+        "pred": "{\"geography\": \"Chile\", \"year\": \"2001\"}",
+        "correct": "year,geography",
+        "verbatim": "year,geography"
       },
       {
-        "mention": "administrative panel data",
-        "gold": "{\"geography\": \"Colombia\"}",
-        "pred": "{\"geography\": \"Colombia\"}",
-        "correct": "geography",
-        "verbatim": "geography"
+        "mention": "1985 Living Standards Survey",
+        "gold": "{\"acronym\": \"CILSS\", \"geography\": \"Côte d'Ivoire\", \"year\": \"1985\"}",
+        "pred": "{\"acronym\": \"CILSS\", \"geography\": \"Côte d'Ivoire\", \"year\": \"1985\"}",
+        "correct": "year,geography,acronym",
+        "verbatim": "year,geography,acronym"
       },
       {
-        "mention": "India Human Development Survey ( IHDS )",
-        "gold": "{\"acronym\": \"( IHDS )\", \"geography\": \"India\"}",
-        "pred": "{\"acronym\": \"IHDS\", \"geography\": \"India\"}",
-        "correct": "geography",
-        "verbatim": "geography,acronym"
+        "mention": "World Bank Survey on Applied Policies in Services",
+        "gold": "{\"producer\": \"The World Bank Research Department\", \"year\": \"2008\"}",
+        "pred": "{\"producer\": \"The World Bank Research Department\", \"year\": \"2008\"}",
+        "correct": "producer,year",
+        "verbatim": "producer,year"
       },
       {
-        "mention": "OECD-PISA 2012 database",
-        "gold": "{\"producer\": \"OECD\", \"year\": \"2012\"}",
-        "pred": "{\"geography\": \"Vietnam\", \"producer\": \"OECD\", \"year\": \"2012\"}",
-        "correct": "producer,year",
-        "verbatim": "producer,year,geography"
+        "mention": "Pesquisa de Orçamentos Familiares",
+        "gold": "{\"acronym\": \"POF\", \"geography\": \"Brazil\"}",
+        "pred": "{\"acronym\": \"POF\", \"geography\": \"Brazil\"}",
+        "correct": "geography,acronym",
+        "verbatim": "geography,acronym"
       },
       {
-        "mention": "DHS figures on access",
-        "gold": "{\"year\": \"1997\"}",
-        "pred": "{\"year\": \"1997\"}",
+        "mention": "administrative data",
+        "gold": "{\"year\": \"2015\"}",
+        "pred": "{\"year\": \"2015\"}",
         "correct": "year",
         "verbatim": "year"
       },
       {
-        "mention": "DHS surveys",
-        "gold": "{\"acronym\": \"DHS\", \"geography\": \"Nigeria\"}",
-        "pred": "{\"geography\": \"Nigeria, Cameroon, and Chad\"}",
-        "correct": "-",
-        "verbatim": "geography"
+        "mention": "Bangladesh Return Migrant Survey",
+        "gold": "{\"acronym\": \"( BRMS )\", \"geography\": \"Bangladesh\", \"producer\": \"World Bank\", \"year\": \"2018 / 19\"}",
+        "pred": "{\"acronym\": \"BRMS\", \"geography\": \"Bangladesh\", \"producer\": \"World Bank\", \"year\": \"2018\"}",
+        "correct": "producer,geography",
+        "verbatim": "producer,year,geography,acronym"
       },
       {
-        "mention": "International Country Risk Guide",
-        "gold": "{\"acronym\": \"( ICRG )\"}",
-        "pred": "{\"acronym\": \"ICRG\"}",
-        "correct": "-",
-        "verbatim": "acronym"
+        "mention": "ACLED data",
+        "gold": "{\"acronym\": \"ACLED\", \"geography\": \"Novosilka city in Donetsk oblast\", \"producer\": \"Armed Conflict Location and Event Data\", \"year\": \"2022\"}",
+        "pred": "{\"acronym\": \"ACLED\", \"geography\": \"Novosilka city\", \"year\": \"2022\"}",
+        "correct": "year,acronym",
+        "verbatim": "year,geography,acronym"
       }
     ]
   }