{"id":"W3093493237","doi":"10.1145/3340531.3412746","title":"The Utility of Context When Extracting Entities From Legal Documents","year":2020,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Cisco Systems (Canada)","funders":"","keywords":"Computer science; Sentence; Context (archaeology); Natural language processing; Process (computing); Named-entity recognition; Artificial intelligence; Sequence (biology); Information retrieval; Layer (electronics); Legal document; Information extraction; Task (project management)","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001539061,0.0008046143,0.0004876367,0.004760795,0.001029905,0.002142995,0.0007441693,0.0009828021,0.003706091],"category_scores_gemma":[0.01321558,0.0004011027,0.0005694767,0.002928061,0.0004125424,0.006538007,0.001401875,0.00118295,0.003745946],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004926978,"about_ca_system_score_gemma":0.001018329,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005199851,"about_ca_topic_score_gemma":0.01562068,"domain_scores_codex":[0.9986626,0.0004200562,0.0001509971,0.0004672566,0.0002141784,0.00008496336],"domain_scores_gemma":[0.993229,0.004329504,0.0005273756,0.0007843757,0.0009342582,0.0001953445],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0005428345,0.0001884795,0.02734211,0.001105367,0.00011364,0.001294553,0.00213641,0.007603334,0.03138652,0.00802013,0.02283856,0.8974282],"study_design_scores_gemma":[0.0001535335,0.0006984881,0.06726465,0.001285925,0.0007723647,0.005872575,0.004582636,0.3779574,0.1330305,0.05631211,0.3516437,0.0004260878],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.3843456,0.01701419,0.4989468,0.004343349,0.0009857152,0.0008412704,0.0136953,0.02374157,0.05608624],"genre_scores_gemma":[0.6383814,0.00359059,0.3412693,0.0004857186,0.0004134721,0.0001890356,0.01019759,0.0006404174,0.004832538],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005199851,"threshold_uncertainty_score":0.01239812,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04434289484593809,"score_gpt":0.2590443196216146,"score_spread":0.2147014247756765,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}