{"id":"W4385570803","doi":"10.18653/v1/2023.findings-acl.187","title":"Automated Refugee Case Analysis: A NLP Pipeline for Supporting Legal Practitioners","year":2023,"lang":"en","type":"article","venue":"","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Named-entity recognition; Pipeline (software); Natural language processing; Artificial intelligence; Refugee; Transformer; Information extraction; Architecture; Domain (mathematical analysis); Matching (statistics); Test set; Deep learning; Information retrieval; Machine learning; Data mining; Law","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002047986,0.001677057,0.0006744784,0.005102703,0.001300367,0.002567162,0.001946158,0.001842558,0.02447924],"category_scores_gemma":[0.01000089,0.0007743335,0.00124737,0.001961436,0.0006206121,0.004681109,0.003389966,0.002203793,0.02082935],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001277074,"about_ca_system_score_gemma":0.002510887,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01171819,"about_ca_topic_score_gemma":0.01778563,"domain_scores_codex":[0.9988883,0.0002083663,0.0001067019,0.0003934207,0.0002979772,0.0001052884],"domain_scores_gemma":[0.9966414,0.001611697,0.0001506247,0.000601294,0.0007925418,0.0002024082],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002780791,0.0005122933,0.006737505,0.0006765368,0.0001059183,0.001419459,0.001351992,0.009019295,0.0261831,0.009326049,0.1076671,0.8367227],"study_design_scores_gemma":[0.0001284698,0.0002222994,0.01179393,0.0003644589,0.0001128067,0.002384177,0.002664993,0.6079508,0.05673639,0.06936862,0.2481232,0.0001498362],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01989754,0.000491962,0.8458756,0.002025222,0.0001324741,0.001390746,0.01195321,0.1065705,0.01166263],"genre_scores_gemma":[0.07348972,0.0004614767,0.8873138,0.0004212329,0.00005770177,0.0004734459,0.02913855,0.001656539,0.006987564],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02447924,"threshold_uncertainty_score":0.08189124,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07342167803979849,"score_gpt":0.4534971779005597,"score_spread":0.3800754998607612,"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."}}