{"id":"W4400725186","doi":"10.1007/978-3-031-64359-0_24","title":"Limitations of the Utility of Categorization in eDiscovery Review Efforts","year":2024,"lang":"en","type":"book-chapter","venue":"Communications in computer and information science","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Categorization; Computer science; Information retrieval; Data science; Artificial intelligence","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":[0.1011314,0.000806152,0.001652087,0.01780667,0.002317522,0.01468172,0.004708216,0.001677467,0.008167513],"category_scores_gemma":[0.3316756,0.0008022846,0.001147403,0.01718678,0.002631905,0.01453671,0.004464862,0.001885595,0.006001499],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004563698,"about_ca_system_score_gemma":0.006840538,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01364063,"about_ca_topic_score_gemma":0.01782058,"domain_scores_codex":[0.9074623,0.05562637,0.006223997,0.00507204,0.02413349,0.001481698],"domain_scores_gemma":[0.533525,0.3713939,0.0111705,0.02037553,0.0609406,0.002594529],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000282778,0.0001842642,0.02065986,0.003775523,0.0003544878,0.0000759894,0.001311369,0.001945794,0.0005815224,0.04756857,0.08589367,0.8373661],"study_design_scores_gemma":[0.0002343506,0.0004771819,0.0514962,0.01344384,0.001258507,0.001221886,0.00967142,0.04977687,0.008422127,0.2409592,0.622696,0.0003424569],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.08588447,0.174792,0.193905,0.1140014,0.004656576,0.002488279,0.01824527,0.006374103,0.399653],"genre_scores_gemma":[0.7343194,0.03789154,0.1733592,0.01988473,0.003046895,0.002104927,0.006223866,0.001053873,0.02211552],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.1011314,"threshold_uncertainty_score":0.5348402,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09306055483829982,"score_gpt":0.2995506511540169,"score_spread":0.2064900963157171,"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."}}