{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008879264,0.0001228565,0.0002273564,0.0006935924,0.00009850112,0.0001029377,0.002797336,0.00008260805,0.000003013753],"category_scores_gemma":[0.0001324754,0.00009616417,0.00005465016,0.001083895,0.0009095195,0.003173841,0.001972866,0.0002654388,0.000006521987],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006759133,"about_ca_system_score_gemma":0.0003267476,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000111967,"about_ca_topic_score_gemma":0.00001906219,"domain_scores_codex":[0.9984414,0.00002270934,0.0008888861,0.0001790689,0.0003651913,0.000102773],"domain_scores_gemma":[0.9969857,0.0002214805,0.0004880529,0.002003915,0.0002819375,0.00001892052],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[2.934592e-7,0.000008730985,0.00006594279,0.0001959783,0.00000181057,2.609756e-8,0.0003690077,0.00001639407,0.000002173146,0.9042227,0.0002460498,0.09487084],"study_design_scores_gemma":[0.0003362947,0.00007354475,0.0156059,0.007298551,0.00003195427,0.00001253878,0.00007794522,0.3155322,0.0003359552,0.3968064,0.2633611,0.0005276088],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.0001792423,0.0199612,0.3892952,0.0197903,0.001214572,0.002375739,0.00005385852,0.0002183525,0.5669116],"genre_scores_gemma":[0.8282842,0.1302464,0.03877341,0.0007854744,0.00001319084,0.00008382103,0.00005809885,0.00001052325,0.001744888],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.828105,"threshold_uncertainty_score":0.5198193,"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."}}