{"id":"W4401508264","doi":"10.1109/dcoss-iot61029.2024.00037","title":"Privacy-Preserving Multi-Party Keyword-Based Classification of Unstructured Text Data","year":2024,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Information retrieval; Unstructured data; Keyword search; Natural language processing; World Wide Web; Data mining; Big data","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":[0.003361109,0.0004074919,0.001325799,0.0008897425,0.001408858,0.002829647,0.001823164,0.00132627,0.00155204],"category_scores_gemma":[0.01244459,0.0002957405,0.001055005,0.002101143,0.001181798,0.004711669,0.003014439,0.001698633,0.001320976],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001089687,"about_ca_system_score_gemma":0.001922617,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005269902,"about_ca_topic_score_gemma":0.0005272409,"domain_scores_codex":[0.9931272,0.001862954,0.0006756504,0.001201852,0.0023282,0.0008040444],"domain_scores_gemma":[0.9882545,0.003714705,0.001316231,0.005455385,0.0009944806,0.0002646644],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002514904,0.0007354965,0.006211543,0.0006447475,0.0003086686,0.00138265,0.001591608,0.1954083,0.08054585,0.2764239,0.01360523,0.4206271],"study_design_scores_gemma":[0.00008136057,0.000174647,0.0007209712,0.00003293512,0.00005256141,0.0007805277,0.0002286375,0.8076518,0.0452931,0.1388084,0.006120268,0.00005480334],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0388968,0.0002494874,0.9563198,0.0008490176,0.00008123767,0.0001989647,0.0004105408,0.000639872,0.002354231],"genre_scores_gemma":[0.8116084,0.0002518285,0.1829363,0.0003255874,0.0001540987,0.0001993897,0.0006633212,0.0000821205,0.003778781],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003361109,"threshold_uncertainty_score":0.01777542,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.10229281079065,"score_gpt":0.3338654759186699,"score_spread":0.2315726651280198,"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."}}