{"id":"W2079796778","doi":"10.1016/j.datak.2013.03.005","title":"Subject-based semantic document clustering for digital forensic investigations","year":2013,"lang":"en","type":"article","venue":"Data & Knowledge Engineering","topic":"Digital and Cyber Forensics","field":"Computer Science","cited_by":26,"is_retracted":false,"has_abstract":false,"ca_institutions":"Concordia University","funders":"","keywords":"Subject (documents); Computer science; Cluster analysis; Information retrieval; Document clustering; Natural language processing; Artificial intelligence; Data science; World Wide Web","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":[0.0008448762,0.0005807148,0.0007863636,0.008139671,0.001340218,0.00204681,0.0009276006,0.001051883,0.001612368],"category_scores_gemma":[0.002447172,0.0002290328,0.001152628,0.006337168,0.0005721767,0.001789851,0.001094313,0.0005333201,0.001627674],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007862542,"about_ca_system_score_gemma":0.002380087,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006236954,"about_ca_topic_score_gemma":0.01110362,"domain_scores_codex":[0.9990849,0.0001380412,0.0001106423,0.0002360342,0.0003279111,0.0001024308],"domain_scores_gemma":[0.9987091,0.0002219579,0.0001330377,0.0003106805,0.0005376413,0.00008769842],"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":[0.0009586053,0.0005538635,0.01166228,0.0005520848,0.0002088973,0.0002875903,0.0007223309,0.02365585,0.06290718,0.01656646,0.01316178,0.8687631],"study_design_scores_gemma":[0.00009333932,0.0004197227,0.02096792,0.0001926764,0.0004667806,0.001282677,0.002038677,0.8120723,0.06550688,0.05618059,0.04064301,0.0001354314],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1325772,0.002047897,0.8477777,0.0003175107,0.0002332499,0.0004362643,0.004327316,0.006385979,0.005896857],"genre_scores_gemma":[0.4289235,0.00103705,0.5560168,0.00008376958,0.0001452619,0.0002209128,0.009022624,0.000303366,0.004246597],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008139671,"threshold_uncertainty_score":0.01240128,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02300508494314149,"score_gpt":0.2321864369908625,"score_spread":0.209181352047721,"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."}}