{"id":"W1886218757","doi":"10.1109/icdm.2002.1183904","title":"Phrase-based document similarity based on an index graph model","year":2003,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":73,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Document clustering; Cluster analysis; Phrase; Vector space model; Information retrieval; Graph; Similarity (geometry); Matching (statistics); Set (abstract data type); Artificial intelligence; Data mining; Mathematics","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":[0.001126039,0.0005454622,0.001229881,0.004967847,0.0005954102,0.001540435,0.001710898,0.001038477,0.001696005],"category_scores_gemma":[0.008384758,0.00025483,0.001070633,0.006765921,0.0008878278,0.004589661,0.00118135,0.0008153256,0.0008462379],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009275427,"about_ca_system_score_gemma":0.0008512974,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002752351,"about_ca_topic_score_gemma":0.003168185,"domain_scores_codex":[0.9979874,0.0005126411,0.0001446471,0.0003644417,0.0009225618,0.0000682548],"domain_scores_gemma":[0.9969439,0.00123474,0.0003234595,0.0006264049,0.0007758754,0.00009559886],"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":[0.0006243394,0.0004234713,0.007424873,0.0007343898,0.0004978825,0.0004106421,0.0007160804,0.2284798,0.04287295,0.1770761,0.01223889,0.5285006],"study_design_scores_gemma":[0.00003639422,0.0001691946,0.001683523,0.00002191839,0.0001011175,0.0004200691,0.00006840036,0.9312703,0.005752809,0.05632987,0.004080491,0.00006589514],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0318128,0.0003892923,0.9641405,0.0001515461,0.00005335997,0.0001462233,0.0004599362,0.000923099,0.001923199],"genre_scores_gemma":[0.4387927,0.0007965512,0.5551353,0.0001608413,0.0001930537,0.0004146661,0.001803828,0.0002665895,0.002436385],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004967847,"threshold_uncertainty_score":0.006729782,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02070779077958386,"score_gpt":0.2583232720134422,"score_spread":0.2376154812338583,"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."}}