{"id":"W2039979754","doi":"10.1109/isspa.2012.6310552","title":"Automatic Document Topic Identification using Wikipedia Hierarchical Ontology","year":2012,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Ontology; Information retrieval; Identification (biology); Document clustering; Cluster analysis; Construct (python library); Hierarchical clustering; Data mining; 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":[0.000943532,0.0005478316,0.0004775235,0.006876561,0.0009907448,0.001611107,0.0006565831,0.0006446639,0.0009228961],"category_scores_gemma":[0.003432041,0.0002486069,0.0008291526,0.00403108,0.0003003465,0.002627577,0.001173595,0.0006706375,0.0007611985],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000862593,"about_ca_system_score_gemma":0.002137508,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01482849,"about_ca_topic_score_gemma":0.01641914,"domain_scores_codex":[0.9991282,0.0001785933,0.00008162215,0.0002401698,0.0002750348,0.00009647405],"domain_scores_gemma":[0.9985238,0.0004957066,0.0002155332,0.0001594377,0.0005203333,0.00008522957],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0005147996,0.0003944548,0.02064338,0.0008555407,0.0002381358,0.0006493907,0.002412258,0.01307061,0.07324719,0.01691242,0.02535161,0.8457102],"study_design_scores_gemma":[0.0001120279,0.0001549446,0.03219533,0.0002868846,0.0003771449,0.001133233,0.003301587,0.7660795,0.07957599,0.03057208,0.08601855,0.0001926629],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1675846,0.001452167,0.8093446,0.0004484639,0.0001453323,0.00066975,0.004728595,0.007024943,0.00860152],"genre_scores_gemma":[0.3807178,0.001005087,0.6027707,0.00007278266,0.00007142871,0.0003766652,0.01146016,0.0003175292,0.003207786],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01482849,"threshold_uncertainty_score":0.02948439,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02265378680570878,"score_gpt":0.3276773295832177,"score_spread":0.3050235427775089,"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."}}