{"id":"W142552274","doi":"10.1007/978-3-642-41062-8_19","title":"Text Categorization via Similarity Search","year":2013,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Ottawa","funders":"","keywords":"Computer science; Artificial intelligence; Similarity (geometry); Cluster analysis; Categorization; Centroid; Outlier; Preprocessor; Metric (unit); Similarity measure; Point (geometry); Text categorization; Class (philosophy); Pattern recognition (psychology); 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":[0.0009776372,0.0008893819,0.001986111,0.007149538,0.001195009,0.002301859,0.001783945,0.001275834,0.01189146],"category_scores_gemma":[0.002960587,0.0003875567,0.001676427,0.007463313,0.0005334347,0.003341535,0.002273896,0.000920768,0.0110454],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006385113,"about_ca_system_score_gemma":0.0011738,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001739849,"about_ca_topic_score_gemma":0.002634834,"domain_scores_codex":[0.9982926,0.0002606822,0.0001737946,0.0004971073,0.0006673152,0.000108549],"domain_scores_gemma":[0.9988027,0.0003624507,0.00008769726,0.0003084304,0.0003813929,0.00005735558],"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":[0.0001484786,0.0002163315,0.0004520182,0.0002403557,0.00008875554,0.00005823773,0.00006922018,0.002239314,0.01605076,0.007908832,0.0173109,0.9552168],"study_design_scores_gemma":[0.0001748755,0.000813357,0.004900764,0.0001757442,0.0004031308,0.002418063,0.0006336159,0.67307,0.07635,0.1510828,0.08982685,0.0001507956],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02117245,0.00284422,0.9552062,0.000437694,0.0003454428,0.0006317257,0.001244916,0.007618515,0.01049884],"genre_scores_gemma":[0.146711,0.001404749,0.8161401,0.0002923211,0.0003192058,0.0004672147,0.006775837,0.0004502808,0.02743921],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01189146,"threshold_uncertainty_score":0.03978086,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02403845612795377,"score_gpt":0.2559175215299839,"score_spread":0.2318790654020301,"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."}}