{"id":"W2093641143","doi":"10.3115/1073445.1073477","title":"Frequency estimates for statistical word similarity measures","year":2003,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":197,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Coordenação de Aperfeiçoamento de Pessoal de Nível Superior","keywords":"Computer science; Word (group theory); Word lists by frequency; Natural language processing; Similarity (geometry); Artificial intelligence; Context (archaeology); Synonym (taxonomy); Semantic similarity; Set (abstract data type); Word2vec; Information retrieval; Sentence; Linguistics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002857899,0.00007827518,0.00009845429,0.00002554549,0.00007943522,0.00009188594,0.0003033869,0.00003655053,0.00004426469],"category_scores_gemma":[0.0005270945,0.0000661494,0.00002651068,0.00006850697,0.00001956437,0.0001668878,0.00003151335,0.00005590311,0.00001448073],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001979223,"about_ca_system_score_gemma":0.00006418024,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000308717,"about_ca_topic_score_gemma":0.00002354287,"domain_scores_codex":[0.9991988,0.00002585544,0.0001427738,0.0002611232,0.0001497256,0.0002217287],"domain_scores_gemma":[0.9992789,0.0002468654,0.00001911335,0.0003144832,0.00006746203,0.00007315562],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[4.051052e-7,0.00001652836,0.0009478842,0.000006559969,0.000004534583,0.000001880945,0.00004120335,0.00007353079,0.0001223096,0.9807369,0.0005869326,0.01746137],"study_design_scores_gemma":[0.0002061303,0.00002506684,0.0007117583,0.000004838299,0.000005503575,0.000006443084,0.000007702261,0.2219824,0.002618686,0.7719571,0.002316861,0.0001575175],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0006352463,0.00005962188,0.9899328,0.000375394,0.0001924441,0.0001255205,0.000002624944,0.0001463073,0.008530051],"genre_scores_gemma":[0.2830199,0.000001097894,0.716642,0.000216724,0.0000131011,0.00001336255,6.384854e-7,0.000003304366,0.00008982992],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2823847,"threshold_uncertainty_score":0.2697494,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05852592657578515,"score_gpt":0.2991449372864159,"score_spread":0.2406190107106308,"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."}}