{"id":"W2974383016","doi":"10.48550/arxiv.1909.07512","title":"Short-Text Classification Using Unsupervised Keyword Expansion","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Sentence; Word (group theory); Artificial intelligence; Natural language processing; Feature (linguistics); Limiting; Process (computing); Speech recognition; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001131478,0.0009842522,0.0006603433,0.00182548,0.0003409637,0.0006168058,0.001008443,0.0007696468,0.001812132],"category_scores_gemma":[0.004084389,0.0002943576,0.0009050278,0.001412285,0.0004244654,0.00208131,0.0007698522,0.000996022,0.002006248],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005130588,"about_ca_system_score_gemma":0.000613977,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001479089,"about_ca_topic_score_gemma":0.002275762,"domain_scores_codex":[0.9991013,0.0002643471,0.00008876177,0.0003350077,0.0001519225,0.00005864373],"domain_scores_gemma":[0.9962354,0.002020989,0.0003767376,0.0004560839,0.0008047507,0.0001060843],"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":[0.001020891,0.0004761655,0.006929855,0.0004587723,0.0001634917,0.0004375403,0.000647795,0.0841071,0.08408882,0.003761262,0.01159312,0.8063152],"study_design_scores_gemma":[0.00005442341,0.0003252648,0.002669354,0.00002447709,0.00005049295,0.0002802524,0.0001380982,0.9588663,0.02717499,0.005609988,0.004765992,0.00004043915],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2174885,0.001024971,0.7652746,0.0003157049,0.0002428146,0.0007461616,0.001874903,0.009675572,0.003356811],"genre_scores_gemma":[0.5811419,0.0004456213,0.400759,0.0002106601,0.0003171104,0.0008886653,0.008168033,0.0004823482,0.007586699],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00182548,"threshold_uncertainty_score":0.00606215,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.196499995122714,"score_gpt":0.2195294196263386,"score_spread":0.02302942450362458,"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."}}