{"id":"W2138954094","doi":"10.1145/1097047.1097059","title":"Narrative text classification for automatic key phrase extraction in web document corpora","year":2005,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Automatic summarization; Information retrieval; Plain text; Key (lock); Natural language processing; Phrase; tf–idf; Artificial intelligence; Ranking (information retrieval); Document clustering; Information extraction; HTML element; Web page; Keyword extraction; Cluster analysis; World Wide Web; Term (time)","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":[0.0003379104,0.0001223924,0.0001556581,0.0002371328,0.00007702831,0.00009196351,0.0003497747,0.00005572673,0.00006631777],"category_scores_gemma":[0.00005061362,0.0001105642,0.00006086158,0.0004438656,0.00002452244,0.001483573,0.00004581281,0.0000853126,0.00005019926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002980529,"about_ca_system_score_gemma":0.00005655519,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001050066,"about_ca_topic_score_gemma":0.0003320111,"domain_scores_codex":[0.998867,0.00004840355,0.0003539338,0.0003541282,0.0001921637,0.0001844223],"domain_scores_gemma":[0.9991518,0.00009365062,0.0002052135,0.0004122195,0.00008830085,0.00004886623],"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":[0.00001263192,0.0003618763,0.0003174042,0.00001910172,0.00002548563,0.000003077579,0.005403575,0.0006599881,0.04071084,0.2299132,0.004983912,0.7175889],"study_design_scores_gemma":[0.00039267,0.00007072539,0.001284575,0.00002067634,0.000007207837,0.000004262358,0.0004689286,0.9514095,0.01820568,0.02024778,0.007676095,0.0002118791],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04118503,0.00002473838,0.9530451,0.002940278,0.00003683452,0.0004721308,4.162337e-7,0.0004843523,0.001811073],"genre_scores_gemma":[0.6076561,0.000008658535,0.3913634,0.0001454497,0.00002346424,0.0001634593,0.000003916133,0.000004881918,0.0006307636],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9507495,"threshold_uncertainty_score":0.4508676,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02461358341141394,"score_gpt":0.3336148891383208,"score_spread":0.3090013057269069,"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."}}