{"id":"W2536295050","doi":"10.1109/fskd.2016.7603424","title":"Chinese term extraction from web pages based on expected point-wise mutual information","year":2016,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Term (time); Lexicon; Point (geometry); Computer science; Information retrieval; Word (group theory); Mutual information; Information extraction; Artificial intelligence; Natural language processing; Precision and recall; Data mining; 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":[0.001170793,0.00111017,0.001443758,0.01141443,0.0008347354,0.001327083,0.001038859,0.0006905856,0.001981395],"category_scores_gemma":[0.006329372,0.0004354045,0.001723039,0.007967515,0.0004995141,0.002627999,0.001255494,0.0006302359,0.001631268],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007014879,"about_ca_system_score_gemma":0.001742513,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003328665,"about_ca_topic_score_gemma":0.003954977,"domain_scores_codex":[0.9976495,0.0002463826,0.0004132114,0.0006308152,0.000883033,0.0001770493],"domain_scores_gemma":[0.9973605,0.0009930026,0.0003932268,0.0002771236,0.000898557,0.00007754852],"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":[0.0006839393,0.0001592926,0.01554296,0.001082455,0.0004316655,0.001216501,0.0007443553,0.0138221,0.06052471,0.009368749,0.01023966,0.8861836],"study_design_scores_gemma":[0.0001303295,0.0005166667,0.05597928,0.0001864533,0.0009906632,0.004165007,0.0004356738,0.7707341,0.1112919,0.03051359,0.02467657,0.0003797583],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.122727,0.002811322,0.8564125,0.0003074667,0.0001219875,0.0005044491,0.004840024,0.006224059,0.006051239],"genre_scores_gemma":[0.4868327,0.001431028,0.4961217,0.0001148741,0.0002405007,0.0006799693,0.009874689,0.000372167,0.004332436],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01141443,"threshold_uncertainty_score":0.006628454,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.006995051740656491,"score_gpt":0.2656118688887918,"score_spread":0.2586168171481353,"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."}}