{"id":"W2123409753","doi":"10.1111/0824-7935.00126","title":"Probability‐Based Chinese Text Processing and Retrieval","year":2000,"lang":"en","type":"article","venue":"Computational Intelligence","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"University of Waterloo; University of Regina","keywords":"Weighting; Computer science; Artificial intelligence; Natural language processing; Probabilistic logic; Word (group theory); Text processing; Term (time); Focus (optics); Relevance (law); Pattern recognition (psychology); Speech recognition; Information retrieval; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.001723614,0.0009585392,0.0009908661,0.004403182,0.001054592,0.0016279,0.001321614,0.0005829544,0.008137869],"category_scores_gemma":[0.007109631,0.0003382202,0.001308806,0.005435944,0.0007853059,0.003842239,0.0008572788,0.0005626644,0.004258061],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001355403,"about_ca_system_score_gemma":0.001563993,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006251512,"about_ca_topic_score_gemma":0.00463141,"domain_scores_codex":[0.9981535,0.0003821661,0.000223674,0.0002944662,0.0008258056,0.0001203437],"domain_scores_gemma":[0.9978517,0.0008106788,0.000210625,0.0003118746,0.0007615825,0.00005349378],"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":[0.0002168101,0.0001453938,0.001796992,0.001082654,0.0001854698,0.0002498588,0.0003316876,0.02328151,0.02744971,0.04037834,0.01522355,0.889658],"study_design_scores_gemma":[0.0001439911,0.0005937963,0.01014298,0.0001334539,0.0005032263,0.001629805,0.0002753282,0.7218625,0.1253987,0.04939311,0.08960221,0.0003209371],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0151883,0.002243943,0.9712799,0.0004254761,0.0002795964,0.0005387391,0.0004890904,0.002328055,0.007227033],"genre_scores_gemma":[0.2209738,0.003743188,0.7574415,0.0004093303,0.0005047112,0.001011898,0.002121903,0.0003179709,0.01347564],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008137869,"threshold_uncertainty_score":0.02722394,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02799854521753289,"score_gpt":0.3141129953579396,"score_spread":0.2861144501404067,"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."}}