{"id":"W2060864821","doi":"10.1300/j104v32n03_08","title":"Word Division in the Transcription of Chinese Script in the Title Fields of Bibliographic Records","year":2001,"lang":"en","type":"article","venue":"Cataloging & Classification Quarterly","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Pinyin; Computer science; Ambiguity; Natural language processing; Information retrieval; Word (group theory); Transcription (linguistics); Orthography; Linguistics; Artificial intelligence; Chinese characters","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":[0.001945405,0.0003616968,0.000347971,0.002049238,0.0006922435,0.0009953756,0.0002386698,0.0002146654,0.004086827],"category_scores_gemma":[0.022985,0.0001830352,0.0002075223,0.003842666,0.0005218812,0.0007632748,0.0005785829,0.0004688603,0.001905826],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009877232,"about_ca_system_score_gemma":0.00167013,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008472159,"about_ca_topic_score_gemma":0.01677606,"domain_scores_codex":[0.9977474,0.001021822,0.0003302353,0.0003621317,0.0004276609,0.0001107667],"domain_scores_gemma":[0.9843435,0.01002544,0.001468193,0.0007319844,0.003166909,0.0002639017],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00347139,0.0002030961,0.07992614,0.002930904,0.00008363745,0.001500121,0.03547338,0.002014279,0.1562513,0.003232877,0.01443296,0.7004799],"study_design_scores_gemma":[0.0002164451,0.001452617,0.5972849,0.0007224117,0.0004287158,0.002021927,0.0397893,0.05318101,0.2330307,0.003871036,0.06771342,0.0002875622],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.9643422,0.0007359066,0.01870108,0.0004537738,0.0002239183,0.0005355703,0.004416099,0.0005475804,0.01004393],"genre_scores_gemma":[0.9513107,0.0004219151,0.03786216,0.00008823931,0.00005361435,0.0002522407,0.005754795,0.0001076624,0.004148658],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.008472159,"threshold_uncertainty_score":0.01684564,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02613999809245618,"score_gpt":0.287983754464532,"score_spread":0.2618437563720759,"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."}}