{"id":"W2593276484","doi":"10.22148/16.016","title":"Topic Modeling the Hàn diăn Ancient Classics","year":2017,"lang":"en","type":"article","venue":"Journal of Cultural Analytics","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Pinyin; Topic model; Pronunciation; Linguistics; Computer science; History; Natural language processing; Artificial intelligence; Chinese characters; Philosophy","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002198717,0.0005015676,0.0005685415,0.002024701,0.000871254,0.002608806,0.0009362695,0.0005965232,0.003063044],"category_scores_gemma":[0.007810347,0.0003381374,0.0008773295,0.002806329,0.0007840871,0.002148851,0.0008891819,0.0008942761,0.0006116273],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001332701,"about_ca_system_score_gemma":0.001044002,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0138652,"about_ca_topic_score_gemma":0.01228507,"domain_scores_codex":[0.9991375,0.0003685257,0.00004611554,0.0002652532,0.0001152912,0.00006722633],"domain_scores_gemma":[0.9971578,0.002221192,0.000117862,0.000199611,0.000220927,0.00008249606],"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":[0.0005950432,0.0002277954,0.0457805,0.0008948742,0.0005444603,0.00105821,0.01239796,0.1971034,0.003590525,0.3305866,0.03227026,0.3749503],"study_design_scores_gemma":[0.00003887504,0.00003980461,0.01470924,0.0001007911,0.0001108935,0.0003053054,0.001544872,0.8425754,0.001340493,0.1004448,0.03873299,0.0000565524],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3903582,0.007669891,0.5675967,0.0046233,0.0003300357,0.0002831144,0.005244947,0.001074815,0.02281902],"genre_scores_gemma":[0.9041876,0.002217297,0.07617199,0.0001936492,0.0003211694,0.0002799815,0.003335346,0.0001786841,0.01311419],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0138652,"threshold_uncertainty_score":0.027569,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1623321213294615,"score_gpt":0.4360665115014897,"score_spread":0.2737343901720282,"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."}}