{"id":"W2784269545","doi":"10.1007/s10115-017-1147-9","title":"Localized user-driven topic discovery via boosted ensemble of nonnegative matrix factorization","year":2018,"lang":"en","type":"article","venue":"Knowledge and Information Systems","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"National Research Foundation of Korea","keywords":"Matrix decomposition; Computer science; Factorization; Matrix (chemical analysis); Non-negative matrix factorization; Mathematics; Artificial intelligence; Algorithm; Physics","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":[0.003228977,0.00179771,0.002594737,0.002582957,0.001307519,0.001894168,0.002627161,0.002087228,0.002144259],"category_scores_gemma":[0.009759893,0.0007555356,0.00203017,0.002609742,0.0007990813,0.003701201,0.002373175,0.002362843,0.002109441],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005862626,"about_ca_system_score_gemma":0.00180176,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005197224,"about_ca_topic_score_gemma":0.01069732,"domain_scores_codex":[0.9972031,0.001004944,0.0001427245,0.0007511267,0.0005854145,0.0003126662],"domain_scores_gemma":[0.9937567,0.003537664,0.0003369015,0.0006434082,0.001377392,0.0003479487],"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":[0.002054002,0.001287676,0.007850456,0.00070859,0.0007308743,0.0004143031,0.0009896205,0.1270237,0.04567176,0.01710174,0.03632962,0.7598377],"study_design_scores_gemma":[0.00002551644,0.00006022161,0.0004610524,0.0000137622,0.00006443336,0.00007319912,0.00004582179,0.9870304,0.002513256,0.008176457,0.001511013,0.00002479822],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02608386,0.00130507,0.9687497,0.0004799293,0.0003150174,0.00009856149,0.0004708659,0.001463213,0.001033853],"genre_scores_gemma":[0.5172196,0.001307504,0.4670597,0.000573029,0.00142973,0.0003934241,0.004744101,0.0004159832,0.006857112],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005197224,"threshold_uncertainty_score":0.01707667,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009477174681492609,"score_gpt":0.2722200081363226,"score_spread":0.26274283345483,"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."}}