{"id":"W4256184506","doi":"10.4018/978-1-60960-126-3.ch004","title":"Heterogeneous Text and Numerical Data Mining with Possible Applications in Business and Financial Sectors","year":2011,"lang":"en","type":"book-chapter","venue":"IGI Global eBooks","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ontario Tech University","funders":"","keywords":"Computer science; Semantics (computer science); Data type; Data science; Data processing; Qualitative property; Data modeling; Data mining; Database; Programming language; Machine learning","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":[0.001218322,0.0008472795,0.0008053001,0.003164822,0.0006612571,0.002778089,0.001542007,0.00073796,0.01165213],"category_scores_gemma":[0.002176699,0.0004312766,0.0009281067,0.009521321,0.0008518184,0.004081449,0.001297676,0.001355249,0.004648193],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007101501,"about_ca_system_score_gemma":0.0007544819,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005674644,"about_ca_topic_score_gemma":0.001140307,"domain_scores_codex":[0.9993687,0.0001164264,0.00004777258,0.0001591514,0.0002815556,0.00002636914],"domain_scores_gemma":[0.999208,0.0005383499,0.000038493,0.00007687614,0.0001030202,0.00003530467],"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":[0.00006413594,0.0001130583,0.0008510321,0.001451405,0.00005118652,0.0004712244,0.0006412313,0.003712998,0.004472488,0.2561487,0.0606855,0.671337],"study_design_scores_gemma":[0.00001767854,0.00003642251,0.0017504,0.001066755,0.00006300115,0.001284791,0.0003645547,0.02167744,0.004349729,0.3461226,0.6232171,0.00004954911],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0096898,0.09953389,0.733394,0.003814297,0.001880941,0.0002993784,0.002502807,0.001614681,0.1472702],"genre_scores_gemma":[0.04861524,0.05692996,0.7867255,0.001715237,0.001660156,0.0003588319,0.005379582,0.000625777,0.09798964],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01165213,"threshold_uncertainty_score":0.03898031,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02756939377203025,"score_gpt":0.2426782168533207,"score_spread":0.2151088230812905,"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."}}