{"id":"W2044194343","doi":"10.4018/jdwm.2012100101","title":"Towards Comparative Mining of Web Document Objects with NFA","year":2012,"lang":"en","type":"article","venue":"International Journal of Data Warehousing and Mining","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science; Web mining; Data Web; Web page; Web modeling; Information retrieval; Web standards; World Wide Web; Web intelligence; Static web page; Web navigation; Database","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008696044,0.000123862,0.0002861249,0.0002872798,0.00006482331,0.0001824,0.001293208,0.0000281279,0.000008112759],"category_scores_gemma":[0.00008982482,0.0000937255,0.00003924208,0.0001862629,0.00007195693,0.002472874,0.0005833093,0.0001164872,0.000001049248],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003596273,"about_ca_system_score_gemma":0.0001842414,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003213146,"about_ca_topic_score_gemma":0.00001092948,"domain_scores_codex":[0.9985283,0.00006508286,0.0004338858,0.0001848117,0.0006016225,0.0001863057],"domain_scores_gemma":[0.9984772,0.0001721898,0.0005957208,0.0003280161,0.0003092348,0.0001176383],"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":[0.0003501978,0.0006714729,0.06613788,0.00007834087,0.003785584,0.0003278278,0.07746766,0.001335528,0.006732796,0.002561865,0.006666624,0.8338842],"study_design_scores_gemma":[0.03005348,0.006674334,0.09047828,0.02091358,0.00348461,0.02233436,0.145774,0.4597222,0.05415137,0.001032676,0.1588925,0.006488613],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9021004,0.001128876,0.09463657,0.0004435628,0.0006182069,0.00002896711,0.00005841133,0.00001927587,0.0009657807],"genre_scores_gemma":[0.8008,0.00005439534,0.1988146,0.00006187973,0.0002291948,2.933105e-7,0.00001924127,0.000005096284,0.00001530908],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8273956,"threshold_uncertainty_score":0.3822015,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07416393370134935,"score_gpt":0.3429523755542617,"score_spread":0.2687884418529124,"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."}}