{"id":"W2519452335","doi":"10.1145/2938503.2938522","title":"Comparative Mining of B2C Web Sites by Discovering Web Database Schemas","year":2016,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science; Web page; Data Web; Document Object Model; Web mining; Web modeling; Information retrieval; World Wide Web; Schema matching; Web mapping; Parsing; Web intelligence; Static web page; Database; Web navigation; Data integration; Artificial intelligence","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":[0.002081621,0.000497635,0.0005871681,0.0170166,0.0009968365,0.002942436,0.001353153,0.00103663,0.003122551],"category_scores_gemma":[0.008310943,0.0003820738,0.001339754,0.01465238,0.0004176387,0.003366409,0.000860844,0.000551718,0.001686489],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001026136,"about_ca_system_score_gemma":0.001423236,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01243191,"about_ca_topic_score_gemma":0.01897885,"domain_scores_codex":[0.9977946,0.0006090822,0.0002564671,0.0004327392,0.0007346137,0.0001724506],"domain_scores_gemma":[0.9935831,0.002664984,0.0004818079,0.001195867,0.001730672,0.0003434066],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001321851,0.00144723,0.4705596,0.001809228,0.001022132,0.002014725,0.002289174,0.02730102,0.02627345,0.0281244,0.01974609,0.4180911],"study_design_scores_gemma":[0.0002205027,0.0006280506,0.3883229,0.0005612441,0.0007520877,0.003884895,0.009012328,0.432717,0.05818339,0.03575268,0.06976087,0.000204048],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8486874,0.002551322,0.09123208,0.0004830316,0.00006091135,0.0009973012,0.02653279,0.001798154,0.02765711],"genre_scores_gemma":[0.7886619,0.001099748,0.1585128,0.0001446518,0.00002421407,0.0003474473,0.04886451,0.0003348226,0.002009765],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.0170166,"threshold_uncertainty_score":0.02471912,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03579551979941218,"score_gpt":0.2790814370437026,"score_spread":0.2432859172442904,"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."}}