{"id":"W2004193186","doi":"10.1145/584931.584940","title":"A framework for web table mining","year":2002,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":37,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Computer science; Table (database); Web page; Information retrieval; World Wide Web; Web mining; Table of contents; Information extraction; Process (computing); Data mining","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":[0.008295611,0.001865742,0.001893133,0.009678522,0.002424001,0.008419906,0.006132294,0.002592888,0.005897294],"category_scores_gemma":[0.02135476,0.001689631,0.005261851,0.01001889,0.003422442,0.009660353,0.004774561,0.00390068,0.004698471],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001853904,"about_ca_system_score_gemma":0.003942176,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007201462,"about_ca_topic_score_gemma":0.008264365,"domain_scores_codex":[0.9912164,0.002539047,0.001144468,0.001429641,0.003341884,0.0003285606],"domain_scores_gemma":[0.9902029,0.004705448,0.0007015966,0.002436866,0.001570432,0.0003826529],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008990621,0.0002488307,0.002647809,0.001042554,0.0002775374,0.000910937,0.001074759,0.03867513,0.002343336,0.5923201,0.03732972,0.3230394],"study_design_scores_gemma":[0.00004708382,0.00005812757,0.0006027419,0.0003705869,0.00007639646,0.0009875651,0.0003315989,0.2066089,0.002091391,0.6479988,0.1407405,0.00008641461],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0003530056,0.0003522526,0.9953374,0.0003607813,0.00004295688,0.0002504944,0.0006239568,0.001637692,0.001041447],"genre_scores_gemma":[0.006604623,0.0003582253,0.9901085,0.0001288751,0.00006463225,0.0003205226,0.001514374,0.0001442571,0.0007560035],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009678522,"threshold_uncertainty_score":0.04387188,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04721870685743441,"score_gpt":0.2677927006460024,"score_spread":0.220573993788568,"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."}}