{"id":"W7095329675","doi":"","title":"Combining Multiple Sources of Evidence in Web Information Extraction. Toronto 2008. In: Foundations of Intelligent Systems","year":2008,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"","keywords":"Information extraction; Ontology; Ranking (information retrieval); Selection (genetic algorithm); Domain (mathematical analysis); Web page; HTML element; Data extraction","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004342691,0.00006383196,0.0001728923,0.0002109396,0.00004050007,0.00003547101,0.0003551063,0.00003280985,0.00001600754],"category_scores_gemma":[0.0002377203,0.00005333831,0.00003517091,0.0004507324,0.00003416217,0.003008919,0.00008438814,0.00005337586,0.000009895182],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007894713,"about_ca_system_score_gemma":0.0001025694,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0125312,"about_ca_topic_score_gemma":0.004148301,"domain_scores_codex":[0.9989205,0.00005671632,0.000573207,0.0001134159,0.0002302917,0.0001058745],"domain_scores_gemma":[0.9990519,0.0002942022,0.0002241668,0.0002963502,0.0001060926,0.00002726895],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002650435,0.0004724565,0.8251199,0.0002409927,0.00006221073,0.000008338814,0.03132443,0.09464307,0.002203729,0.009922067,0.00112171,0.03485462],"study_design_scores_gemma":[0.0002419906,0.00005999483,0.03169112,0.0002983641,0.000004729387,0.00001386422,0.00365713,0.9612331,0.001569109,0.00001448604,0.001098567,0.0001175659],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6087403,0.0006870674,0.3885586,0.0001060087,0.0002068264,0.0001556016,0.000004769505,0.00004479941,0.001495986],"genre_scores_gemma":[0.9935773,0.000275619,0.006028957,0.00001010273,0.000007082972,0.000008913306,0.000005725974,0.000001474529,0.00008482305],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.86659,"threshold_uncertainty_score":0.9940444,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1072378682874622,"score_gpt":0.3088743515709479,"score_spread":0.2016364832834857,"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."}}