{"id":"W2927581056","doi":"10.1007/978-3-030-14799-0_7","title":"Robust Web Data Extraction Based on Unsupervised Visual Validation","year":2019,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université du Québec à Montréal","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Extractor; Robustness (evolution); Artificial intelligence; Data extraction; Information extraction; Information retrieval; Data mining; Machine learning","routes":{"ca_aff":true,"ca_fund":true,"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":["metaepi_narrow","open_science"],"consensus_categories":[],"category_scores_codex":[0.001528811,0.0005143325,0.0005071468,0.001268758,0.000249892,0.001018751,0.005760861,0.0003185728,0.00006407912],"category_scores_gemma":[0.0001670146,0.0004714309,0.000115043,0.0008934349,0.0002350387,0.001511995,0.001518382,0.0007934712,0.0002516586],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002361124,"about_ca_system_score_gemma":0.001120612,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002681882,"about_ca_topic_score_gemma":0.0000386108,"domain_scores_codex":[0.9949993,0.00007559266,0.0005106398,0.002460624,0.001426032,0.0005278136],"domain_scores_gemma":[0.9947106,0.0007036473,0.0003359583,0.003909971,0.0001929467,0.0001469418],"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":[0.00001159642,0.00008825898,0.0001209916,0.00004106289,0.00001853164,0.00004839635,0.00008104999,0.4428529,0.000388199,0.001537424,0.0002122894,0.5545993],"study_design_scores_gemma":[0.0002909577,0.0001672006,0.00008279558,0.0003145661,0.0000213551,0.00001108702,1.30749e-7,0.9945709,0.0005776931,0.001182418,0.002244058,0.0005368135],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001289831,0.00005325288,0.9945608,0.0008732747,0.001576915,0.0002375148,0.00004743822,0.0001842472,0.002337578],"genre_scores_gemma":[0.3783059,0.00004944811,0.6154259,0.003081483,0.001154386,0.000007025208,0.0008818171,0.00008394365,0.001010078],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.5540625,"threshold_uncertainty_score":0.9997737,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05360116226821633,"score_gpt":0.2863784460066002,"score_spread":0.2327772837383839,"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."}}