{"id":"W4223909614","doi":"10.13052/jwe1540-9589.2146","title":"Integrated-Block: A New Combination Model to Improve Web Page Segmentation","year":2022,"lang":"en","type":"article","venue":"Journal of Web Engineering","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Web page; Segmentation; Block (permutation group theory); Information retrieval; Gestalt psychology; Context (archaeology); Similarity (geometry); Artificial intelligence; Data mining; World Wide Web; Perception; Image (mathematics)","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":[0.001800909,0.001625409,0.002119938,0.00500955,0.0009183322,0.001971254,0.002721216,0.002094685,0.004246732],"category_scores_gemma":[0.003237262,0.0007439854,0.002584241,0.003843904,0.0009865579,0.004735338,0.001776848,0.001475355,0.002453414],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001496021,"about_ca_system_score_gemma":0.001659532,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01298431,"about_ca_topic_score_gemma":0.01651346,"domain_scores_codex":[0.9981086,0.0002977057,0.0001361581,0.0006818677,0.0005915398,0.000184103],"domain_scores_gemma":[0.9983247,0.0004858641,0.0001925613,0.0003063009,0.0005445155,0.0001460619],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001398551,0.0006344194,0.0136391,0.0005769581,0.000575305,0.0003965475,0.0006377402,0.173244,0.02717908,0.01017961,0.01535807,0.7561807],"study_design_scores_gemma":[0.0000418932,0.00016954,0.002488595,0.00002741456,0.0001478569,0.00015441,0.00005550942,0.9788289,0.005104729,0.00589769,0.007050008,0.00003334126],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03863754,0.001648757,0.9475656,0.0003330662,0.0001618295,0.000308955,0.0007113986,0.006795011,0.003837752],"genre_scores_gemma":[0.3548365,0.0009741093,0.6299925,0.0004251159,0.000234141,0.0006099103,0.003563849,0.00128729,0.00807664],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01298431,"threshold_uncertainty_score":0.02581745,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01158827969045619,"score_gpt":0.2223981886257918,"score_spread":0.2108099089353356,"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."}}