{"id":"W1594382369","doi":"10.28945/509","title":"HTML Tags as Extraction Cues for Web Page Description Construction","year":2003,"lang":"en","type":"article","venue":"Informing Science The International Journal of an Emerging Transdiscipline","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":23,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Information retrieval; World Wide Web; Web page; HTML element","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":[],"consensus_categories":[],"category_scores_codex":[0.003189857,0.0001255405,0.0001455375,0.0005413833,0.0005114325,0.0006682325,0.001985938,0.00003336575,0.00001702906],"category_scores_gemma":[0.0003563735,0.00008848122,0.00014748,0.0005705276,0.000267461,0.005814309,0.0000514008,0.0001857258,0.000004460995],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001193628,"about_ca_system_score_gemma":0.0004802412,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001478733,"about_ca_topic_score_gemma":0.00001739085,"domain_scores_codex":[0.9978676,0.00003507966,0.0005908921,0.0002114368,0.001049237,0.000245714],"domain_scores_gemma":[0.9981986,0.0001059338,0.000495158,0.0002667141,0.0008220437,0.0001115127],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002143845,0.0003032475,0.0009652866,0.00002033964,0.0002689444,0.00004823936,0.01365564,0.01468797,0.5456177,0.1296242,0.0003588987,0.2942352],"study_design_scores_gemma":[0.003920715,0.001298967,0.002498605,0.0004895581,0.0002425972,0.005326375,0.01545792,0.6906124,0.1706464,0.0730566,0.03538243,0.001067515],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6251068,0.00002777491,0.3691182,0.002357173,0.002831127,0.00006377146,0.000005577382,0.00003003643,0.0004594829],"genre_scores_gemma":[0.9220325,0.00003239426,0.07745366,0.0001572903,0.0002450685,0.00000317671,0.000003774206,0.000005541634,0.0000666187],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6759244,"threshold_uncertainty_score":0.6443781,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02587826375411731,"score_gpt":0.3316387600285208,"score_spread":0.3057604962744035,"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."}}