{"id":"W32038199","doi":"10.3389/fnhum.2019.00463","title":"XML Query Support for Web Information Extraction: A Study on HTML Element Depth Distribution ?","year":2005,"lang":"en","type":"article","venue":"Frontiers in Human Neuroscience","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Information retrieval; XML; Database; Information extraction; XML database; Document Structure Description; Web query classification; Point (geometry); Database query; Data mining; Search engine; Web search query; World Wide Web; Mathematics","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007033052,0.0001383648,0.0001569219,0.0002798318,0.0003482183,0.0003430205,0.0008527042,0.00003185933,0.00000294197],"category_scores_gemma":[0.0001127872,0.0001352775,0.00005765154,0.0006375855,0.00005736571,0.002318428,0.0001166094,0.0001375786,0.000005462757],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001616573,"about_ca_system_score_gemma":0.00007652576,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008540816,"about_ca_topic_score_gemma":0.00002380072,"domain_scores_codex":[0.9983233,0.00006090188,0.0003932366,0.0004425847,0.0004469347,0.0003330492],"domain_scores_gemma":[0.9991473,0.0000248593,0.0001712262,0.00052763,0.00005040938,0.00007860495],"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":[0.0001275253,0.004406453,0.2277142,0.00007343289,0.00003672275,0.00005598118,0.006512986,0.03742964,0.006194792,0.01764251,0.3300268,0.3697789],"study_design_scores_gemma":[0.001630703,0.001556477,0.1576758,0.00002711952,0.0000251482,0.00001064687,0.0008170265,0.6294112,0.001340564,0.0002615917,0.2067081,0.0005357062],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08210642,0.000004958084,0.9155542,0.000416326,0.00101605,0.0004056815,0.00003971628,0.0001017107,0.0003548969],"genre_scores_gemma":[0.992101,0.000003619774,0.007011799,0.0004809064,0.00007357873,0.00008412699,0.00005115847,0.000003944053,0.0001899335],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9099945,"threshold_uncertainty_score":0.5516455,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02458695769088694,"score_gpt":0.2978415342857985,"score_spread":0.2732545765949115,"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."}}