{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003866053,0.0006244658,0.0008699847,0.004921817,0.0007895717,0.002185482,0.0007748994,0.0009730871,0.03599919],"category_scores_gemma":[0.02564164,0.0003530517,0.0008185764,0.005867163,0.0003132468,0.004166025,0.001521525,0.0005020885,0.01449959],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005561721,"about_ca_system_score_gemma":0.001395484,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009378744,"about_ca_topic_score_gemma":0.007940364,"domain_scores_codex":[0.9974986,0.0008686942,0.0004945481,0.0003629355,0.0006253592,0.0001498897],"domain_scores_gemma":[0.9789691,0.01646156,0.0008311243,0.001206241,0.002174743,0.000357258],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.002825925,0.001214618,0.06697768,0.008310268,0.0001904214,0.002514631,0.008784467,0.0005099105,0.01283691,0.004080547,0.1194398,0.7723148],"study_design_scores_gemma":[0.0006228997,0.001075181,0.2369374,0.002426693,0.0008399337,0.005637427,0.01210969,0.0226782,0.01911931,0.005703909,0.6925781,0.0002712076],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5884387,0.008049925,0.06947382,0.004947411,0.0002957962,0.0057214,0.1905856,0.03440826,0.0980791],"genre_scores_gemma":[0.6020726,0.00595968,0.1182329,0.002642754,0.0002398943,0.002705326,0.2075273,0.008028701,0.05259085],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03599919,"threshold_uncertainty_score":0.1204293,"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."}}