{"id":"W2138757450","doi":"","title":"Structure and content scoring for XML","year":2005,"lang":"en","type":"article","venue":"","topic":"Advanced Database Systems and Queries","field":"Computer Science","cited_by":141,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo; University of Toronto","funders":"","keywords":"Computer science; XML; Information retrieval; Path (computing); Document Structure Description; Data mining; World Wide Web","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":[0.001827722,0.0007863416,0.0007786578,0.00387835,0.0007033438,0.001799091,0.001335893,0.0009551534,0.005243284],"category_scores_gemma":[0.01658827,0.0003126331,0.0006826518,0.003980415,0.0006739586,0.002972547,0.001378384,0.0007096981,0.002601563],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009779301,"about_ca_system_score_gemma":0.001072415,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002780962,"about_ca_topic_score_gemma":0.002926135,"domain_scores_codex":[0.9973413,0.0006631713,0.0002569599,0.0003219948,0.001276733,0.0001397767],"domain_scores_gemma":[0.994696,0.002068621,0.0005097742,0.001019111,0.001557179,0.0001494443],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003771005,0.0001121101,0.008672372,0.0005060075,0.00007807301,0.0002831151,0.0004531524,0.04782709,0.02252017,0.08922878,0.01905646,0.8108855],"study_design_scores_gemma":[0.00008512305,0.0002941484,0.004844822,0.00009296589,0.0000668841,0.0009027658,0.0002998294,0.7899024,0.03464661,0.1333624,0.03542045,0.00008147518],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02660325,0.0006114143,0.9624734,0.0002956537,0.0000728286,0.0002227707,0.001137815,0.004551981,0.004030873],"genre_scores_gemma":[0.2358012,0.0005024999,0.7524949,0.000105755,0.0001466188,0.0003103229,0.004091637,0.0005135306,0.006033586],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005243284,"threshold_uncertainty_score":0.01754051,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03575117782653841,"score_gpt":0.2541037418922822,"score_spread":0.2183525640657438,"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."}}