{"id":"W2137580272","doi":"10.1007/s11280-005-1544-y","title":"Studying the XML Web: Gathering Statistics from an XML Sample","year":2005,"lang":"en","type":"article","venue":"World Wide Web","topic":"Advanced Database Systems and Queries","field":"Computer Science","cited_by":76,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"Institut national de recherche en informatique et en automatique (INRIA)","keywords":"Computer science; XML validation; Efficient XML Interchange; XML Schema Editor; XML Base; Streaming XML; Document Structure Description; XML Signature; XML Encryption; Information retrieval; World Wide Web; XML database; cXML; XML; Database","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":[0.00637815,0.000619735,0.0008321802,0.0154419,0.0009221823,0.001837293,0.0008439264,0.0009035427,0.002490565],"category_scores_gemma":[0.05966192,0.0004400713,0.0005672657,0.01250546,0.0005470302,0.002427479,0.001359418,0.0008229456,0.002224816],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005091887,"about_ca_system_score_gemma":0.001602065,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004158435,"about_ca_topic_score_gemma":0.004572113,"domain_scores_codex":[0.9937411,0.002131283,0.0009984407,0.0007623306,0.002065803,0.0003010466],"domain_scores_gemma":[0.9152558,0.05405458,0.006135259,0.01096479,0.01156157,0.002028003],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0005443902,0.001116412,0.7510663,0.0005131999,0.0003351653,0.0005404495,0.003314149,0.00557311,0.01244583,0.004953764,0.02837483,0.1912224],"study_design_scores_gemma":[0.0001712047,0.001544345,0.7506283,0.0002320505,0.0006429473,0.002322506,0.01238585,0.09675982,0.03256545,0.02192013,0.0806281,0.0001993748],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8108255,0.0006182778,0.1201333,0.000585046,0.00004374174,0.0009576763,0.05804168,0.002633093,0.006161732],"genre_scores_gemma":[0.7689009,0.0007268259,0.1081321,0.0002989931,0.0001652566,0.001914315,0.1163044,0.0005810222,0.002976209],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0154419,"threshold_uncertainty_score":0.03373128,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02279868175895083,"score_gpt":0.2698407601171572,"score_spread":0.2470420783582064,"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."}}