{"id":"W2077439276","doi":"10.1145/1806916.1806920","title":"Exploring XML web collections with DescribeX","year":2010,"lang":"en","type":"article","venue":"ACM Transactions on the Web","topic":"Advanced Database Systems and Queries","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa; Carleton University; University of Toronto","funders":"","keywords":"Computer science; XPath; Document Structure Description; Information retrieval; Schema (genetic algorithms); World Wide Web; XML; XML Schema Editor; RSS; XML validation","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001372864,0.0001354452,0.0001072509,0.0001201986,0.0009796661,0.00009501484,0.0005826778,0.00002998027,0.00009026778],"category_scores_gemma":[0.00002947516,0.00008588839,0.00005500628,0.0008295061,0.00008992138,0.0008948377,0.00001555604,0.0003927089,0.00008385716],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002130988,"about_ca_system_score_gemma":0.0001157241,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006084619,"about_ca_topic_score_gemma":0.001279282,"domain_scores_codex":[0.9991314,0.00003951104,0.0001372454,0.0002701004,0.0001964893,0.000225289],"domain_scores_gemma":[0.998259,0.0002247035,0.00004243938,0.001346274,0.00006420597,0.00006339556],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002173212,0.001072286,0.0001516238,0.00007813126,0.000504871,0.00008693759,0.006459376,0.007172239,0.1918724,0.6296836,0.00969691,0.1530043],"study_design_scores_gemma":[0.001061827,0.0004206353,0.0002014526,0.0001243145,0.00005128215,0.0003683819,0.001265322,0.01206422,0.07485662,0.002276234,0.9065317,0.0007780362],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02904557,0.000008081724,0.9640115,0.003431087,0.0009116463,0.0002120879,0.00003685867,0.0003037229,0.002039407],"genre_scores_gemma":[0.8894494,0.00006213823,0.1072843,0.000264558,0.00008089037,0.0003599112,0.000001383892,0.00001988926,0.002477464],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8968348,"threshold_uncertainty_score":0.7534899,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06726977821672953,"score_gpt":0.2344276321629866,"score_spread":0.1671578539462571,"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."}}