{"id":"W2102247866","doi":"10.1109/dexa.2006.62","title":"Finding Syntactic Similarities Between XML Documents","year":2006,"lang":"en","type":"article","venue":"","topic":"Advanced Database Systems and Queries","field":"Computer Science","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; XML; Document Structure Description; Edit distance; XML validation; Information retrieval; Tree (set theory); Efficient XML Interchange; Simple API for XML; Tree structure; XML Schema (W3C); XML database; Document type definition; Theoretical computer science; Data mining; XML Signature; Algorithm; Programming language; Data structure; World Wide Web; Mathematics","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":[0.0001144106,0.0001017461,0.0001311949,0.00006087643,0.0001520823,0.0000978301,0.0002503782,0.00002905576,0.00003152194],"category_scores_gemma":[0.00001384989,0.00008575951,0.00003369384,0.0001515404,0.00002360366,0.001312379,0.0001892355,0.00006276834,0.0001096274],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003206707,"about_ca_system_score_gemma":0.00001912875,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004539732,"about_ca_topic_score_gemma":0.00004515751,"domain_scores_codex":[0.999121,0.00002420078,0.0001913934,0.0002350921,0.0001916331,0.0002367335],"domain_scores_gemma":[0.9994256,0.00009427683,0.00005597688,0.0003589883,0.00002451044,0.00004059548],"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":[9.202986e-7,0.00001713082,0.01997091,0.00002845474,0.00001323849,0.00002138592,0.00007832437,0.00004780529,0.0003197731,0.9707378,0.005014383,0.00374991],"study_design_scores_gemma":[0.000865848,0.0001255278,0.05032653,0.0001686752,0.00002264997,0.00004942803,0.0003034383,0.002083125,0.026444,0.03803592,0.8804193,0.001155558],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01190377,0.00007748731,0.9544133,0.000299046,0.000221456,0.00008477586,0.00001208158,0.0002361945,0.0327519],"genre_scores_gemma":[0.8785772,0.000003691137,0.1151221,0.0001174893,0.0001977706,0.00001019894,0.00001655829,0.000007546804,0.005947421],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9327018,"threshold_uncertainty_score":0.3497171,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01781436096921012,"score_gpt":0.2641840172389015,"score_spread":0.2463696562696914,"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."}}