{"id":"W2075429730","doi":"10.1145/1151454.1151483","title":"An extensible text extraction tool for learning objects","year":2006,"lang":"en","type":"article","venue":"","topic":"Advanced Database Systems and Queries","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of New Brunswick","funders":"","keywords":"Computer science; XML; Interoperability; Document Structure Description; World Wide Web; Information retrieval; Context (archaeology); Information extraction; XML Base; Efficient XML Interchange; XML validation; XML framework; Document type definition","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":[0.001574631,0.00113163,0.0008454625,0.003976679,0.0007170226,0.002138231,0.001714624,0.0009543074,0.008024794],"category_scores_gemma":[0.004829873,0.0006687221,0.001069326,0.002620163,0.0003538238,0.003225523,0.001889772,0.001190945,0.006882676],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004825405,"about_ca_system_score_gemma":0.001215368,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001732013,"about_ca_topic_score_gemma":0.002127259,"domain_scores_codex":[0.9987805,0.0001417908,0.0002919266,0.0002429937,0.0004819588,0.00006075202],"domain_scores_gemma":[0.9981163,0.0007806711,0.0001819388,0.0003349351,0.000501061,0.00008520314],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004315908,0.0002376057,0.002582512,0.001652456,0.00009644452,0.001530224,0.0008990644,0.002984156,0.06353666,0.01466626,0.04518939,0.8661937],"study_design_scores_gemma":[0.0002683862,0.0002709647,0.007448005,0.0007336552,0.0002309896,0.003959738,0.0004888624,0.07071251,0.150703,0.01842556,0.7465122,0.0002461612],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.006391115,0.0004327355,0.9053619,0.0002023406,0.0001361917,0.000590326,0.005771191,0.07703573,0.004078424],"genre_scores_gemma":[0.03593843,0.0006440317,0.9155935,0.0002381809,0.00007011132,0.0005619526,0.02757488,0.006355868,0.01302314],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008024794,"threshold_uncertainty_score":0.02684557,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01049118711209562,"score_gpt":0.2735226169605028,"score_spread":0.2630314298484072,"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."}}