{"id":"W2603902857","doi":"10.22230/src.2017v8n1a268","title":"Automatically Determining Versions of Scholarly Articles","year":2017,"lang":"en","type":"article","venue":"Scholarly and Research Communication","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Arcadia Fund","keywords":"Pairwise comparison; Computer science; Information retrieval; Simple (philosophy); Data science; Data mining; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.003705859,0.00006590864,0.0001081775,0.0001606334,0.002558451,0.005436992,0.003436649,0.00007153966,0.000005684023],"category_scores_gemma":[0.002672013,0.00005980852,0.00002664702,0.0002161742,0.0004727991,0.01016461,0.002482889,0.0006492275,0.00003276931],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002137989,"about_ca_system_score_gemma":0.0001052871,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001212694,"about_ca_topic_score_gemma":0.00002243831,"domain_scores_codex":[0.9985085,0.0002918641,0.0002201666,0.000230424,0.0005251442,0.000223866],"domain_scores_gemma":[0.9956078,0.0004330931,0.0001300837,0.002935676,0.000759155,0.0001341733],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.000005577564,0.0001600931,0.007759904,0.00002753013,0.00002386471,0.000002316728,0.001916258,9.906861e-7,0.01000607,0.261402,0.0005568634,0.7181386],"study_design_scores_gemma":[0.001217653,0.0003733126,0.7094072,0.000578384,0.00001875565,0.00002619907,0.001168443,0.1553072,0.01156432,0.105568,0.01436732,0.000403151],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8826314,0.001195699,0.07828483,0.02556094,0.00007494242,0.00046447,0.00002042737,0.0001651949,0.01160208],"genre_scores_gemma":[0.8334022,0.0003102334,0.1661032,0.0000266102,0.00001024598,0.00002531412,0.000003649464,0.000004523974,0.0001140611],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7177354,"threshold_uncertainty_score":0.9987401,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.137466202228515,"score_gpt":0.4056484116769164,"score_spread":0.2681822094484014,"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."}}