{"id":"W4386919265","doi":"10.1007/978-3-031-43849-3_5","title":"PreprintResolver: Improving Citation Quality by Resolving Published Versions of ArXiv Preprints Using Literature Databases","year":2023,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Academic Publishing and Open Access","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Preprint; Computer science; Field (mathematics); Citation; Information retrieval; Quality (philosophy); World Wide Web; Sample (material); 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":["metaresearch","metaepi_narrow","scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.01543673,0.0004681922,0.0008136839,0.001777639,0.0004408657,0.003057971,0.006247127,0.0005142566,0.0001685377],"category_scores_gemma":[0.03017912,0.0003884916,0.0002110591,0.002755651,0.0008307227,0.004903849,0.00447712,0.001722097,0.00004964035],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002840588,"about_ca_system_score_gemma":0.000749959,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004955934,"about_ca_topic_score_gemma":0.0001379181,"domain_scores_codex":[0.991183,0.0002688609,0.001724682,0.002745809,0.003447331,0.0006302918],"domain_scores_gemma":[0.9883553,0.005964318,0.001620141,0.002521161,0.001301643,0.0002374597],"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":[0.0001992832,0.0001651443,0.0364518,0.0004169807,0.00009716404,0.00009804603,0.008513859,0.1359161,0.01357056,0.01352945,0.007537919,0.7835037],"study_design_scores_gemma":[0.0008619854,0.00006470521,0.006441297,0.003758932,0.00005294044,0.00003519341,0.00002901078,0.6692381,0.008134804,0.3067671,0.002793289,0.001822641],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0154226,0.0003215031,0.9780643,0.0005755512,0.002207353,0.000531599,0.0002424438,0.000124112,0.002510554],"genre_scores_gemma":[0.879586,0.0000812028,0.1142267,0.0008012517,0.0007463391,0.00001269745,0.0001908939,0.0001033088,0.00425165],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8641633,"threshold_uncertainty_score":0.9998567,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1402033635676231,"score_gpt":0.3910124678096511,"score_spread":0.2508091042420281,"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."}}