{"id":"W4295630651","doi":"10.31274/jlsc.14421","title":"Free and Open-Source Automated Open Access Preprint Harvesting","year":2022,"lang":"en","type":"article","venue":"Journal of Librarianship and Scholarly Communication","topic":"Research Data Management Practices","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"","keywords":"Upload; World Wide Web; Computer science; Metadata; Identifier; JavaScript; License; Preprint; Unique identifier; Open source; MIT License; Process (computing); Scholarly communication; Publishing; Software; Political science; Operating system","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":["scholarly_communication","open_science"],"consensus_categories":["scholarly_communication","open_science"],"category_scores_codex":[0.008885168,0.00009033785,0.0001873761,0.0002526181,0.00102772,0.1140259,0.03015141,0.00002665476,0.00002191859],"category_scores_gemma":[0.001937818,0.0000854562,0.00001956813,0.0005773611,0.00005084979,0.3384196,0.0819072,0.001001741,5.963203e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004563528,"about_ca_system_score_gemma":0.0002266599,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003711958,"about_ca_topic_score_gemma":0.00001644746,"domain_scores_codex":[0.9965843,0.002086474,0.0004554039,0.0002434001,0.0004773187,0.0001530795],"domain_scores_gemma":[0.9963667,0.0007119132,0.0007900483,0.00184869,0.0001303817,0.0001522989],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","study_design_scores_codex":[0.0003362572,0.0006595137,0.1345733,0.00008504249,0.0003257011,0.00009098763,0.007191164,0.0008496274,0.001279539,0.6429393,0.0245004,0.1871692],"study_design_scores_gemma":[0.003852127,0.0006250067,0.5502833,0.0002143592,0.00004344817,0.0005035315,0.001932586,0.04023834,0.0002041536,0.04720837,0.3544514,0.0004433913],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5260912,0.006851469,0.2144137,0.2001819,0.0003585025,0.002628246,0.00003282449,0.0004024859,0.04903962],"genre_scores_gemma":[0.9014086,0.001198004,0.09533482,0.0009845505,0.00001943354,0.00003350031,0.000008394145,0.0000141009,0.0009986704],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5957309,"threshold_uncertainty_score":0.9750959,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2453683543165483,"score_gpt":0.4174049630888566,"score_spread":0.1720366087723082,"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."}}