{"id":"W2940481073","doi":"10.1002/gdj3.62","title":"From books to bytes: A new data rescue tool","year":2019,"lang":"en","type":"article","venue":"Geoscience Data Journal","topic":"Time Series Analysis and Forecasting","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Social Sciences and Humanities Research Council of Canada; McGill University","keywords":"Metadata; Computer science; Data mapping; Open data; World Wide Web; Traceability; Linked data; Data science; Schema (genetic algorithms); Unstructured data; Context (archaeology); Information retrieval; Database; Data mining; Semantic Web; Software engineering; Big data","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"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":[0.007019712,0.001298751,0.0009227668,0.006959612,0.002016222,0.008563608,0.00420406,0.002255179,0.07616404],"category_scores_gemma":[0.02889423,0.001205641,0.001131115,0.005113652,0.002235559,0.01685907,0.01176968,0.003641143,0.03981312],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001276884,"about_ca_system_score_gemma":0.002240674,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002322265,"about_ca_topic_score_gemma":0.003462981,"domain_scores_codex":[0.9947246,0.001154886,0.0007164413,0.0007684637,0.002422641,0.0002130496],"domain_scores_gemma":[0.976777,0.01103274,0.0008685106,0.007055365,0.00304093,0.001225397],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002546802,0.0002080549,0.002268235,0.0009107007,0.00003993123,0.0008564396,0.006975127,0.001240073,0.004304464,0.05611542,0.4463008,0.4805262],"study_design_scores_gemma":[0.00005240351,0.00004850837,0.000718793,0.0003839043,0.00001651972,0.0004590695,0.001266431,0.004726262,0.004933144,0.02089565,0.9664033,0.00009604741],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.007872102,0.001081292,0.5952316,0.007180048,0.001507076,0.001259679,0.02806686,0.2702306,0.08757079],"genre_scores_gemma":[0.04475615,0.001318787,0.7739624,0.002642018,0.0006357136,0.001782624,0.03835168,0.05533195,0.0812187],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.07616404,"threshold_uncertainty_score":0.2547941,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0669314535514649,"score_gpt":0.2863183443244633,"score_spread":0.2193868907729984,"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."}}