{"id":"W4280650547","doi":"10.5281/zenodo.6564285","title":"D3.7 Technical report on information extraction from heterogeneous data using TDM","year":2022,"lang":"en","type":"report","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Neural Networks and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Canarie","funders":"European Commission","keywords":"Extraction (chemistry); Computer science; Data extraction; Information retrieval; Data science; Data mining; Chromatography; Chemistry; MEDLINE","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":[0.01605384,0.002057941,0.001598312,0.005490558,0.001668018,0.00850716,0.003631047,0.001823701,0.05553755],"category_scores_gemma":[0.04464716,0.001729752,0.00344689,0.004809089,0.001173188,0.007702809,0.009582696,0.003843159,0.08480541],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002033372,"about_ca_system_score_gemma":0.007178319,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008900067,"about_ca_topic_score_gemma":0.005089577,"domain_scores_codex":[0.9833789,0.003330227,0.001957316,0.002509382,0.008096207,0.0007278849],"domain_scores_gemma":[0.9711604,0.008738071,0.0009158079,0.009360868,0.008947285,0.000877573],"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":[0.0004366089,0.0003728443,0.002210993,0.003179051,0.0003360798,0.0009838117,0.001725738,0.005696891,0.0254117,0.02790435,0.3844897,0.5472522],"study_design_scores_gemma":[0.0001712531,0.0001847193,0.002413185,0.0008016514,0.000111829,0.0006860119,0.0008924107,0.03421309,0.04943601,0.02833833,0.8825685,0.0001830612],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"other","genre_scores_codex":[0.003724477,0.0008578237,0.8459162,0.002979134,0.001074931,0.002157291,0.0561812,0.0655255,0.0215835],"genre_scores_gemma":[0.01456009,0.001357139,0.7429399,0.001038355,0.0004142297,0.002919238,0.1929536,0.01559744,0.02822001],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.05553755,"threshold_uncertainty_score":0.1857916,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1310883463557256,"score_gpt":0.3304839633735825,"score_spread":0.1993956170178569,"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."}}