{"id":"W3012108798","doi":"10.2196/17644","title":"Document-Level Biomedical Relation Extraction Leveraging Pretrained Self-Attention Structure and Entity Replacement: Algorithm and Pretreatment Method Validation Study","year":2020,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":10,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Science Foundation of Guangdong Province; National Natural Science Foundation of China","keywords":"Relationship extraction; Computer science; Preprocessor; Relation (database); Semantics (computer science); Sentence; Context (archaeology); Noise (video); Artificial intelligence; Natural language processing; Heuristic; Data mining; Biomedical text mining; Information retrieval; Machine learning; Pattern recognition (psychology); Text mining; Image (mathematics)","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":[],"consensus_categories":[],"category_scores_codex":[0.0005243075,0.0001784819,0.0001949573,0.00005457078,0.0001311491,0.00006756237,0.00009577096,0.0003274266,0.0000505602],"category_scores_gemma":[0.0002658005,0.0001457315,0.00003439172,0.0001292885,0.00009558153,0.00003412755,0.0001559649,0.0002445006,0.000002266841],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003118194,"about_ca_system_score_gemma":0.00006437704,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008577334,"about_ca_topic_score_gemma":0.000001579477,"domain_scores_codex":[0.9983382,0.000138842,0.0004996231,0.0002526613,0.000576811,0.0001939009],"domain_scores_gemma":[0.9992543,0.00004323015,0.0001966763,0.0001515905,0.00005135281,0.0003028823],"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":[0.0002811528,0.0006477502,0.02346947,0.000557536,0.0007090922,0.00001796913,0.01938481,0.000006802188,0.007120795,0.00005397841,0.00263681,0.9451138],"study_design_scores_gemma":[0.03998706,0.02329222,0.4028825,0.0006345161,0.001533841,0.0009409758,0.04972009,0.3764274,0.0271198,0.001291007,0.07289694,0.003273678],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7902299,0.00008220013,0.2082474,0.0006740167,0.0001203374,0.0005227774,0.00002116752,0.00006401546,0.00003821049],"genre_scores_gemma":[0.883943,0.0001273184,0.1147727,0.0004100508,0.0001995786,0.00004317703,0.0004515534,0.0000123969,0.00004031841],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9418402,"threshold_uncertainty_score":0.5942758,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02101473833240067,"score_gpt":0.3257688454509995,"score_spread":0.3047541071185988,"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."}}