{"id":"W4399174990","doi":"10.1145/3654984","title":"Unstructured Data Fusion for Schema and Data Extraction","year":2024,"lang":"en","type":"article","venue":"Proceedings of the ACM on Management of Data","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Information retrieval; Schema (genetic algorithms); Component (thermodynamics); Data mining; Table (database)","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":["open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.0009438187,0.0001011667,0.0001171113,0.00007548944,0.00006279305,0.0001794931,0.0171442,0.00003104007,0.000002791011],"category_scores_gemma":[0.0002407534,0.00007355378,0.00001343678,0.0002220584,0.00003341998,0.002482767,0.03822575,0.00008721198,9.929338e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009587436,"about_ca_system_score_gemma":0.00001154154,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006028885,"about_ca_topic_score_gemma":8.390851e-7,"domain_scores_codex":[0.9984418,0.000003119946,0.0002345824,0.0008566557,0.0003379596,0.0001258831],"domain_scores_gemma":[0.9941567,0.00005810408,0.0001369203,0.005583527,0.00004182901,0.00002291744],"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":[0.00003592201,0.00005884479,0.0001102138,0.003100174,0.0001928661,0.000001219808,0.0001085568,0.00001075283,0.00563001,0.1537425,0.08083028,0.7561786],"study_design_scores_gemma":[0.0003451138,0.0000446164,0.0004438191,0.0007839248,0.0001514738,0.000007694895,0.0001515442,0.8964323,0.002833656,0.03247177,0.06617129,0.0001627998],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08366098,0.003380919,0.8359582,0.05840151,0.003742253,0.004393767,0.004914106,0.0006420108,0.004906295],"genre_scores_gemma":[0.2687187,0.0003669214,0.7301885,0.0000838699,0.0001161075,0.000005722396,0.0002706895,0.00001524087,0.0002342473],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8964216,"threshold_uncertainty_score":0.9881735,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1355387895140462,"score_gpt":0.3611646660040275,"score_spread":0.2256258764899813,"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."}}