{"id":"W4381161845","doi":"10.20944/preprints202306.1378.v1","title":"Algorithm-based Data Generation (ADG) Engine for Data Analytics","year":2023,"lang":"en","type":"preprint","venue":"Preprints.org","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Western University","funders":"","keywords":"Big data; Computer science; Analytics; Data analysis; Domain (mathematical analysis); Data science; Data mining","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":["metaepi_narrow","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.002298285,0.0003971013,0.000434046,0.0002328454,0.0002384456,0.0003099058,0.01360879,0.000274943,0.0000315629],"category_scores_gemma":[0.0006275373,0.0004344847,0.00009335025,0.0004690746,0.00006348074,0.0006538201,0.02393473,0.0005582595,0.0006623094],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008756751,"about_ca_system_score_gemma":0.000587988,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003028059,"about_ca_topic_score_gemma":0.00005135646,"domain_scores_codex":[0.9950871,0.00007475963,0.0006912972,0.003188877,0.0004984289,0.0004594959],"domain_scores_gemma":[0.9810461,0.0002480266,0.0003897867,0.01787958,0.0002558365,0.000180683],"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":[0.00001365519,0.001112595,0.007524566,0.0007104905,0.0009651326,0.00004080334,0.0004350601,0.04980901,0.001797412,0.00569634,0.08264455,0.8492504],"study_design_scores_gemma":[0.0002569308,0.000009902973,0.001966561,0.0000580815,0.00008546562,0.000001828198,0.000005495969,0.9473891,0.001272329,0.001737486,0.04678558,0.0004311899],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0006016858,0.00006077046,0.9842067,0.004481083,0.001274327,0.0009646725,0.007635556,0.0007127576,0.00006243048],"genre_scores_gemma":[0.006284161,0.0001285311,0.9279581,0.0002810626,0.001173794,0.0004467769,0.06264441,0.00008087737,0.001002249],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8975801,"threshold_uncertainty_score":0.9998107,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5410040128402656,"score_gpt":0.4267004535432325,"score_spread":0.1143035592970331,"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."}}