{"id":"W3198122215","doi":"10.1145/3472163.3472171","title":"Rigorous Measurement Model for Validity of Big Data: MEGA Approach","year":2021,"lang":"en","type":"article","venue":"","topic":"Big Data and Business Intelligence","field":"Business, Management and Accounting","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Big data; Data science; Computer science; Data quality; Quality (philosophy); Context (archaeology); Credibility; Data mining; Engineering","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":[0.04790025,0.001650921,0.002228272,0.01049729,0.002004103,0.008215865,0.004506553,0.00372364,0.003107157],"category_scores_gemma":[0.1629341,0.001118431,0.003180874,0.007350494,0.01334197,0.01631762,0.006692208,0.006625127,0.0007286661],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.008370513,"about_ca_system_score_gemma":0.007120849,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004641159,"about_ca_topic_score_gemma":0.002300943,"domain_scores_codex":[0.9196643,0.05171015,0.004141746,0.006006118,0.0166801,0.001797677],"domain_scores_gemma":[0.803547,0.1522072,0.01376067,0.01265676,0.01621233,0.001616102],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002393585,0.00005053369,0.003978475,0.0001845907,0.0001027337,0.00008824788,0.0006489883,0.01455711,0.0001519391,0.966499,0.0007943696,0.01292008],"study_design_scores_gemma":[0.00002388232,0.00007924816,0.00178462,0.0001928955,0.00005593241,0.0001158974,0.0003264902,0.1201477,0.0002568997,0.8738992,0.003072052,0.00004502914],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.009046994,0.0005106512,0.9764091,0.003803236,0.00007250701,0.0003015521,0.000254349,0.0001486135,0.009453018],"genre_scores_gemma":[0.6286763,0.0009449651,0.3647528,0.0009878521,0.0003407666,0.002131299,0.0005403403,0.00009545234,0.001530245],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04790025,"threshold_uncertainty_score":0.2533237,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5630573622897133,"score_gpt":0.3311868205412617,"score_spread":0.2318705417484516,"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."}}