{"id":"W2585000520","doi":"10.1109/bigdata.2016.7840938","title":"Software engineering for big data projects: Domains, methodologies and gaps","year":2016,"lang":"en","type":"article","venue":"","topic":"Big Data and Business Intelligence","field":"Business, Management and Accounting","cited_by":30,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Big data; Computer science; Data science; Context (archaeology); Milestone; Scalability; Software; Software engineering; Database; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04189386,0.0007364408,0.001050758,0.01436228,0.001603459,0.008711026,0.002753376,0.002539162,0.002031231],"category_scores_gemma":[0.06847782,0.0007022734,0.00101527,0.02200729,0.003766297,0.01230046,0.00349795,0.00265923,0.0009218674],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.005470644,"about_ca_system_score_gemma":0.01666553,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002607867,"about_ca_topic_score_gemma":0.002607824,"domain_scores_codex":[0.9706891,0.01685099,0.003756019,0.001354178,0.006702505,0.0006471997],"domain_scores_gemma":[0.8167537,0.1444726,0.009603375,0.004759913,0.02156934,0.002840951],"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":[0.0000681696,0.0003100677,0.01198474,0.02418707,0.0001422752,0.000398616,0.008417,0.004792232,0.0007853546,0.1737344,0.01533686,0.7598432],"study_design_scores_gemma":[0.00007894123,0.0004033487,0.01677717,0.06981435,0.0001734249,0.001976267,0.04470944,0.01791206,0.002961913,0.3558589,0.4891543,0.0001799504],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.05430694,0.6136855,0.2035458,0.07898195,0.001348907,0.0009716815,0.0005622283,0.000488718,0.04610828],"genre_scores_gemma":[0.3079551,0.4746631,0.2038982,0.005844934,0.0009095878,0.002364988,0.0009026005,0.000314397,0.003147151],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.04189386,"threshold_uncertainty_score":0.2215585,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3419614475189196,"score_gpt":0.3394599206551451,"score_spread":0.002501526863774528,"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."}}