{"id":"W2611926229","doi":"","title":"Beyond the Dataset: Understanding Sociotechnical Aspects of the Knowledge Discovery Process Among Modern Data Professionals","year":2017,"lang":"en","type":"dissertation","venue":"UWSpace (University of Waterloo)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Waterloo; Simon Fraser University","keywords":"Sociotechnical system; Data science; Process (computing); Knowledge management; Computer science","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","sts"],"consensus_categories":[],"category_scores_codex":[0.05823795,0.0004771526,0.0006028374,0.006386802,0.008450444,0.01528962,0.002052025,0.002574714,0.001412828],"category_scores_gemma":[0.1061619,0.0006926595,0.0005184507,0.004802518,0.01460396,0.01981893,0.009281198,0.00287106,0.0003375899],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006689593,"about_ca_system_score_gemma":0.01240813,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008985182,"about_ca_topic_score_gemma":0.009753685,"domain_scores_codex":[0.9553595,0.03478755,0.001315298,0.002491781,0.004509055,0.001536955],"domain_scores_gemma":[0.7533233,0.2232188,0.008063155,0.00619371,0.006661057,0.00253991],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.00003537777,0.0000867592,0.01777373,0.0002221413,0.00001594826,0.0003246845,0.9230875,0.0002726486,0.0007571943,0.02577072,0.0007087826,0.03094456],"study_design_scores_gemma":[0.0000180584,0.00005242586,0.01112497,0.0004314836,0.00001927398,0.0002634392,0.8929711,0.005092531,0.0009240884,0.06468002,0.02438348,0.00003912584],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.846199,0.000935048,0.1111107,0.0184409,0.00007684116,0.0004960394,0.0001958215,0.00008683866,0.02245872],"genre_scores_gemma":[0.9599973,0.0005945194,0.03671697,0.0009029591,0.00003483557,0.0003324739,0.0001516699,0.00004974247,0.001219551],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9915496,"threshold_uncertainty_score":0.3079954,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1558935873536148,"score_gpt":0.3803340841839857,"score_spread":0.2244404968303708,"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."}}