{"id":"W3081885737","doi":"10.5539/mas.v14n9p12","title":"Big Data Ingestion and Preparation Tools","year":2020,"lang":"en","type":"article","venue":"Modern Applied Science","topic":"Big Data and Business Intelligence","field":"Business, Management and Accounting","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Big data; Ingestion; Data science; Point (geometry); Data mining; Medicine","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01121354,0.002064036,0.001754099,0.009116433,0.001799826,0.008605929,0.003347419,0.001556638,0.02668087],"category_scores_gemma":[0.03697579,0.001784831,0.00261761,0.008568103,0.001272314,0.009965752,0.006816278,0.003222492,0.01883445],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008296383,"about_ca_system_score_gemma":0.003325314,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001791778,"about_ca_topic_score_gemma":0.001200792,"domain_scores_codex":[0.994024,0.001243345,0.0008682089,0.0007815278,0.002734859,0.0003480577],"domain_scores_gemma":[0.9702065,0.01518567,0.001718912,0.007040344,0.005113121,0.0007355292],"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":[0.0006012045,0.0002848874,0.007082448,0.0049985,0.0002980994,0.001138619,0.002635506,0.003414243,0.006606461,0.03987443,0.238683,0.6943826],"study_design_scores_gemma":[0.0001449929,0.0001442755,0.007251881,0.002648694,0.0002058724,0.001768229,0.001635771,0.01676105,0.02441567,0.06216625,0.882543,0.0003143976],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.005898859,0.00496992,0.7834756,0.003109326,0.0009078733,0.00213762,0.0158538,0.1435027,0.04014425],"genre_scores_gemma":[0.04355945,0.006285541,0.8754343,0.002374061,0.0007611535,0.00213796,0.03315223,0.01890317,0.01739211],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02668087,"threshold_uncertainty_score":0.08925635,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2707234784005533,"score_gpt":0.316525427836894,"score_spread":0.04580194943634069,"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."}}