{"id":"W6950058172","doi":"10.5281/zenodo.3465855","title":"Large-scale Topic-based Document Retrieval Experiments","year":2019,"lang":"en","type":"other","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"","field":"","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Virtual Materials Group (Canada)","funders":"","keywords":"Document retrieval; Vector space model; Document clustering; Set (abstract data type); Field (mathematics)","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":[0.008035642,0.003273701,0.002877891,0.003599507,0.002127081,0.001747919,0.003327145,0.003125769,0.01155669],"category_scores_gemma":[0.01917921,0.000729104,0.00189532,0.00517858,0.001378025,0.002980384,0.00261033,0.001988811,0.007547092],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001688407,"about_ca_system_score_gemma":0.002479116,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02038042,"about_ca_topic_score_gemma":0.01616203,"domain_scores_codex":[0.992165,0.003287018,0.001144096,0.001330593,0.001366643,0.0007067016],"domain_scores_gemma":[0.9837549,0.009106165,0.0004913281,0.00299929,0.002893288,0.0007549744],"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":[0.02669788,0.02308917,0.0101108,0.01518941,0.002539486,0.002128848,0.002363623,0.1167958,0.07758152,0.004477806,0.3481949,0.3708307],"study_design_scores_gemma":[0.02214121,0.02228865,0.06627134,0.0008444494,0.003442202,0.004243342,0.004402243,0.5160469,0.1867106,0.01095903,0.1613281,0.001322096],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7883778,0.008254858,0.03806955,0.001451448,0.001684453,0.00640378,0.1080437,0.02376278,0.02395161],"genre_scores_gemma":[0.5490371,0.001990515,0.09765113,0.0008237027,0.0005903176,0.005038973,0.3247491,0.002214024,0.01790505],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02038042,"threshold_uncertainty_score":0.04249704,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02708428529971325,"score_gpt":0.2715849584263811,"score_spread":0.2445006731266679,"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."}}