{"id":"W2052842594","doi":"10.1145/2590988","title":"Modeling Term Associations for Probabilistic Information Retrieval","year":2014,"lang":"en","type":"article","venue":"ACM Transactions on Information Systems","topic":"Information Retrieval and Search Behavior","field":"Computer Science","cited_by":51,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"Natural Sciences and Engineering Research Council of Canada; International Business Machines Corporation","keywords":"Term Discrimination; Computer science; Term (time); Bigram; Divergence-from-randomness model; Probabilistic logic; Ranking (information retrieval); Query expansion; Information retrieval; Data mining; Artificial intelligence; Web search query; Search engine; Concept search; Trigram","routes":{"ca_aff":true,"ca_fund":true,"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":[0.004333611,0.001486326,0.00192702,0.003842405,0.0009274848,0.002524682,0.003274251,0.002701929,0.002743546],"category_scores_gemma":[0.01983828,0.0009725412,0.002314277,0.006262755,0.001449513,0.007917346,0.002072444,0.002196685,0.002253713],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002035314,"about_ca_system_score_gemma":0.001486177,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007621221,"about_ca_topic_score_gemma":0.007163209,"domain_scores_codex":[0.9954612,0.001514216,0.0003651852,0.0009759381,0.001329714,0.0003538473],"domain_scores_gemma":[0.9926813,0.004638723,0.0009394166,0.0008443351,0.0007646119,0.0001316315],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0002366491,0.0001323733,0.002965892,0.0003922682,0.0001976537,0.0003128555,0.0004096749,0.6860224,0.00564169,0.15638,0.004194378,0.1431141],"study_design_scores_gemma":[0.00001199592,0.00004251246,0.0002737198,0.00001112723,0.00003313255,0.0001126681,0.00001852565,0.9534232,0.0004575095,0.04411577,0.001472638,0.00002716037],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01150598,0.001416278,0.9841663,0.0003388581,0.00005689228,0.00009500788,0.0002578175,0.0006733853,0.001489493],"genre_scores_gemma":[0.6159866,0.003796335,0.3650089,0.0004725867,0.0005371277,0.0009243336,0.002049478,0.0003684755,0.01085615],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007621221,"threshold_uncertainty_score":0.02291864,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03054072446624775,"score_gpt":0.267857013344392,"score_spread":0.2373162888781443,"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."}}