{"id":"W2070507812","doi":"10.1145/1148170.1148285","title":"Term proximity scoring for ad-hoc retrieval on very large text collections","year":2006,"lang":"en","type":"article","venue":"","topic":"Information Retrieval and Search Behavior","field":"Computer Science","cited_by":162,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Information retrieval; Term (time); Term Discrimination; Artificial intelligence; Natural language processing; Search engine; Concept search; Web search query","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":[0.003938238,0.001669488,0.002402649,0.006628517,0.001847408,0.002268425,0.002526033,0.001462613,0.004204764],"category_scores_gemma":[0.01630929,0.0005090956,0.001019285,0.006845159,0.0007911081,0.003382941,0.001927836,0.001361929,0.005136847],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008916894,"about_ca_system_score_gemma":0.00188833,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004948084,"about_ca_topic_score_gemma":0.008248072,"domain_scores_codex":[0.9966312,0.001048723,0.0003062296,0.0005100026,0.001355898,0.0001478655],"domain_scores_gemma":[0.9943522,0.002508192,0.0004005027,0.001078378,0.001433149,0.0002275426],"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":[0.0003402031,0.000546916,0.002152586,0.0006662645,0.0002642459,0.0002063913,0.0002361366,0.02220717,0.03450051,0.003220544,0.01868981,0.9169692],"study_design_scores_gemma":[0.0002184449,0.001099594,0.008558518,0.0001143432,0.0003979893,0.00107384,0.0003721338,0.8942028,0.04293935,0.03187007,0.01893513,0.0002177284],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.03812081,0.002727755,0.9465234,0.0003299712,0.0002846963,0.0007223025,0.0005769706,0.007870689,0.002843411],"genre_scores_gemma":[0.2073316,0.001347596,0.7814657,0.000247898,0.0006686189,0.000872903,0.002601748,0.0004932196,0.004970776],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006628517,"threshold_uncertainty_score":0.02082765,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02238626999339941,"score_gpt":0.2708168976675244,"score_spread":0.248430627674125,"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."}}