{"id":"W2102838806","doi":"10.1016/j.ipm.2008.01.003","title":"Lexical cohesion and term proximity in document ranking","year":2008,"lang":"en","type":"article","venue":"Information Processing & Management","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":32,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Query expansion; Ranking (information retrieval); Computer science; Cohesion (chemistry); Information retrieval; Collocation (remote sensing); Relevance (law); Term (time); Query optimization; Natural language processing; Artificial intelligence; Machine learning","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":[0.005765066,0.0004686474,0.001888398,0.01458578,0.002366853,0.004167299,0.0009420876,0.001567272,0.002916908],"category_scores_gemma":[0.03846732,0.0006150901,0.0009216817,0.01263542,0.001815726,0.007520295,0.00238887,0.001417875,0.0008045201],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00106415,"about_ca_system_score_gemma":0.001241538,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004461339,"about_ca_topic_score_gemma":0.006525444,"domain_scores_codex":[0.9945151,0.002403614,0.0006844114,0.00060776,0.001447597,0.0003415981],"domain_scores_gemma":[0.9592744,0.0347773,0.001571361,0.001239882,0.002411858,0.0007252264],"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":[0.004106771,0.0009498762,0.09645715,0.001660081,0.0008439483,0.001052313,0.003140299,0.03893217,0.01787482,0.09893999,0.0121563,0.7238863],"study_design_scores_gemma":[0.0006093816,0.001339855,0.1201491,0.0003669608,0.001423687,0.002181933,0.003470624,0.5549489,0.01403083,0.2856083,0.01545458,0.0004157588],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6481903,0.01108225,0.3278551,0.000916437,0.000334324,0.000248121,0.001006674,0.00094801,0.009418789],"genre_scores_gemma":[0.9302239,0.00107055,0.06453741,0.00007311178,0.0004562446,0.0001395467,0.001185162,0.0001283704,0.002185801],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01458578,"threshold_uncertainty_score":0.03048897,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0155754297426092,"score_gpt":0.2417716080082823,"score_spread":0.2261961782656731,"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."}}