{"id":"W2103291633","doi":"10.1109/dcc.1991.213370","title":"Models for compression in full-text retrieval systems","year":2002,"lang":"en","type":"article","venue":"","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Calgary","funders":"","keywords":"Lexicon; Computer science; Word (group theory); Natural language processing; Coding (social sciences); Artificial intelligence; Compression (physics); Information retrieval; Data compression; Mathematics; Statistics","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":[0.001887413,0.001009435,0.001251352,0.001373583,0.0008140256,0.003176445,0.001950686,0.002400599,0.008009781],"category_scores_gemma":[0.01192872,0.0006719889,0.001035693,0.00139932,0.001682049,0.006070722,0.001367231,0.001719416,0.003017363],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0023778,"about_ca_system_score_gemma":0.001291953,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006621629,"about_ca_topic_score_gemma":0.00425345,"domain_scores_codex":[0.9983338,0.0006011759,0.000114731,0.000216686,0.0005640771,0.0001696088],"domain_scores_gemma":[0.9950413,0.003483774,0.0002376406,0.0004650271,0.0006854962,0.00008688145],"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":[0.0002038705,0.00007434496,0.0005083655,0.0002066091,0.00004280993,0.000199775,0.000338333,0.7113449,0.001686483,0.2406499,0.004317751,0.04042684],"study_design_scores_gemma":[0.00002600147,0.00002953225,0.00007387939,0.00001865252,0.00001207663,0.00006669904,0.00002878525,0.9179722,0.0006049309,0.07843195,0.002717934,0.00001730542],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02457454,0.001890937,0.9562918,0.0017922,0.0001813759,0.0002639303,0.0006009602,0.001446344,0.01295789],"genre_scores_gemma":[0.7330446,0.004146353,0.2203176,0.0007537895,0.0004086574,0.001379683,0.001654778,0.0006021954,0.03769242],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008009781,"threshold_uncertainty_score":0.02679545,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04978943089054107,"score_gpt":0.2527467752001725,"score_spread":0.2029573443096314,"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."}}