{"id":"W4243041937","doi":"10.1017/cbo9780511801686","title":"Analyzing Linguistic Data","year":2008,"lang":"en","type":"book","venue":"Cambridge University Press eBooks","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3194,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Variety (cybernetics); Computer science; Construct (python library); Variation (astronomy); Visualization; Natural language processing; Range (aeronautics); Artificial intelligence; Measure (data warehouse); Statistical model; Corpus linguistics; Linguistics; Data science; Data mining; Engineering; Programming language","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":[0.005411941,0.000937312,0.0008152378,0.005406387,0.001433904,0.007209702,0.001476171,0.0009199222,0.05099769],"category_scores_gemma":[0.02692075,0.0004977309,0.0008235945,0.005818343,0.001672746,0.00523484,0.002571823,0.002213044,0.03767369],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001400145,"about_ca_system_score_gemma":0.002211506,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001365242,"about_ca_topic_score_gemma":0.00198636,"domain_scores_codex":[0.9949432,0.00173065,0.0004572741,0.0006536764,0.002130448,0.00008487212],"domain_scores_gemma":[0.9877977,0.007480753,0.0004641383,0.001837347,0.002269476,0.0001505522],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002696401,0.00004157253,0.001648363,0.001235269,0.00006349267,0.0002513525,0.004097162,0.001048478,0.002271621,0.1001514,0.2350129,0.6541514],"study_design_scores_gemma":[0.000007934844,0.00003234542,0.001965866,0.0008052386,0.00002612755,0.0004072351,0.002405562,0.002437725,0.001461121,0.08536851,0.9050379,0.00004441099],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01072259,0.007347953,0.6669684,0.01276175,0.002731959,0.001383744,0.03114793,0.01014175,0.256794],"genre_scores_gemma":[0.04971389,0.01052129,0.7470227,0.004213885,0.001356526,0.001858593,0.03546619,0.004877033,0.1449699],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.05099769,"threshold_uncertainty_score":0.1706043,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03352730678943484,"score_gpt":0.2424501016676691,"score_spread":0.2089227948782343,"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."}}