{"id":"W1996527383","doi":"10.7202/038312ar","title":"Is Bigger Better? Corpus and Dictionary Use in the Search for Compounds, Collocations, Derived Forms and Fixed Expressions","year":2009,"lang":"en","type":"article","venue":"Meta Journal des traducteurs","topic":"Lexicography and Language Studies","field":"Arts and Humanities","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Computer science; Natural language processing; Artificial intelligence; Bilingual dictionary; Corpus linguistics; Machine-readable dictionary; Speech recognition","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":true,"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":[0.01060951,0.0004161256,0.001058978,0.01424175,0.003026592,0.006987889,0.001057722,0.0009561313,0.008587775],"category_scores_gemma":[0.04046158,0.0006358639,0.0004303337,0.02103346,0.002434854,0.0116697,0.002750972,0.001148274,0.001812077],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001419292,"about_ca_system_score_gemma":0.003088056,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005483777,"about_ca_topic_score_gemma":0.0123129,"domain_scores_codex":[0.9900068,0.005853313,0.00101561,0.001079363,0.001857217,0.0001875682],"domain_scores_gemma":[0.9513904,0.03383442,0.002822393,0.005538606,0.00571269,0.000701464],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0005538938,0.0001199166,0.02650357,0.002288353,0.000162875,0.000787704,0.02036778,0.0008028495,0.007466966,0.08880944,0.01576351,0.8363732],"study_design_scores_gemma":[0.0002148795,0.0004226512,0.04359527,0.005227451,0.0004707978,0.007152858,0.04631793,0.0111548,0.01709617,0.0845412,0.7833913,0.0004146732],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.4382112,0.05351213,0.310161,0.02127312,0.00151818,0.0009821063,0.008371344,0.002262475,0.1637085],"genre_scores_gemma":[0.4154834,0.02022906,0.5455743,0.001263477,0.0004066714,0.0007827722,0.004379579,0.001194612,0.01068618],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01424175,"threshold_uncertainty_score":0.05610913,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08673062436678199,"score_gpt":0.2818169938120741,"score_spread":0.1950863694452922,"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."}}