{"id":"W2799804203","doi":"10.5539/ijel.v8n4p282","title":"Academic Vocabulary Use in Doctoral Theses: A Corpus-Based Lexical Analysis of Academic Word List (AWL) in Major Scientific Disciplinary Groups","year":2018,"lang":"en","type":"article","venue":"International Journal of English Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Vocabulary; Word (group theory); Discipline; Word list; Lexical density; Concordance; Computer science; Corpus linguistics; Linguistics; Natural language processing; Artificial intelligence; Lexical item; Sociology; Social science; Medicine","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.00222434,0.0002846699,0.0004701497,0.01051904,0.001277317,0.001999162,0.0003899972,0.000468873,0.002442347],"category_scores_gemma":[0.01413412,0.0001941854,0.000309361,0.010783,0.001225625,0.001999408,0.002674954,0.0006759788,0.0005920612],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009400402,"about_ca_system_score_gemma":0.001585378,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006048999,"about_ca_topic_score_gemma":0.01108677,"domain_scores_codex":[0.9981238,0.0005059395,0.0003847509,0.0004127952,0.0004852035,0.00008735438],"domain_scores_gemma":[0.9897781,0.007060489,0.001288662,0.0003549212,0.001312526,0.0002052771],"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":[0.0007757944,0.0004776688,0.2797491,0.01268364,0.0001961172,0.003255981,0.2944582,0.0008621509,0.0345971,0.008001297,0.02442337,0.3405195],"study_design_scores_gemma":[0.00007911289,0.0002410842,0.6767305,0.002223106,0.0002619603,0.002386349,0.1832893,0.004086225,0.007236813,0.002810692,0.1205028,0.0001520542],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9780461,0.001851327,0.003418135,0.0003177572,0.00006792736,0.0003968936,0.01038827,0.00007677809,0.005436763],"genre_scores_gemma":[0.9541947,0.001659841,0.01619162,0.0001404252,0.00005918674,0.001474475,0.02388957,0.0001049586,0.002285269],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.989481,"threshold_uncertainty_score":0.01202762,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03791565145728037,"score_gpt":0.3492451197885814,"score_spread":0.311329468331301,"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."}}