{"id":"W4402684015","doi":"10.18653/v1/2024.acl-long.188","title":"Collaboration or Corporate Capture? Quantifying NLP’s Reliance on Industry Artifacts and Contributions","year":2024,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Natural language processing; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.07418799,0.0006483766,0.0008571914,0.01835699,0.002392566,0.01867045,0.002112574,0.002680255,0.005257343],"category_scores_gemma":[0.3187877,0.0007538197,0.000880947,0.03432789,0.003777667,0.02777255,0.01055148,0.002256023,0.002656962],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002286222,"about_ca_system_score_gemma":0.00375317,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002208821,"about_ca_topic_score_gemma":0.003294314,"domain_scores_codex":[0.9252914,0.03708281,0.007774902,0.006991289,0.02108421,0.001775347],"domain_scores_gemma":[0.4565975,0.4127725,0.04114981,0.05647649,0.02911032,0.003893299],"domain_codex":null,"domain_gemma":"incentives","domain_candidate":"incentives","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0003071032,0.0002141079,0.3322673,0.004924719,0.0007411717,0.001793475,0.04209641,0.005353348,0.006345226,0.09681822,0.02852904,0.4806099],"study_design_scores_gemma":[0.00009184668,0.0004307653,0.2435131,0.005437903,0.0009571946,0.00489042,0.05938551,0.02635598,0.01740691,0.114486,0.526717,0.0003273663],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.566673,0.0309325,0.1206441,0.03705717,0.001182842,0.0004857496,0.007365937,0.001549809,0.2341089],"genre_scores_gemma":[0.9503938,0.008529264,0.02818437,0.00150327,0.0008667775,0.0003242412,0.004088179,0.0006651926,0.005444873],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.981643,"threshold_uncertainty_score":0.3923483,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05251448698163644,"score_gpt":0.3398627932147888,"score_spread":0.2873483062331523,"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."}}