{"id":"W6958173769","doi":"10.60692/j695m-0rz06","title":"Phrase-aware Unsupervised Constituency Parsing","year":2022,"lang":"en","type":"article","venue":"Greater South Information System","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Parsing; Phrase; Dependency grammar; Generality; Grammar; Task (project management); Regularization (linguistics); Natural language; Benchmark (surveying)","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":[0.001237164,0.001185057,0.0009152165,0.001276329,0.0007103423,0.0008999094,0.001841291,0.001011339,0.004819952],"category_scores_gemma":[0.003252279,0.0005760693,0.001229314,0.001350295,0.0007447986,0.002413725,0.00141535,0.002135531,0.003745466],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000711725,"about_ca_system_score_gemma":0.001693801,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002091013,"about_ca_topic_score_gemma":0.005285514,"domain_scores_codex":[0.9989839,0.0003692384,0.00005872532,0.0003611636,0.0001603362,0.00006668134],"domain_scores_gemma":[0.9976979,0.001178862,0.0001395371,0.0006246098,0.0003019056,0.00005728586],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002447637,0.000234705,0.004087033,0.0005555312,0.0001762059,0.0003885955,0.0006202944,0.07812057,0.09510709,0.03801935,0.02559749,0.7568483],"study_design_scores_gemma":[0.00002753063,0.00008877526,0.001555527,0.00002442915,0.00005023169,0.0002052603,0.00009471153,0.9151228,0.03335157,0.03925185,0.01019218,0.0000351356],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01843596,0.0002276409,0.9665824,0.0001868645,0.00004428417,0.00008753303,0.0007118089,0.0113459,0.002377554],"genre_scores_gemma":[0.2566359,0.0002429397,0.7264355,0.000342922,0.00009760789,0.0002563545,0.007988679,0.002796065,0.005204027],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004819952,"threshold_uncertainty_score":0.01612437,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02521848210352971,"score_gpt":0.2179249548280836,"score_spread":0.1927064727245539,"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."}}