{"id":"W4307534208","doi":"10.1145/3550356.3561592","title":"Towards automatically extracting UML class diagrams from natural language specifications","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":35,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Class diagram; Computer science; Unified Modeling Language; UML tool; Applications of UML; Communication diagram; Programming language; Activity diagram; Object Constraint Language; Class (philosophy); Systems Modeling Language; Precision and recall; Software engineering; Natural language processing; Artificial intelligence; Software","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.000548312,0.0002995101,0.0003189852,0.0002275855,0.0001572923,0.000730412,0.003452785,0.0001752788,0.001500527],"category_scores_gemma":[0.001695593,0.0002929562,0.0001859175,0.0004481463,0.00004067233,0.0002273163,0.004936897,0.002174386,0.0002446059],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004074027,"about_ca_system_score_gemma":0.0003584737,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006411815,"about_ca_topic_score_gemma":0.00003148218,"domain_scores_codex":[0.9969459,0.0001397196,0.0004113833,0.0009383556,0.001074028,0.000490668],"domain_scores_gemma":[0.9957293,0.001766157,0.0001201074,0.002103904,0.0001024224,0.0001780629],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001482785,0.000792388,0.006010882,0.0004574306,0.000731279,0.001095397,0.01897628,0.01463354,0.004258428,0.07969271,0.01918793,0.8541489],"study_design_scores_gemma":[0.0002416786,0.00002989476,0.1586907,0.00008760292,0.00001941088,0.0000164446,0.000374265,0.8295919,0.0006533618,0.003460543,0.005873279,0.000960979],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06597953,0.001015785,0.9151457,0.002705822,0.003362783,0.0005384936,0.00005120436,0.002934314,0.008266356],"genre_scores_gemma":[0.7121096,0.00001680554,0.2850667,0.0000956862,0.0003129067,0.0001851732,0.000104247,0.00003407828,0.002074807],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8531879,"threshold_uncertainty_score":0.9999523,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04266183570983247,"score_gpt":0.3121412878482224,"score_spread":0.2694794521383899,"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."}}