{"id":"W4362575620","doi":"10.22215/etd/2023-15426","title":"Knowledge Graph Generation for Unstructured Data Using Data Processing Pipeline","year":2023,"lang":"en","type":"dissertation","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"Unstructured data; Computer science; Pipeline (software); Information extraction; Knowledge graph; Coreference; Graph; Information retrieval; The Internet; Knowledge extraction; Data mining; Natural language processing; Artificial intelligence; Data science; Resolution (logic); Big data; World Wide Web; Theoretical computer science; Programming language","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":[0.00148212,0.0009816567,0.0006273256,0.003745075,0.0007553653,0.001484178,0.001582416,0.0008912117,0.004358451],"category_scores_gemma":[0.006270119,0.000541401,0.001695915,0.003218238,0.0005112289,0.003081447,0.00213409,0.001523031,0.002044664],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001111787,"about_ca_system_score_gemma":0.001959972,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01067949,"about_ca_topic_score_gemma":0.01740525,"domain_scores_codex":[0.9991537,0.0001416232,0.00007105237,0.0003161983,0.0002591338,0.00005816461],"domain_scores_gemma":[0.9970288,0.001493659,0.000116929,0.000722243,0.0005585476,0.0000799526],"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":[0.0003283696,0.0004447259,0.004764374,0.0006912994,0.0001819827,0.0005591401,0.0009318975,0.06968354,0.02170684,0.02739981,0.0453455,0.8279625],"study_design_scores_gemma":[0.00005926092,0.00008475711,0.001302905,0.00004458062,0.00006895155,0.0001894383,0.000296836,0.8902264,0.02779612,0.04990652,0.02998198,0.00004221356],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01733471,0.0001381292,0.945179,0.0004571602,0.00005376096,0.0004312424,0.004245085,0.02997951,0.002181398],"genre_scores_gemma":[0.1043612,0.0001531667,0.8743642,0.0001425309,0.00002253305,0.0003851627,0.01778226,0.000856329,0.001932637],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01067949,"threshold_uncertainty_score":0.02123463,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.269369695870474,"score_gpt":0.4050623587275948,"score_spread":0.1356926628571208,"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."}}