{"id":"W4402569628","doi":"10.1109/access.2024.3462635","title":"Knowledge Graph Generation and Application for Unstructured Data Using Data Processing Pipeline","year":2024,"lang":"en","type":"article","venue":"IEEE Access","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"Cistel Technology (Canada); Carleton University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Unstructured data; Pipeline (software); Graph; Data modeling; Graph database; Data mining; Big data; Database; Theoretical computer science; Programming language","routes":{"ca_aff":true,"ca_fund":true,"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":[0.002319048,0.001049028,0.00057174,0.00610941,0.0008884409,0.001821234,0.001538785,0.0008973693,0.003801198],"category_scores_gemma":[0.01316196,0.0005284774,0.001338594,0.004987593,0.0005922452,0.002947785,0.002541857,0.001537771,0.00154348],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001014971,"about_ca_system_score_gemma":0.001899762,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007469419,"about_ca_topic_score_gemma":0.009282236,"domain_scores_codex":[0.998544,0.0003381346,0.0001554683,0.000453086,0.0004379992,0.00007136339],"domain_scores_gemma":[0.9924741,0.004194427,0.0003642182,0.001453651,0.001335088,0.00017852],"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":[0.0005013212,0.0005469506,0.005291405,0.001222711,0.0001892915,0.001163337,0.002283902,0.05255251,0.02722309,0.01754728,0.03424292,0.8572352],"study_design_scores_gemma":[0.0001651274,0.0002832309,0.003560049,0.0001482446,0.0001364057,0.0005479787,0.00155475,0.8031346,0.061439,0.05941271,0.0694935,0.0001243541],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03833137,0.0002821075,0.9027525,0.0008410457,0.0001496615,0.001162605,0.009782233,0.04260692,0.004091606],"genre_scores_gemma":[0.09143727,0.0002151268,0.8887227,0.000137092,0.00002299092,0.0004987594,0.01682623,0.0006538668,0.001486027],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007469419,"threshold_uncertainty_score":0.01485187,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.208256562006371,"score_gpt":0.4276106568465334,"score_spread":0.2193540948401625,"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."}}