{"id":"W4294959418","doi":"10.1109/wi-iat55865.2022.00049","title":"Impact of Injecting Ground Truth Explanations on Relational Graph Convolutional Networks and their Explanation Methods for Link Prediction on Knowledge Graphs","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Advanced Graph Neural Networks","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Thales (Canada)","funders":"","keywords":"Ground truth; Leverage (statistics); Computer science; Knowledge graph; Benchmark (surveying); Theoretical computer science; Graph; Node (physics); Black box; Artificial intelligence; Machine learning","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":[0.005113952,0.001596571,0.0007348498,0.001418821,0.0005533505,0.001591188,0.00188574,0.002458495,0.002754852],"category_scores_gemma":[0.04536204,0.0004766981,0.000870175,0.0009797567,0.001507252,0.005311671,0.002231026,0.002993496,0.0005285849],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002211941,"about_ca_system_score_gemma":0.001798278,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01144826,"about_ca_topic_score_gemma":0.01755201,"domain_scores_codex":[0.9972458,0.001190922,0.0001404465,0.0007724073,0.0004668875,0.0001837121],"domain_scores_gemma":[0.9689736,0.02470016,0.0014257,0.003039468,0.001388889,0.0004720781],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00109676,0.0003246697,0.0290236,0.0004794406,0.0003559281,0.0002510272,0.0002994809,0.6737905,0.002440355,0.01594027,0.008462109,0.2675359],"study_design_scores_gemma":[0.00002915342,0.00009514736,0.001286843,0.00007661778,0.00005574982,0.00005016639,0.0000607891,0.9819267,0.002336952,0.01331577,0.0007534694,0.00001257395],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.6087261,0.005791424,0.3645974,0.003937554,0.0004488244,0.0001985184,0.002367044,0.007256516,0.00667649],"genre_scores_gemma":[0.9364967,0.0005689688,0.0574633,0.0004333906,0.0000707849,0.00005355123,0.003200331,0.0002253521,0.001487545],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.01144826,"threshold_uncertainty_score":0.02704555,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05978387932113884,"score_gpt":0.3664902829082578,"score_spread":0.306706403587119,"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."}}