Impact of Injecting Ground Truth Explanations on Relational Graph Convolutional Networks and their Explanation Methods for Link Prediction on Knowledge Graphs - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Impact of Injecting Ground Truth Explanations on Relational Graph Convolutional Networks and their Explanation Methods for Link Prediction on Knowledge Graphs

Résumé

Relational Graph Convolutional Networks (RGCNs) are commonly applied to Knowledge Graphs (KGs) for black box link prediction. Several algorithms, or explanations methods, have been proposed to explain the predictions of this model. Recently, researchers have constructed datasets with ground truth explanations for quantitative and qualitative evaluation of predicted explanations. Benchmark results showed state-of-theart explanation methods had difficulties predicting explanations. In this work, we leverage prior knowledge to further constrain the loss function of RGCNs, by penalizing node embeddings far away from the node embeddings in their associated ground truth explanation. Empirical results show improved explanation prediction performance of state-of-the-art post hoc explanations methods for RGCNs, at the cost of predictive performance. Additionally, we quantify the different types of errors made both in terms of data and semantics.
Fichier principal
Vignette du fichier
WI_2022.pdf (249.51 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03771424 , version 1 (07-09-2022)

Identifiants

  • HAL Id : hal-03771424 , version 1

Citer

Nicholas Halliwell, Fabien Gandon, Freddy Lecue. Impact of Injecting Ground Truth Explanations on Relational Graph Convolutional Networks and their Explanation Methods for Link Prediction on Knowledge Graphs. WI-IAT 2022 - The 21st IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology, Nov 2022, Niagara Falls / Hybrid, Canada. ⟨hal-03771424⟩
110 Consultations
112 Téléchargements

Partager

Gmail Facebook X LinkedIn More