The latest generations of Link Prediction (LP) models rely on embeddings to tackle incompleteness in Knowledge Graphs, achieving great performance at the cost of interpretability. Their opaqueness limits the trust that users can place in them, hindering their adoption in real-world applications. We have recently introduced Kelpie, an explainability framework tailored specifically for embedding-based LP models. Kelpie can be applied to any embedding-based LP model, and supports two explanation scenarios that we have called necessary and sufficient. In this demonstration we showcase Kelpie’s capability to explain the predictions of models based on vastly different architectures on the 5 major datasets in literature.

Rossi, A., Merialdo, P., Firmani, D., Teofili, T. (2022). Kelpie: an Explainability Framework for Embedding-based Link Prediction Models. PROCEEDINGS OF THE VLDB ENDOWMENT, 15(12), 3566-3569 [10.14778/3554821.3554845].

Kelpie: an Explainability Framework for Embedding-based Link Prediction Models

Rossi A.;Merialdo P.;Teofili T.
2022

Abstract

The latest generations of Link Prediction (LP) models rely on embeddings to tackle incompleteness in Knowledge Graphs, achieving great performance at the cost of interpretability. Their opaqueness limits the trust that users can place in them, hindering their adoption in real-world applications. We have recently introduced Kelpie, an explainability framework tailored specifically for embedding-based LP models. Kelpie can be applied to any embedding-based LP model, and supports two explanation scenarios that we have called necessary and sufficient. In this demonstration we showcase Kelpie’s capability to explain the predictions of models based on vastly different architectures on the 5 major datasets in literature.
Rossi, A., Merialdo, P., Firmani, D., Teofili, T. (2022). Kelpie: an Explainability Framework for Embedding-based Link Prediction Models. PROCEEDINGS OF THE VLDB ENDOWMENT, 15(12), 3566-3569 [10.14778/3554821.3554845].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11590/418220
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact