Application of Concepts of Neighbours to Knowledge Graph Completion

Tracking #: 622-1602

Authors:

NameORCID
Sébastien FerréORCID logo https://orcid.org/0000-0002-6302-2333


Responsible editor: 

Tobias Kuhn

Submission Type: 

Research Paper

Abstract: 

The open nature of Knowledge Graphs (KG) often implies that they are incomplete. Knowlege graph completion (aka. link prediction) consists in infering new relationships between the entities of a KG based on existing relationships. Most existing approaches rely on the learning of latent feature vectors for the encoding of entities and relations. In general however, latent features cannot be easily interpreted. Rule-based approaches offer interpretability but a distinct ruleset must be learned for each relation. In both latent- and rule-based approaches, the training phase has to be run again when the KG is updated. We propose a new approach that does not need a training phase, and that can provide interpretable explanations for each inference. It relies on the computation of Concepts of Nearest Neighbours (CNN) to identify clusters of similar entities based on common graph patterns. Different rules are then derived from those graph patterns, and combined to predict new relationships. We evaluate our approach on standard benchmarks for link prediction, where it gets competitive performance compared to existing approaches.

Manuscript: 

Tags: 

  • Reviewed

Data repository URLs: 

Date of Submission: 

Saturday, February 29, 2020

Date of Decision: 

Monday, April 20, 2020


Nanopublication URLs:

Decision: 

Undecided

Solicited Reviews:


1 Comment

Meta-Review by Editor

The reviewers agree that the paper has its merits, but also point to a number of shortcomings. In particular, scalability should be better addressed, the formal terminology and use of symbols should be simplified, the bigger picture should be made clearer throughout, and the explainability property should be better justified. These points need to be fixed before the paper can be accepted.

Tobias Kuhn (http://orcid.org/0000-0002-1267-0234)