Tensor Decompositions for temporal knowledge base completion

TL;DR

This paper introduces a tensor decomposition approach for temporal knowledge base completion, extending ComplEx with new regularization, and provides a large new dataset for evaluating temporal link prediction methods.

Contribution

It presents a novel tensor-based method for temporal link prediction and introduces a large, new Wikidata-derived dataset for benchmarking.

Findings

Achieves state-of-the-art performance on temporal link prediction tasks

Introduces a new large-scale dataset for knowledge base completion

Extends ComplEx with effective regularization schemes

Abstract

Most algorithms for representation learning and link prediction in relational data have been designed for static data. However, the data they are applied to usually evolves with time, such as friend graphs in social networks or user interactions with items in recommender systems. This is also the case for knowledge bases, which contain facts such as (US, has president, B. Obama, [2009-2017]) that are valid only at certain points in time. For the problem of link prediction under temporal constraints, i.e., answering queries such as (US, has president, ?, 2012), we propose a solution inspired by the canonical decomposition of tensors of order 4. We introduce new regularization schemes and present an extension of ComplEx (Trouillon et al., 2016) that achieves state-of-the-art performance. Additionally, we propose a new dataset for knowledge base completion constructed from Wikidata, larger than previous benchmarks by an order of magnitude, as a new reference for evaluating temporal and non-temporal link prediction methods.