Knowledge Graph Embeddings for Causal Relation Prediction

Abstract

Recently, there has been an increasing interest in knowledge graphs (KGs) of causal relations between events. Such KGs can be used for event analysis and forecasting in a variety of applications. In this paper, we study the problem of enriching an existing causal KG of news events using KG embeddings-based link prediction techniques. We perform a thorough evaluation of the performance of five different methods using classic accuracy measures as well as a novel scheme for manual evaluation. Our study provides insights on the strengths and weaknesses of different link prediction methods.