Embedding Models for Episodic Knowledge Graphs

TL;DR

This paper extends static knowledge graph embedding models to temporal knowledge graphs, introducing a new tensor model called ConT, and demonstrates improved performance on real-world datasets, with implications for modeling episodic and semantic memory.

Contribution

The paper generalizes existing static knowledge graph models to temporal graphs and introduces ConT, a novel tensor model with superior generalization capabilities.

Findings

ConT outperforms existing models on GDELT and ICEWS datasets.

Temporal embeddings can model episodic memory effectively.

Episodic-to-semantic memory projection is validated on ICEWS.

Abstract

In recent years a number of large-scale triple-oriented knowledge graphs have been generated and various models have been proposed to perform learning in those graphs. Most knowledge graphs are static and reflect the world in its current state. In reality, of course, the state of the world is changing: a healthy person becomes diagnosed with a disease and a new president is inaugurated. In this paper, we extend models for static knowledge graphs to temporal knowledge graphs. This enables us to store episodic data and to generalize to new facts (inductive learning). We generalize leading learning models for static knowledge graphs (i.e., Tucker, RESCAL, HolE, ComplEx, DistMult) to temporal knowledge graphs. In particular, we introduce a new tensor model, ConT, with superior generalization performance. The performances of all proposed models are analyzed on two different datasets: the Global Database of Events, Language, and Tone (GDELT) and the database for Integrated Conflict Early Warning System (ICEWS). We argue that temporal knowledge graph embeddings might be models also for cognitive episodic memory (facts we remember and can recollect) and that a semantic memory (current facts we know) can be generated from episodic memory by a marginalization operation. We validate this episodic-to-semantic projection hypothesis with the ICEWS dataset.