GausSetExpander: A Simple Approach for Entity Set Expansion

Entity Set Expansion (ESE) is an important task in Natural Language Processing that aims at expanding a small set of entities into a larger one with items from a large pool of candidates. The problem implicitly requires the definition of the notion of similarity between the given entities and the candidates. In this paper, we propose GausSetExpander, an unsupervised approach for the task of ESE based on optimal transport techniques. We propose to re-frame the problem as choosing the entity that best completes the input set. For this, we interpret a set as an elliptical distribution with a centroid which represents the mean and a dispersion that serves as the spread of the variance. The best candidate entity is the one that increases the spread of the set the least. We analyze the strength and the weaknesses of the proposed solution in order to assess the validity of our proposed approach.