1cecc7a77928ca8133fa24680a88d2f9-Reviews.html

First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. The authors propose a simple and scalable approach to modeling multi-relational data using low-dimensional vector embeddings of entities, with the relationships between embeddings captured using offset vectors. The embeddings are learned by training a margin-based ranking model to score the observed entity1,relationship,entity2 triples higher than the unobserved ones. Though the proposed model can be seen as a special case of several existing models (e.g. The approach is well motivated and clearly described. The empirical evaluation is reasonably well done, but the write up could be better.