Semantic similarity expresses the commonality among concepts; therefore, it may play a critical role in a wide range of computational applications. The authors propose a method (wpath) to measure “the semantic similarity between concepts in knowledge graphs (KGs).” It combines the two major approaches for semantic similarity measurement, namely path-based and information content (IC)-based solutions. Wpath adopts the former approach to evaluate the difference between two given concepts, while the latter provides an estimation of the commonality between them.

The experiments, conducted on well-known datasets, show high performance for wpath in statistical terms (“significant improvement of correlation between computed similarity scores and human judgments”) as well as in terms of accuracy.

The paper is well written and properly structured. I found it interesting and informative. It is definitely a solid contribution, well supported by experiments.