Similarity of Authors' Profiles and Its Usage for Reviewers' Recommendation

This article describes an algorithm to facilitate the proper assignment of reviewers by finding an author's profile. It uses an original approach to analyzing publications published in digital libraries to get additional keywords based on NLP (natural language processing) techniques. Comparing profiles and finding similarities between them are performed afterwards in the vector space model in generally known ways. The result of our work is an algorithm to help conference organizers assign reviewers to registered publications. In many cases, the organizers have to know about the research area of all program committee members and have to select the appropriate opponent manually. The results of the algorithm functionality are verified by real data from two conferences, a local and global one.