Method of text similarity detection based on comparing their headers using assignment problem

Abstract

The main goal of the article is the development of the method for comparing news articles. It is suggested to compare news articles based on their titles. This problem was reduced to the problem of comparing of the short texts (namely, to the equivalency detection        problem). This problem was reduced to an assignment problem – classical computer science problem, that can be solved with Hungarian algorithm or using the algorithms, that find minimum cost maximum flow. The method can be used for news articles clasterization and for news aggregators.