Abstract
In the area of online news articles ltering, two types of methods have been used. Either the collaborative ltering, or the content based ltering supported by infor- mation retrieval techniques.
In this paper we introduce a novel approach to information ltering exploiting semantic information contained in each news article. The seman- tic information is in our case represented by named enti- ties extracted from the article.
We propose a system using named entities recognition to build a user pro le for infor- mation ltering. We developed the model of the user pro le that is being built automatically based on a user feedback.