DC proposal: Model for news filtering with named entities

Abstract

In this paper we introduce the project of our PhD thesis. The subject is a model for news articles filtering. We propose a framework combining information about named entities extracted from news articles with article texts. Named entities are enriched with additional attributes crawled from semantic web resources. These properties are then used to enhance the filtering results. We described various ways of a user profile creation, using our model. This should enable news filtering covering any specific user needs. We report on some preliminary experiments and propose a complex experimental environment and different measures. © 2011 Springer-Verlag.