Maciej Kiewra

Learn More
The World Wide Web is the most popular area to which information retrieval and recommending systems are applied. The majority of techniques uses the Web site content and usage as a source of data. Nevertheless, modern Websites cooperate with more structured relational databases which can successfully enrich traditional approaches. In This work a new(More)
The paper describes RankFeed a new adaptive method of recommendation that benefits from similarities between searching and recommendation. Concepts such as: the initial ranking, the positive and negative feedback widely used in searching are applied to recommendation in order to enhance its coverage, maintaining high accuracy. There are four principal(More)
A method of iterative preferences discovering is presented in this paper. It is based on the vector space model and a fuzzy classification of the on-line user’s session to precalculated clusters. As a result, the preference vector is created. It measures the user’s willingness to see web pages, products in an e-commerce site, masterpieces in a virtual(More)
  • 1