2008 | OriginalPaper | Buchkapitel
A Novel Web Page Analysis Method for Efficient Reasoning of User Preference
verfasst von : Seunghwa Lee, Minchul Jung, Eunseok Lee
Erschienen in: Computer-Human Interaction
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
The amount of information on the Web is rapidly increasing. Recommender systems can help users selectively filter this information based on their preferences. One way to obtain user preferences is to analyze characteristics of content that is accessed by the user. Unfortunately, web pages may contain elements irrelevant to user interests (e.g., navigation bar, advertisements, and links.). Hence, existing analysis approaches using the TF-IDF method may not be suitable. This paper proposes a novel user preference analysis system that eliminates elements that repeatedly appear in web pages. It extracts user interest keywords in the identified primary content. Also, the system has features that collect the anchor tag, and track the user’s search route, in order to identify keywords that are of core interest to the user. This paper compares the proposed system with pure TF-IDF analysis method. The analysis confirms its effectiveness in terms of the accuracy of the analyzed user profiles.