A novel web page analysis method for efficient reasoning of user preference

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

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.

Original languageEnglish
Title of host publicationComputer-Human Interaction - 8th Asia-Pacific Conference, APCHI 2008, Proceedings
Pages86-93
Number of pages8
DOIs
StatePublished - 2008
Event8th Asia-Pacific Conference on Computer-Human Interaction, APCHI 2008 - Seoul, Korea, Republic of
Duration: 6 Jul 20089 Jul 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5068 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th Asia-Pacific Conference on Computer-Human Interaction, APCHI 2008
Country/TerritoryKorea, Republic of
CitySeoul
Period6/07/089/07/08

Keywords

  • Recommendation system
  • TF-IDF
  • User preference
  • User profile

Fingerprint

Dive into the research topics of 'A novel web page analysis method for efficient reasoning of user preference'. Together they form a unique fingerprint.

Cite this