본문 바로가기
  • Home

Design and Implementation of a Responsive Website for Personalized News Article Recommendation and Newspaper Production System

  • Journal of Knowledge Information Technology and Systems
  • Abbr : JKITS
  • 2016, 11(2), pp.205-214
  • Publisher : Korea Knowledge Information Technology Society
  • Research Area : Interdisciplinary Studies > Interdisciplinary Research
  • Published : April 30, 2016

Ho-Sang Park 1 EOM. I. K 1 Kim young in 1

1부산대학교

Accredited

ABSTRACT

Recently, there has been an increasing demand in personalized web news article recommendation service that customizes the newspaper according to the user’s preference. However, it can be true that the current research in news services through extracting the people with common interests and similar characteristics is still lacking. This study will give guidance to news recommendation system and creating a customized newspaper by using collaborative filtering techniques and newspaper articles framing techniques. To implement this system, we built a server that collects the users’ scrapped articles information and ratings of the users’ scrapped news articles; the server performs collaborative filtering algorithms to recommend the articles by the similarity measurement. News articles are collected from the web sites of each media outlet by using crawling techniques, and embodied search function by using web portal site’s API. In order to make this service practically usable, user interface was put in consideration during the development, and enabled to be used on PC, tablet and smartphone by embodying responsive web design. The proposed recommend system perform scoring users based on their interests area, and then recommend new newspaper articles to the users by using collaborative filtering techniques considering user scrapped news articles. Experimental results demonstrate that the proposed system can provide useful news to users by using responsive website design, user's newspaper article scrap information and collaborative filtering algorithm. It is highly expected that using the proposed collaborative filtering system would provide more accuracy in selecting suitable news for each individual, which will increase the level of user satisfaction.

Citation status

* References for papers published after 2023 are currently being built.