Indonesian News Harvester and Recommender System

To provide convenience for the user that frequently read the news, a system to gather, classify, and provide news from several news websites in one place
was needed. This system utilized a recommender system to provide only relevant news to the user. This research proposed a system architecture that used vector space model, and Rocchio relevance feedback to provide specifc news recommendation to user’s feedback. The results are that the proposed system architecture can achieve the goal by using fve levels of feedback from the user. However, the time needed to gather news is increasing exponentially in line with the number of terms gathered from articles.