A Recommender System for Political Information Filtering
Recommender system has been widely used and showcased its successful stories in e-business area for the last decade. It assists in making profits within a lot of companies by recommending their products that the customers would be interested in. Compared to many successful stories in e-business and industries, however, a recommender system has not been fully exploited in non-profit activities where people need information that is unbiased, accurate, up-to-date and mostly relevant to their interest, especially in politics. Even though choosing a right candidate with appropriate and accurate information is required to voters, it is not easy for them to keep up with the political issues due to the massive amounts of online media and its speed. To address these issues, we suggest a politician recommender system by using two widely used filtering: collaborative filtering and content-based filtering.
In order to build the recommendation system, we first collect public profile of current congress members in Korea and people’s preference ratings to these politicians. These data are preprocessed and used in filtering methods to recommend politicians that a user would be favorable for. We compare the experimental results, and combine the two filtering whether the hybrid approach shows better performance than two individual methods. We anticipate this saves people’s time and effort to obtain information to support their decision and makes people actively participate in political issues.
KeywordsRecommendation system Data mining Content-based filtering Collaborative filtering
- 1.Hahsler, M.: recommenderlab: A Framework for Developing and Testing Recommendation Algorithms. Southern Methodist University, Texas (2011)Google Scholar
- 2.Gorakala, S.K., Usuelli, M.: Building a Recommendation System with R. Packt Publishing, Birmingham (2015)Google Scholar
- 3.Yao, G., Cai, L.: User-Based and Item-Based Collaborative Filtering Recommendation Algorithms Design. University of California, San DiegoGoogle Scholar
- 5.Niwattanakul, S., et al.: Using of Jaccard coefficient for keywords similarity. Paper presented at International MultiConference of Engineers and Computer Scientists, Hong Kong, 13–15 March 2013Google Scholar
- 6.Hamilton, H.J.: Knowledge Discovery in Databases. University of Regina School of Computer Science (2012). http://www2.cs.uregina.ca/~dbd/cs831/index.html. Accessed 10 Feb 2017
- 7.Hyndman, R.J., Athanasopoulos, G.: Forecasting: principles and practice. OTexts (2013). https://www.otexts.org/fpp/2/5. Accessed 10 Feb 2017
- 9.Kim, C.I., et al.: On the development of a course recommender system: a hybrid filtering approach. Entrue J. Inf. Technol. 14(2), 71–82 (2015)Google Scholar
- 10.Shani, G., et al.: Evaluating Recommendation Systems. Microsoft Research (2009)Google Scholar
- 11.Wang, R.: Building a Movie Recommendation Engine with R (2015). https://muffynomster.wordpress.com/2015/06/07/building-a-movie-recommendation-engine-with-r/. Accessed 10 Feb 2017
- 12.Suh, B.W.: Development of Content Recommendation System Algorithm. Broadcasting Trend & Insight, vol. 5 (2016)Google Scholar
- 13.Information of 300 Politicians. http://watch.peoplepower21.org/. Accessed 10 Feb 2017