Evaluation of Clustering Techniques for News Recommendation

The information overloading is one of the most significant issues nowadays. It can be seen in different domains, consisting of business, particularly in news. This is more important in link to news websites and web, where the news websites reliability is typically determined by quantity of news added to the portal. Then the most popular news websites include numerous of news articles daily. The classical solution generally used to address the information overloading is a recommendation. In this book we evaluate the MapReduce k-means and fuzzy k-means clustering for a content-based recommendation for news articles, based on Euclidian distance and cosine similarity search. This approach consists of two phases of operation.