A Content Recommendation System Based on Category Correlations

In Web 2.0, there is no clear distinction between users and webmasters. Many internet users are not only information consumers but also information providers. There are lots of information in the Web and most people can find what they want by searching the Web. One problem of the large number of data in the Web is that we often spend most of our time to find a correct result from search results. Thus, people start looking for a better system that can suggest relevant information instead of letting users go through all search results: We call such systems recommendation systems. A recommendation system is often based on collaborative filtering (CF). A traditional CF approach requires lots if user data so that a recommendation system can compute the similarity of user preferences and suggest items based on the computed preferences between users. This approach has two problems: sparsity and cold start. We propose a different CF approach based on the category correlation of contents. First, we show how to compute the category correlations of contents. Then, we design a new recommendation algorithm based on the category correlations to a user with certain preferences. Note that our approach does not require computing the preference similarity between users.