Recommend Items for User in Social Networking Services with CF

摘要

In this paper we focus on the algorithm for prediction task involves predicting whether or not a user will follow an item that has been recommended to the user in social networking services. Items can be person, organizations or groups, which is sponsored by Ten cent Weibo as KDD Cup 2012. We evaluate a range of different profiling and recommendation strategies, based on a subset of large dataset from KDD Cup 2012.