Recommendation systems can take advantage of social media in various ways. One common example is combining social relationship into neighborhood-based recommendation systems, under the assumption that social relationship affects individuals’ interest or preference. Although this assumption may not be always true, this paper presents a realistic application, personalized academic paper recommendation system, which social relationship can be closely related to taste. There is an increasing number of academic papers being published each year, but most researchers rely on keyword-based search or browsing through proceedings of top conferences and journals to find their related work. Personalized academic paper recommendation system is designed to reduce their workload. With a collaborative-filteringbased approach, it recommends potentially preferred articles for each researcher in personalized manner. Both computational evaluation and user study demonstrate that our system recommends a useful set of research papers.