Picking a place to eat is a difficult decision, especially when there are numerous choices in cities such as NYC. Traveltant1 is a social network-based recommender system that suggests places to eat based on the following factors: the user’s preference, their friends’ preferences, and the popularity of the restaurant. Traveltant mines the preferences of the users and their friends from their Facebook accounts and also obtains a list of restaurants and their popularity from the location-based social network Yelp2 to recommend dining places cross the two social networks. We evaluated our system using volunteer subjects who rated the recommendation results using 14 different models representing different combinations of factors. The results showed that personal preferences play the most significant role in restaurant selection process for most of the people; yet, friends’ preferences can be effective to a recommender system to overcome some scenarios in recommender systems where personal preferences do not exist as in the cold start problem.