Recent advancements in mobile technologies have designed a new kind of device: the smart phone. Smart phones have become quintessential for a large fraction of people in developed as well as developing countries in their day to day life. The device integrates a number of sensor technologies for automatic observation which can provide data with high accuracy & precision, can be programmed to interact with the user, and can have communication with remote researchers. This allows relentless and cost-effective access to erstwhile inaccessible history of data on everyday social behavior, such as intimacy of people, phone calls, and movement patterns. In this paper, the concept and applications of reality mining are summarized and its significance towards human behavior analysis is illustrated. The sensor data collected from smart phones is experimented with survey data by applying machine learning algorithms to predict human behavior in social networks. This survey conjointly highlights reality mining applications, future issues and challenges.