Social graph based location recommendation using users' behavior: By locating the best route and dining in best restaurant

Abstract

Study of users' behavior is increasingly becoming a topic of research because of innovations in web. Out of many research areas, check-ins on Facebook is a rather best way to connect with users' places of interest. Services such as location recommendation would definitely benefit from such research. There are times when people are stuck in situations where they are completely new to a particular place due to lack of location information and trends based recommendations, basically they don't have idea to explore a specific location. Therefore many times users don't have information to plan their trip as to fetch maximum benefits from the trip. The prime objective of this research work is to understand, analyze and suggest location and restaurants on the basis of user behavior. To achieve this, we developed an application `Travel Best' which focuses on extracting check-ins from the Facebook accounts of the users and generate trends based suggestion for users to travel best. Within this, users can explore nearby trending regions of places within a city according to facebook checked-in data. To add on a cross domain flavor to the system, we also analyze the restaurants data based on ratings and reviews for the searched location according to Zomato extracted information. In our proposed approach, we map location based data to suggest location on the basis of shortest path, longest path and most traveled route and visualized our result using graph database. Further to suggest restaurants of the chosen location or chosen path, we apply opinion mining to provide relevant ratings on the basis of different categories namely food quality, service and ambiance unlike overall ratings usually provided by other applications.