Developing a mobile location-based collaborative Recommender System for GIS applications

Abstract

Collaborative recommender system recommends items to the user based on what actions that other users in the same environment did in the past. This paper focuses on recommending popular places based on behavior of Global Positioning System (GPS) traces to multiple users. Collecting different reviews of popular places from multiple visitors is an explicit way to help Recommender System to rate these places. A Genetic Algorithm is used in predicting the interest of the user for unvisited locations, based on implicit and explicit ratings. Then, Recommender System sorts the rated places in descending order by their rates. The places that have the highest rates in most number of users are recommended by the system. Collaborative Recommender system advises the user to go to the best place via mobile according to other users' preferences in places and the nearest place from user's location. Then the system shows the direction to this place on the map via user's mobile.