An Enhanced Travel Package Recommendation System based on Location Dependent Social Data

Abstract

Location based social networking systems add location as main dimension to the social connections which gives necessities for personalized location recommendations. Travel packages can be personalized by obtaining user preferences, POI attractions and patterns between them from LBSNs. Existing recommendation systems concentrate mostly either recommending locations , travel packages to a single user or not precise enough, just recommending a list of possibly suitable packages to select by a user group. In our system travel packages are personalized to a user group by considering their common interests, social connections among them along with their individual interests, constraints. Recommendations made precise by considering multiple metrics which varies in degree of personalization and time period of evaluation. We built a prototype system and evaluated results based on data obtained from foursquare site. Experimental results prove system recommends effectively in single user scenario and also adapted well to user group scenario.