Point-of-interest (POI) recommendation systems provide recommendation of places to users based on their behavior or activities. Checking behavior features from many Location Based Social Network (LBSN) applications combined with POI recommendation systems, provides better location-based services and benefits consumers and businesses in many areas. For users, they assist consumers in discovering interesting places, while for industries, they distribute commercials to target consumers and increase industry benefits. LBSNs may also use a POI recommendation system to have more target customers in return. This research aims to improve the precision of the POI recommender system for individuals as well as social groups in LBSNs by overcoming limitations of current models. The revised model was designed to support individual as well as group recommendations. In terms of individual recommendations, the proposed model is intended to take friendships between users into consideration and their impact on LBSNs and their POI ratings. Furthermore, for group recommendations, consideration was given to the aggregation of individual user recommendations. The improved model was implemented on a Gowalla dataset and results were compared with current models. The experimental results showed higher precision in POI recommendations for individuals in LBSNs.
|Number of pages||12|
|Journal||IAENG International Journal of Computer Science|
|Publication status||Published - Sep 2020|