Recent development of internet and its applications has become the source for research in service recommendation. Among them, point of interest (POI) recommendation based on user behaviour has enticed a decent attention. Along with quality of recommendations, the utilization of user feedback has grown to be a key part in the POI recommendations. While implementing similarity-based methods in conventional recommender systems, it faces various issues such as trustworthiness, sparsity and cold-start. The commonness and popularity of social network facilitate people to interact with different users and generate massive data such as user relationships, ratings and interactions. Thus, integration of trust relationship of user in location based social network along with their feedback for POI recommendation is the motivation of this work. In this article, we present a POI recommendation method based on trust enhancement in social networks known as social pertinent trust walker (SPTW). Initially, the level of trust between users in social networks is calculated through matrix factorization technique. Then, SPTW with high probability location category algorithm helps to generate POIs as list of recommendations. Experiments on real-world datasets are conducted to evaluate proposed algorithm for accuracy. Results reveal the effectiveness of approach and quality of recommendations is better, when compared to existing algorithms. © 2017, Springer Science+Business Media, LLC.