It is difficult for the users to reach the most appropriate and reliable item for them among vast number of items and comments on these items. Recommendation systems and trust/reputation systems are one of the solutions to deal with this problem with the help of personalized services. These systems suggest items to the user by estimating the ratings that user would give to them. Use of trust data for giving recommendation has emerged as a new way for giving better recommendations. In the literature, it is shown that trust based recommendation approaches perform better than the ones that are only based on user similarity, or item similarity. In this paper, a comparative review of recommendation systems, trust/reputation systems, and their combined usage is presented. Then, a sample trust based agent oriented recommendation system is proposed and its effectiveness is justified with the help of some experiments.