Problem statement: Nowadays, web based social networks have gained a lot of attention. Over one billion users participate, communicate and share information on these networks based on trust. Approach: One important issue on online social networks is how much one person can trust another person in the network whom they are not directly connected. Scholars around the world have introduced a variety of models and algorithms for inferring trust in online social networks. This study takes a deep look at current methods, their advantages and disadvantages. Results: All current methods have weaknesses and cannot be considered the best solution for inferring trust. Conclusion: Considering the fact that web based social networks are based on trust upon participants, a more accurate and fast algorithm is needed to infer trustworthiness and help users gain valuable information.