The ubiquity of mobile phones and web technology has facilitated continuous communication within Online Social Networks (OSNs) such as Twitter, Instagram, Facebook, and LinkedIn. In OSNs, individuals have the capability to share posts and photos with their relatives and friends, as well as to promote and publicize their businesses. However, this also opens up the potential for malicious users to misuse personal information. Consequently, researchers have turned their attention toward proposing trust-related methods aimed at ensuring a secure environment within social networks. Trust management involves the observation and rating of each user based on his activities, ensuring that only trustworthy users are granted permission to share their activities. In this paper “User Activities Based Trust Management Framework for Online Social Networks-TMFOSN” is proposed. Where users trust values are managed and keep getting updated through the different phases like trust computation, aggregation, propagation, and updations. In TMFOSN, user direct and indirect trust is evaluated with reference of Twitter social network dataset. The performance of TMFOSN is compared with other trust estimation methods, where it is outperformed with increased number of user profiles.