Abstarct: Today, there are media or social networks to spread ideas and views or a tool for entertainment and fun in various forms and tixtles, but the main purpose of all of them is to communicate and exchange information with each other and information is moving in different formats (events). The purpose of this study is to identify events that are published in cyberspace and to analyze their content. Every event on a social network is created and published by a person or a group of people in a certain period, and after publication, it is reflected by users (for example, its contents may be reduced or added to it, or changes in its nature may occur). . To follow the events of a social network, the path of that event must be traced between users as network nodes. Recent studies have used methods such as neural networks and Bayesian networks or decision trees to classify documents by type of content, which often works well with large amounts of data over time. In the proposed method, using the ant algorithm, each ant is considered as a crawling event on the social network. Each Ant has a memory in which information is collected about the path that emerges from the trajectory to analyze connections and similarities. This algorithm can also work properly with little data. baxsed on the obtained results, the ant algorithm will give us the ability to simulate the behavior of event navigation, with high execution cost and acceptable accuracy percentage, therefore, the proposed approach of using genetic algorithm and refrigeration, to achieve higher accuracy and lower cost baxsed on users' survey behaviors. The average accuracy obtained in the results of the proposed method has an identification accuracy of 94.25%.