Abstract: Aimed at the problems of high complexity and low efficiency of existing technology for Web of Things (WoT) resources management, an intelligent WoT resources management mechanism based on probabilistic topic model is proposed. This mechanism automatically extracted topic information from the WoT resource description, used machine learning algorithms to classify the topic, and published the resources in the P2P network constructed by the WoT gateway based on the classification results. Simulation results show that this mechanism has higher management efficiency and accuracy than text-based and semantic-based methods. Distributed storage of resources based on the topic also improves the usability of the application. An example of this mechanism in the intelligent management of WoT resources was given.