Manufacturing Resource Classification (MRC) is a prerequisite for optimal manufacturing resource allocation in the Industrial Internet Platform (IIP). Some key challenges are the inaccurate description of digital manufacturing resources (MRs) and capabilities, the large scope of resource search, and the slow response time of the cloud platform. To address these issues, this paper proposes a framework for classifying resources based on Video Semantic Description (VSD) and Event Description (ED) and related implementations to support the optimal allocation of MRs in the IIP. Firstly, we propose a framework for accurately and efficiently classifying MRs, the structure of which mainly consists of real-time digital descriptions of MRs, construction of manufacturing event clouds, and similar classification of manufacturing events. In addition, we propose a real-time numerical description method for MRs to address the real-time transmission and accurate description of IIP platforms, an event cloud construction method to shorten the search scope of resource allocation, and an event cloud similarity classification method to reduce the associated computational redundancy, which can support the accurate and efficient classification of MRs.