With the rapid development of the WoT, acquisition requirements to the data grow with each passing day. However, the existing data acquisition systems usually use the upload mode. Uploading data not only takes a lot of time, but also requires staff to continuously monitor on duty. Aiming at the problem of enterprise management data acquisition, a WoT-based enterprise management data acquisition and analysis system is constructed. The system obtains multiple monitoring data through the data acquisition module and integrates these data through the processing module, which can solve the problems of traditional data acquisition systems. In order to effectively fuse data, based on canonical correlation analysis (CCA), the multiple order statistics correlation discrimination analysis algorithm is proposed as a fusion strategy. Experimental results show that the fusion strategy based on CCA can effectively integrate monitoring data and accurately identify different characters.