Application of Hierarchical Optimal Random Walk Algorithm in Data Specification of College Sports Training Intelligent Platform

This paper studies the application of hierarchical optimal random walk algorithm in data specification of college sports training intelligent platform. The algorithm proposed here is aimed at the representation of a neighbor node. Since the user may have multiple neighbor nodes, the final core vector representation of the user is the fusion of all its neighbor nodes. The representative sample network of the original social network can be obtained by sampling technology, so as to achieve the purpose of reducing the scale of network data. Hence, with these supporting technologies, the designed model is considered with the hierarchical optimal random walk algorithm. The experimental results have presented the effectiveness.