Research on classification query optimization algorithm in data stream

The classification query method of data stream can not only improve the efficiency of data stream query, also achieve data stream query in the best matching state. The difficulty of classification query of data stream difficulty is how to achieve data matching in the optimal matching degree, the traditional classification query method for data stream is the method based on keyword matching, the effect on a single condition is better, but when there are more query conditions, query efficiency is low and matching degree is poor. To this end, a classification query optimization method on data stream is proposed based on improved TFIDF algorithm, the information entropy between data characteristics and the information entropy within characteristic are viewed as weighting factors of data classification query, nonlinear mapping ability of neural network is adopted to realize weight calculation and the fuzzification of TFIDF algorithm, so as to solve classification query problems of data streams. With actual database to process classification query, experimental results show that, the proposed algorithm for classification query on data stream have greatly improved query efficiency, which has good application value.