SMS marketing is growing at a fast rate in the past few years. As this business is widely used, the requirement for the filtering and matching function of the short message is becoming increasingly higher. Having considering some characteristics of SMS advertising, we present an algorithm for effective text filtering and matching in this article. More specifically, the algorithm is briefly described as followed: firstly, text is filtered based on its eigenvalue and secondly using bit-vector method and Levenshtein Distance algorithm to match two candidate short messages thereby realizing the efficiency of mass short message filtering and matching function. This paper describes the details on calculation of text eigenvalue and presents the analysis on the effectiveness and correctness of this algorithm.