Many data mining techniques have been proposed for mining useful patterns in text documents. However, how to effectively use and update discovered patterns is still an open research issue, especially in the domain of text mining. Since most existing text mining methods adopted term-based approaches, they all suffer from the problems of polysemy (word having multiple meaning) and synonymy (A word having same or nearby same meaning as another word).Instead of keyword based approach which is typically used in this field, pattern based model containing frequent sequential pattern is employed to perform the same concept of task. In this study we propose two approaches based on the use of pattern deploying and pattern evolving strategies. The performance of the pattern deploying algorithms and pattern evolution algorithm for text mining is investigated on the Reuters dataset RCV1 and the results show that the effectiveness is improved by using our proposed pattern refinement approaches.