With development of computer and network technology, string matching are widely used in information retrieval, intrusion detection, network data analysis and other fields. String matching algorithms and experimental analysis of previous studies are set in the pattern scale to tens of thousands of circumstances, for the large-scale pattern set algorithm was not conducted under the in-depth analysis. This paper studies the classic multi-pattern matching algorithm Wu-manber algorithm is proposed under large-scale pattern set algorithm based on Wu-manber three improved algorithms: hash based on the shortest length of the key algorithm, the value of pre-screening algorithm multi-hash, group comparison algorithm. Experiment results show that these improvements can enhance the capability and performance of classic Wu-Manber algorithm.
