基于海量数据的多模式匹配算法研究

摘要

Traditional pattern matching algorithm will significantly reduce its efficiency when dealing with massive data and massive keywords. The construction of keywords status tree makes the completion time of matching algorithm independent from the scale of the keyword,therefore the speed of analysing customer characteristics from text content information has been accelerated one hundred thousand times. The triggering marketing engine developed with this algorithm will launch the marketing activities at right time and on right customers automatically through the corresponding characteristics of customers gaining from feature database and the default marketing motivation strategies.%传统模式匹配算法在面对海量数据和海量关键字时效率都会大幅降低,通过构造关键字状态树,使完成匹配的时间跟关键字的规模无关,在文本内容信息分析客户特征时的速度提升了十万倍.应用该算法开发的触发式营销引擎,通过在特征库获取客户的相关特征和预设的营销激励策略,在适当的时间向适当的客户自动发起营销行为.