The Research And Implementation Of Online Customer Service Robot Based On Lucene

In recent years, with the increasing amount of various new businesses in enterprises, the growing operational complexity of business, and the rising volume of users, almost all customer service departments in enterprises are under enormous pressure: the increasing number of users, and the current single model in dealing with customer service have caused man-service channels to be overcrowded which results in call congestion. Therefore, the service can not meet the needs of users, which reduce users' satisfaction with customer service. Enterprise has to increase the size of service system and the scale of service personnel to meet users' needs. It will definitely increase the cost of training service staff.Purely from the business point of view, the traditional man-service can not totally meet the requirements of customer service in enterprise. So there is an urgent need of using computer automated responds to solve the problem. The introduction of online customer service robot, which answers the simple, repetitive questions, can greatly speed up customer service response time, expand system access capacity, and improve the efficiency of customer service.The thesis researches and implements the system of online customer service robot. This system is based on the full-text retrieval engine package of Apache Lucene. The thesis elaborates on the technology of full-text retrieval, Apache Lucene and the concerned theories of Chinese participle, analyses the current status and demand of enterprise customer service, and makes clear the ultimate goal of the system. The system is analyzed and modeled by UML, describing by use case diagram and activity diagram. Then, the thesis carries on the system's overall architecture, network-building programs as well as the various components of the design scheme in detail. Meanwhile, the thesis improves the participle algorithm of language analyzer in Lucene, and designs the search strategy for this system. At last, it also analyzes the system which has been tested in a company, and shows the running effect of the system.