The Research And Application Of Real-Time Search Engine For Large-Scale Enterprise System

The development of information technology promoted the rapid development of enterprise informatization. Information is always very large and the growth and update of the data is very fast in the large enterprise information management system. There are defects in both real-time and accuracy for data retrieval in the traditional SE. On the other hand, the retrieval speed will be greatly reduced under the condition of huge amount of information. And the performance can’t satisfy the needs of users, even to influence other normal use of the database. According to the above situation, this paper construct a real-time retrieval system based on Lucene.This paper first analyzes the problems existing in the retrieval system in the enterprise information systems, gives the necessity of real-time retrieval system design. Further study on the structures and the key technology of real-time retrieval system, including index, participle, combined index segments, etc. In addition, discussed the main techniques and implementation methods. On this basis, analyze the demand of the system, probes into the construction of real-time search and optimization scheme, put forward the overall solution of the system and software hierarchical design.For index offline working condition in ordinary retrieval system. This system creative propose the query and index into the same process. Using multiple indexes to work together to ensure the document in the add, remove, update state corresponding to the immediacy index update. Furthermore, this system design a core system, it provides many methods for the use of the outside, encapsulates a series of methods. This module control both data fields and query service, realized the real-time retrieval. In addition, this system use the AJAX, JSP to construct a index interaction. On the basis of the data structure of this system, build a flexible query and management platform.Finally, the system test the effect of the query, and to this effect, compared with the traditional database query performance and effect, proves that the system has reached the expected research purpose and requirements.