Research And Application Of Vertical Search Engine Based On Lucene

Since the built-in language processor in the open-source software package Lucene is not friendly to Chinese word segmentation, designers have to design their own Chinese word segmentation model to fit for the system. Currently there are three major Chinese word segmentation methods such as mechanical segmentation, in which the maximum matching method (MM) plays a leading role and is widely researched and applied.After focusing on research into MM, this thesis proposes a new MM. The new MM increases the precision of word segmentation and adheres to the long-term priority principle fully. The efficiency and precision of the method is simulated by experiment, and the experimental data shows that the method increases the word segmentation precision while the efficiency is acceptable, thus the validity and practicability of the new MM is proved.On the basis of the new MM, we design and implement a bus routes query system in terms of the structure of vertical search engine system, which can run on the Tomcat server. The whole structure of the system is given, and the course of it is depicted in detail. In the final, we evaluate the effect of the running system and its capability.