Research And Implementation Of Vertical Search Engine Techniques

In information society, how to acquire the information that is useful to us from Information Sea has been an important problem. So Web search engine emerged and began to have an effect on our life.At present, many search engins apply arithmetic and technology of key words information searching and feedback thousands of web pages. For examples, using Baidu search for "search engine" and find 51,200,000 web pages; Using Google search for "search engine" and find 14,600,000 web pages. There are so many results and users cannot look through all of them at all. Also many of them are useless to users. This shows that search engine only has broadness without depth and it cannot meet the users'demands.This problem can be solved by vertical search engine techniques. Vertical search engine is to search for the valuable information and service that provides to some specific area, crowd or demand. The characters of vertical search engine are special, accurate and profound, meanwhile, it has professional coloring. It is distinct with general search engine.Vertical search engine in this thesis is according to open source code Lucene and Heritrix. Major research is composition of search engine, composition of lucene and heritrix and how to implement a vertical search engine.In this thesis, the first step is to introduce the history of search engine, secondly to introduce the structure of search engine and as follow to introduce the composition and major parts of Hritrix and Lucene. And also make some research on depth-first finding, broadness-first finding and trie arithmetic and put forward idea of making inversion index by trie into reality. At last present the whole process of setting up vertical search engine with examples.