Theory Analysis And Application Research Of Intelligent Search Engine

With the geometrical increase in knowledge on the Internet, the search engine, as a tool that the user get information with from the Internet, has been paid more and more intention in people's daily life and work. To put it simply, with the program named "Web Spider", the search engine saves the information gathered from WWW into a database in a defined manner and the database is maintained and updated all day. When the user put in a query, the search engine will find the information that is correlated with the search's key words in the query from its own database and return it to user.This article firstly introduces the search models such as Boolean information retrieval model, vector space model, and probabilistic information retrieval model. By analyzing both the limitations of traditional search engine technology and the solutions to the limitations, it then introduces the principles of intelligent search engines. This article also analyzes and dicusses several major techniques such as knowledge-based Chinese segmentation, automatic summarization, search results ranking, multi-level intelligent search agents, which are used in popular search engines.Now, full-text search is a mature technology. It has good recall rate, but is weak in precision. To address this program, this article discusses the reasons of poor accuracy such as Chinese segemantation, and then implements a news search engine based on Lucene full-text search engine combined with Spider program.Intelligentization is a direction of search engine development. Now the study on it is still at the start stage and needs theoretical breakthrough in certain aspects. The last chapter of this article provides a prospect of search engine's future development.