Research On Semantic Annotation In Search Engine

Search engine technology brings significant impacts on people's life, work and study. With the accumulating information on the Internet, people have to spend more time on screening information. Hence, how to use semantic matching technology to optimize the efficiency of search engines has become an increasingly pressing issue.This thesis first introduces the search engine operating principle based on Crawler, Inverted Index and Query Order, and introduces problems and challenges concerning those three parts. In the analysis of keyword searching method combined with the background of the increasing webpage on the internet, this thesis analyzes the influencing factors of recall ratio and precision ratio respectively, presents the ideas to solve these problems, and which drives two development estimation about the search engine:one is that the search engine needs to be interconnected; and the other is that the precision ratio will be the main obstacle.In the issue of raising accuracy, this thesis intends to employ the semantic matching technique for the compensation of lack of keyword. Referring to some related literature, this thesis also analyzes the properties of semantic analysis and some previous achievements in this field, such as plain semantic analysis method, semantic field and semantic network. In the later section, this thesis discusses the research findings of semantic definitions and semantic measurements.In the issue of semantic annotation, this thesis summarizes the semantic web technology in a popular way and highlights its advantages in dealing with massive data in search engine. To solve the problem of accuracy limitation, this thesis artfully brings the semantic annotation technique into the search engine and realizes it in a program. In the last section, it discusses the research thought in automatic semantic annotation and introduces the procedure and implementation of achieving automatic annotation based on statistics.