Natural language understanding for query in web search - 2

Abstract

Nowadays, internet has becoming more and more popular, and search engine is becoming a gateway to the World Wide Web. The objective of the project is to apply natural language processing and to understand and restructure the web query, in order to facilitate better Web search result. We also want to successfully establish a software system based on Windows for natural language processing of Web search query, written in Microsoft .Net environment. In the project, we decided to use the c#, one of the applications of .Net environment, to develop our project. And we also decided to restructure the web query based on the relevance on the nutrition.We want to use the googleAPI to extract the search results from google, when we entered the query. Through passing the query to the google, a list of search results, the summary, title, and the hyperlink of the web page, ranked by google, can be obtained. Next, we can use the web documents get from google to re-rank the documents. Vector space modeling will be adopted by us in ranking the web documents. We tried to rank the documents in 3 approaches, that is the static approach, the dynamic approach, and the dynamic approach with feedback. And the results obtained from these approaches will be evaluated using statistical approach.