A Web Search Assistant on Google-hard Problems

• Main Authors: Teng-Yuan Chan, 詹登淵
• Other Authors: Yuh-Pyng Shieh
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/10328230178146588116

碩士 === 銘傳大學 === 資訊工程學系碩士班 === 98 === Search engine has been an indispensable tool in our daily lives. However, search engine cannot always satisfy users’ information needs. There are still many problems hard to be solved by Google. Such problems are called “Google-hard problems” in this paper. Information extraction scheme was built to resolve the Google-hard problems. The scheme extracts information from depth-zero and depth-one web pages. First, text is split into sentences and sentences are split as a term sequence by longest term first method. Second, sentences perform Chinese word segmentation by N-gram method. The scheme provides a relationship term list. Another capability uses features and benefits of Term-Clip algorithm. Text mining is performed through learning mode to collect appositional terms with the same properties as the query terms. The scheme helps users to speed up the search process. Users can handle the term list to change search direction. This is a new concept of search modes.