A Chinese Information Retrieval System Based On Boolean Model And Extended Boolean Model

With the rapid development of Internet in China, more andmore Chinese documents are readily online. The Internet provides importantand convenient repositories for reference information, but it is very difficultto find the relevant information form Internet. Information retrieval systemsare used to help people to find the information they want. IR model is thecore issue, the model accuracy reflected the connection of the user retrievalrequest and the user’s needs documents, which directly affects the IR systemperformance and efficiency. It brings about the development of linguisticprocessing technology. As an example of ideographic languages, they searchthe network, find data information, process and extract useful information forthe user Chinese is very different from Indo-European languages. In thisthesis, we discuss some key issues in Text Retrieval. The major content andcontributions are as follows: this paper focuses on the relevant issues ofChinese Text Retrieval. Based on the system structure of IR model, thispaper gives a logical model of Chinese Text Retrieval based on Boolean andExtended Boolean Model. Text feature extraction and representation is thefundamental operation for Chinese Text Retrieval. Four phases of documentrepresentation are word segmentation, virtual-word removal, featureextraction and feature item weighting. This paper provides a method offeature item weighting based on Extended Boolean Model weightingalgorithm. User query processing accepts Chinese word sequence typed intothe system by the user, segments the words, calculates the similarities ofindexed documents to the query, and returns results of retrieval to the user.