The Research Of Full-Text Retrieval And Its Relative Security Technology For Chinese

At present, full-text retrieval technique becomes the focus among domestic and foreign scholars. With text message as the search object, full-text retrieval technique permits users to obtain some concerned sections, paragraphs or sentences from the text directly. It is the most effective means to locate required message accurately among masses of message. With the number of information increasing rapidly, full-text retrieval technique is added to many application systems so as to find related message quickly from vast information.This thesis aims to add full-text retrieval technique to SECBase database, which can heighten search efficiency and solve the problem of searching long fields without the support of SECBase database. Based on the deep analysis of full-text retrieval system, this thesis puts forward the architecture of full-text retrieval SECBase database engine. It analyzes the functions of each module in the architecture and relations among each other, then generalizes its specific function into four processing courses, with explanation of every step in detail.The core of full-text retrieval is full-text index. With the base of inverted index, a full-text index model-B~+ -Lists is raised, which takes its feature and frequency into consideration. According to the model, the full-text index is divided into two levels for storage and operation. The first level is the index of index in the structure of B~+ tree. And the second one is location information related to index in the form of Lists.On the base of full-text retrieval in theory, the thesis puts its function of index into practice. It starts with data chart of index, then ends in the pretreatment of text message (here it refers to long field message stored in the data base) and the establishment of full-text retrieval, with some optimizing measures.The thesis touches the security of full-text retrieval data base for the first time, probing into its security technique. With the analysis of its hidden danger as the base, it comes to a settlement that is the storage encryption of full-text retrieval.