부산대학교 도서관

In order to further improve the operation and management efficiency and quality of the bid evaluation expert database of power grid enterprises, comprehensively improve the professional skills and quality of bid evaluation experts, and promote the healthy development of bidding work of power grid enterprises. This project plans to build a power engineering bid evaluation expert database based on LSI(Latent semantic index). And LSI technology is applied to the retrieval technology of expert database. First of all, after the original data information of the expert database is labeled and weighted, a series of feature vectors are formed, and the combination of these vectors becomes a word-document matrix. Secondly, the data source document matrix is processed, features are extracted and related features are clustered to form a potential semantic space. Finally, an LSI is established and searched through the latent semantic space. The test results show that this method can retrieve more valuable information in accurate meaning vocabulary retrieval, and the precision rate is higher than that of traditional retrieval methods. LSI technology is applied to the expert database of power engineering bid evaluation, and the results obtained are incomparable to those obtained by ordinary indexes, which really improves the search efficiency of the system.