Research On Fuzzy Querying Based On Related Database

Almost existing databases are related databases. It is a waste of time and energy to change it into a fuzzy database. So use these related databases is more pratical and significant in fuzzy querying. Traditional related database is only used to present and dispose precise data. Using its condition combination and expansion is not possible to query fuzzy concepts in nature language. In order to dispose fuzzy query conditions, we have to expand the SQL. literature[8] has already expanded "select" expression in SQL and realized the fuzzy query with weight. Although raised fuzzy querying efficiency a lot using weight, facing the complexity and uncertainties of objectives and disconsidering the vagueness of human thoughts but depending on the precise values which user nominates does not fit the pratical situation.Base on literature[8], further research on weight concept starts mainly from 2 points: first, user can obtain the weight of each propoty by establishing 2-dimension matrix elements which was achieved after comparing the importances between each pair of propoties without initial weights. In another fact, in order to fit the vagueness of human thoughts mentioned above, this paper offers the weight concept illustrated by interval weight, calculates the weight by one of Interval Weight Analysis Method - Weight Determination by Group Value Statistics, and then processes fuzzy querying with fuzzy weights by the using matching fomular of weighted conjunction querying and weighted extracting querying and membership function. While raising the efficiency, it comes closer to the nature language of human querying.