| Summary: | 碩士 === 國立暨南國際大學 === 資訊管理學系 === 100 === After fuzzy theory was propounded, how to apply it to database systems has been widely discussed. It makes database systems to have the capability to process vague data and queries. Most researches do not store membership functions in database systems, which manipulate membership grade in client-side. Therefore, Database Management System (DBMS) has no concept of membership grade.
In this thesis we propose a fuzzy database platform with three features. First, a table of membership functions and three fuzzy operators are defined to allow data and queries be numerical or fuzzy linguistic. Secondly, methods converting fuzzy operators to standard SQL syntax by “nested select” clauses are discussed. For the converting process is tedious and error prone, we provide a library implemented in C# to automate it. By these three contributes, any Relational Database Management Systems (RDBMS) which support standard SQL syntax can be extended to become a fuzzy database platform. A Plants-Recommending System has also been constructed to demonstrate the practicability of the proposed fuzzy database platform.
|
|---|
