tailieunhanh - Fuzzy improvement of the SQL

Structured Query Language (SQL) is used to obtain data from relational databases. Fuzzy improvement of SQL queries has advantages in cases when the user cannot unambiguously define selection criteria or when the user wants to examine data that almost meet the given criteria. In this paper we examine a realisation of the fuzzy querying concept. For this purposes the fuzzy generalized logical condition for the WHERE part of the SQL is created. | Yugoslav Journal of Operations Research 21 (2011), Number 2, 239-251 DOI: FUZZY IMPROVEMENT OF THE SQL Miroslav HUDEC Institute of Informatics and Statistics, Bratislava, Slovakia hudec@ Received: December 2007 / Accepted: July 2011 Abstract: Structured Query Language (SQL) is used to obtain data from relational databases. Fuzzy improvement of SQL queries has advantages in cases when the user cannot unambiguously define selection criteria or when the user wants to examine data that almost meet the given criteria. In this paper we examine a realisation of the fuzzy querying concept. For this purposes the fuzzy generalized logical condition for the WHERE part of the SQL is created. It allows users to create queries by linguistic terms. The proposed model is an extension of the SQL so that no modification inside databases has to be undertaken. Keywords: SQL, fuzzy query, generalized logical condition, database. MSC: 94D05, 68P99, 90B99. 1. INTRODUCTION SQL is a standard query language for relational databases. The SQL was initially presented by Chamberlin and Boyce [3]. Since then, it has been used for data selection in many relational databases and information systems. SQL is one of the key factors for accepting relational databases and for their use in storing and retrieving vast amount of data. Its advantages among others are: optimised work with relational database management systems and understandable interpretation for users. The SQL uses two-valued logic (crisp logic) in querying process. It means that a small error in data values or in cases when user cannot unambiguously define the criteria by crisp values, may involve some inadequately selected data. This limitation of the SQL can be avoided by fuzzy logic. The fuzzy set theory was initiated by Zadeh [12]. Since then, many researches and applications have been published. Fuzzy queries have appeared in the last 30 years to cope with the necessity to 240 M. Hudec / Fuzzy .