摘要
本文建立了非第一范式历史数据库的一个代数结构，在保持标准关系代数的基础上，实现了不同时间属性值的比较，从而能通过代数运算实现属性动态变化情况的查询．
This paper deals with query languages of N1NF historical databases and obtain a goodalgebraic query language. Historical database is a kind of temporal database that can store theentity's history of attributes throughout certain period, and NINF historical databases are the onesin the form of N1NF. Because of N1NF, all attributes of the same entity during the whole period ofinterest are stored in only one tuple and this is called the property of integrity which makes it highprospect of applications. However, studies should be further carried out before it can be applied.Query languages is one of the most important subjects in the study of databases and severallanguages have been suggested for N1NF historical databases, which can be divided into two groups:the algebraic one and the calculus based one. Through comparison, it can be found that thealgebraic ones are weaker than the calculus based ones and this has impeded the implementation of this kindof databases. To solve this problem, a new algebraic language is proposed. This language iscomposed of the following algebraic operators: set union, set difference, projection, IF selection,WHEN selection, renaming, thita join, adding serial number, decomposition and computation.Through this algebra, all the query abilities in the standard relational database is maintained so thatall information at each given time instant can be retrieved. Moreover, due to the fact thatcomparison of attributes at different time instant can be accomplished, the history of dynamic attributechanges can also be retrieved.
出处
《计算机学报》
EI
CSCD
北大核心
1998年第S1期190-194,共5页
Chinese Journal of Computers
