| Data dependency means all kinds of connections among data. For example, key is a kind of dependency. The happening of data redundance is connected with data dependency. Data dependency is a central notion in database design theory through which we can mormalize attributes, instrinsic semantic restrinction. Because traditional data dependency can not describe danamic dependency relation among attributes well in single granularity for temporal database, that is, it can not adapt to the danamic property of temporal data. So, in temporal database, in order to describe danamic semantic restriction well among attributes, this paper will learn from and expand traditional dependency theory and research temporal database dependency question in multiple granularities.The purpose of a good database logical design is to eliminate dataredundancy and insertion, deletion and update anomalies. Temporal database is the same case. In daily life and real production application, the granularity of system records the time of data efficiency and event happening usually is minute, hour, day, month, year and so on, this kind of multiple granularities usually consists in temporal database, but these granularities meet a kind of totally ordered relation. A temporal scheme with totally ordered temporal type sets has more super properties than general temporal type sets. On the basis of the process of normalization of relational database, the concepts of totally ordered temporal functional dependency sets and totally ordered temporal multivalued dependencies(TOTMVDs) with multiple time granularities based on the theory of data dependency with multiple time granularities are introduced. An axiomatization for TOTMVDs is given. At the same time, the definition of totally ordered temporal scheme, totally ordered temporal module also and their properties are given, then, define totally ordered temporal forth form and give the algorithm for totally ordered temporal forth form decomposition and prove the algorithm. |
PDF Full Text Request