B. Stantic, J. Terry, and A. Sattar (Australia)
Temporal databases, Indexing methods, Point approach, Layered architecture
Most database applications contain some amount of time dependent data because the majority of database applica tions are temporal in nature. While a lot of work has fo cussed on temporal data models and query languages, only a little work has addressed access methods. Research has proved the benefits of adding time dimensions to data and keeping the history of changes, however, the informational benefits of temporal data can, without effective manage ment, be easily outweighed by the costs of poor access times and difficulties in formulating queries. This paper proposes a logical query transformation that relies on the POINT representation of current time. Logical query trans formation enables off-the-shelf Spatial indexes to be used. Testing results indicate a significant computational advan tage with this approach. Further we propose to combine this approach with the previously introduced layered ap proach for temporal DBMS implementations to simplify the formulation of queries and to achieve the full benefit.
Important Links:
Go Back
