Mae---a system model and environment for managing architectural evolution
ACM Transactions on Software Engineering and Methodology (TOSEM)
| Hi-index | 0.00 |
Most of the existing schema versioning approaches are developed for single time-dimension temporal databases. For Bi-temporal databases, which allow both retroactive and proactive updates to the schema and database, these approaches introduce many problems and perform inefficiently. In this paper, we first review two schema versioning approaches, single table versioning (STV) and multiple table versioning (MTV) approaches, and the proposed partial multiple table versioning (PMTV) approach. We prove the correctness of PMTV by showing it producing equivalent results to STV, and then compare these three approaches in database conversion and temporal query processing. The comparison shows that the proposed PMTV approach not only solves the problems but also performs more efficiently than the other approaches when applied to bi-temporal databases.