An intention-based approach to the updatability of views in relational databases
Yoshifumi Masunaga · 2017
Views are essential for relational databases in that they could provide a high degree of logical data independence. However, although views are a useful tool for queries, they present significant problems if insertions, deletions, or rewrites are expressed using views. This is due to the fact that views are not base relations stored in a relational database but virtual in the sense of being merely definitions of queries issued to a relational database. While the view update problem has attracted the attention of many scientists and business people, a complete resolution to this problem is still an open issue unfortunately. In this paper, in contrast to the traditional approaches such as the syntax-based approach and the semantics-based approach, an intention-based approach is presented to resolve this problem. By introducing the pro forma guessing of update intention approach, Cartesian product views and join views become updatable while they are not updatable in the traditional sense. This is due to the fact that in certain cases the user's view update intention can be guessed uniquely by checking the extension of each view update transformation candidate, which is calculated using temporarily materialized views. In addition to the above results, the updatability of other basic views such as union views, difference set views, intersection views, projection views, and selection views is re-examined under the intention-based approach. Based on the result, an algorithm is presented to determine whether a given view, defined arbitrarily by using relational algebra operations recursively, is updatable or not. Compared to the traditional approach, it is shown that the intention-based approach realizes richer solution to view updatability problem.