View Maintenance Aspects
Database systems often define views in order to provide conceptual subsets of the data to different users. Each view may be very complex and require joining information from multiple base relations, or other views. A view can simply be used as a query modification mechanism, where user queries referring to a particular view are appropriately modified based on the definition of the view. However, in applications where fast response times to user queries are essential, views are often materialized by storing their tuples inside the database. This is extremely useful when recomputing the view from the base relations is very expensive. When changes occur to their base relations, materialized views need to be updated, with a process known as view maintenance, in order to provide fresh data to the user.
The use of relational views has long been proposed in relational database systems. The notion of materialized views, or snapshots, was first proposed in ....
- 1.Adiba B, Lindsay B. Database snapshots. In: Proceedings of the 6th International Conference on Very Data Bases; 1980. p. 86–91.Google Scholar
- 2.Colby L, Kawaguchi A, Lieuwen D, Mumick IS, Ross KA. Supporting multiple view maintenance policies. In: Proceedings of the ACM SIGMOD Conference on Management of Data; 1997. p. 405–16.Google Scholar
- 4.Gupta A, Mumick IS. Maintenance of materialized views: problems, techniques, and applications. IEEE Data Eng Bull. 18(2):3–19. Special Issue on Materialized Views and Data Warehousing, June 1995.Google Scholar
- 6.Gupta A., Mumick IS, Subrahmanian VS. Maintaining views incrementally. In: Proceedings of the ACM SIGMOD International Conference on Management of Data; 1993. p. 157–66.Google Scholar
- 7.Jagadish HV, Mumick IS, Silberschatz A. View maintenance issues in the chronicle data model. In: Proceedings of the 14th ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems; 1995. p. 113–24.Google Scholar
- 9.Labrinidis A, Roussopoulos N. WebView materialization. In: Proceedings of the ACM SIGMOD International Conference on Management of Data; 2000. p. 367–78.Google Scholar