Query processing for integration of information sources based on a self-descriptive object model
✍ Scribed by Koichi Munakata
- Publisher
- John Wiley and Sons
- Year
- 1999
- Tongue
- English
- Weight
- 635 KB
- Volume
- 30
- Category
- Article
- ISSN
- 0882-1666
No coin nor oath required. For personal study only.
✦ Synopsis
When data are obtained from heterogeneous distributed information sources, all of the data do not always have the same structure, but they are sometimes different in data structure from instance to instance. As a data structure to integrate such semistructured data, we present the self-descriptive object model. This is the model in which every instance of the data has a label that represents the self-meaning of the instance. In this paper, we describe a query processing method for a mediator based on the self-descriptive object model. According to this method, the mediator can obtain data from the heterogeneous distributed information sources, convert, join, and select them, and create the query result. The mediator is given a declarative description for creating a view with a language called RMSL. In compliance with a query request, on the basis of the RMSL and a query, the mediator creates and executes a query processing plan which is composed of five kinds of nodes: a source node, a predicate node, a conversion node, a view node, and a union node. It is hereby shown that it is possible to integrate flexibly the semistructured data which were difficult to deal with in a conventional relational database and an object-oriented database.