Orsborn, K. (1993). Modeling of Product Data Using an Extensible O-O Query Language. Technical Report LiTH-IDA-R-93-15, Department of Computer and Information Science, Linköping University, Sweden. (bibtex),
Abstract: This work shows how a state-of-the-art object-oriented extensible query language, AMOSQL, can be used to model and manage product information. An example presents an outline of a product model for single sheet metal parts. It is further shown how the query language can be used to model, retrieve and update the product model. It is also shown how the query language can be extended with additional functionality, required by the domain. The implemented mod el describes a subclass of product models for mechanical designs which for instance are valid within the aircraft indus try. The work has been carried out in cooperation with Saab Military Aircraft AB. The concept of product models and data management systems for engineering data are becoming increasingly impor tant in industry to support rational product development and manufacturing. It is possible to make considerable reduc tions in production costs and development time, and an efficient information management is becoming a strategic issue. A product model should be able to represent a sufficient level of structure and behaviour of the real product in order to make fair judgements of its characteristics in the real world and to be able to provide it with these characteris tics. The advantages of the product model concept for managing engineering data includes - besides these explicit ad vantages in using such a system - an increase in efficiency for development and maintenance of these systems. Object-oriented techniques, including object-oriented databases and query languages are well suited to reduce sys tem complexity. Their applicability are especially suitable to engineering applications which consists of large amounts complex data and relationships. Specifically, a main-memory object-oriented database, like AMOS, is combining high-level modeling with high execution efficiency. The use of deductive object-oriented queries offers several advan tages over conventional programming. A declarative description of product data is more transparent and flexible than a procedural and thus becomes more efficient. Declarative models are easier to describe, inspect and understand and thus become more transparent. A declarative modeling with an object-oriented query-language is compact and (de)composable and make the domain modeling very flexible and powerful. This problem-oriented modeling approach will naturally be isomorphic to the problem domain. Query languages also makes it possible to make advanced ad hoc queries concerning the contents of the database. This might be demanded by advanced users and is quite useful since it is impossible to foresee the complete information need. Advanced object-oriented query-languages also provides ob ject views capabilities, supporting data independence and evolution. The extensibility of AMOSQL provide powerful means for flexible domain modeling and a rule system is being included that provides a mechanism for knowledge management. The advantages of these features are, of course, varying for different phases in the system or application life-cycle and for different types of application users and system developers. However, the declarative modeling using an object- oriented extensible query language supports an incremental and iterative development, maintenance and evolution of product models as well as product modeling systems. It will also facilitate a reuse and evolution of design, including domain conceptualizations and knowledge, in its ability of application and data independent representation.
CS Dept TR Overview