We show how quantified constraints expressed in a sub-language of first-order logic, againstashareddatamodelthatisfreetoevolve, provideanexcellentwayof transporting domain-specific semantics along with the data. In this form it can be processed automatically by various intelligent components, instead of requiring human intervention, as it would do if expressed in natural language. It can also be combined with other constraints, by algebraic transformation against a common data model, and then passed to an appropriate solver. These techniques have been tested in a classic e-business application scenario: configuring a product from parts selected from e-vendors' catalogues, whilst conforming to requirements specific to the parts, expressed as mobile constraints.
Kit-ying Hui, Peter M. D. Gray, Graham J. L. Kemp,