Semistructured Computation

Abstract

In conclusion, we have argued that semistructured data and mobile computation are naturally related, because of a hidden similarity in the problems they are trying to solve.

From our point of view, we have discovered that the Ambient Calculus can be seen as a computational model over semistructured data. As a consequence, type systems already developed for the Ambient Calculus can be seen as weak schemas for semistructured data. Moreover, the Ambient Logic, with some modifications, can be seen as a query language for semistructured data.

We have also discovered that it should be interesting to integrate ideas and techniques arising from semistructured databases into the Ambient Calculus, and in mobile computation in general. For example, the generalization of the Ambient Calculus to graph structures, the use of database techniques for modelchecking, and the use of semistructured query languages for network resource discovery.

We hope that, conversely, people in the semistructured database community will find this connection interesting, and will be able to use it for their own purposes. Much, of course, remains to be done.