Semistructured Computation

This paper is based on the observation that the areas of semistructured databases [1] and mobile computation [3] have some surprising similarities at the technical level. Both areas are inspired by the need to make better use of the Internet. Despite this common motivation, the technical similarities that arise seem largely accidental, but they should still permit the transfer of some techniques between the two areas. Moreover, if we can take advantage of the similarities and generalize them, we may obtain a broader model of data and computation on the Internet. The ultimate source of similarities is the fact that both areas have to deal with extreme dynamicity of data and behavior. In semistructured databases, one cannot rely on uniformity of structure because data may come from heterogeneous and uncoordinated sources. Still, it is necessary to perform searches based on whatever uniformity one can find in the data. In mobile computation, one cannot rely on uniformity of structure because agents, devices, and networks can dynamically connect, move around, become inaccessible, or crash. Still, it is necessary to perform computations based on whatever resources and connections one can find on the network. We will develop these similarities throughout the paper. As a sample, consider the following arguments. First, one can regard data structures stored inside network nodes as a natural extension of network structures, since on a large time/space scale both networks and data are semistructured and dynamic. Therefore, one can think of applying the same navigational and code mobility techniques uniformly to networks and data. Second, since networks and their resources are semistructured, one can think of applying semistructured database searches to network structure. This is a well-known major problem in mobile computation, going under the name of resource discovery.