On-line information retrieval systems often offer their users some means to tune the query to match the level of granularity of the information request. Users can be offered a far greater range of possibilities, however, if documents are indexed with coordinated index concepts. Coordinated index concepts are compound index concepts that express a relation between concepts that function as simple subject descriptors, as in cures(headache, aspirin). We show the point by giving a functional specification of a query engine that can handle both simple and coordinated index concepts, Boolean combinations, and superconcept expansion. The Condorcet Query Engine, a prototype query engine that can be run over the World Wide Web (WWW), demonstrates the feasibility of these ideas. The combination of known and novel techniques ensures a smooth migration path from existing query engines to more advanced engines that handle coordinated index concepts.

Presents a particular way of building ontologies that proceeds in a bottom-up fashion. Concepts are defined in a way that mirrors the way their instances are composed out of smaller objects. The smaller objects themselves may also be modeled as being composed. Bottom-up ontologies are flexible through the use of implicit and, hence, parsimonious part-whole and subconcept-superconcept relations. The bottom-up method complements current practice, where, as a rule, ontologies are built top-down. The design method is illustrated by an example involving ontologies of pure substances at several levels of detail. It is not claimed that bottom-up construction is a generally valid recipe; indeed, such recipes are deemed uninformative or impossible. Rather, the approach is intended to enrich the ontology developer's toolkit.

The estimation of time delay between simultaneously recorded EEG (electroence-phalographic) signals is important for the determination of the area in the brain which is responsible for the synchronous activity seen during seizures in epilepsy patients. Methods based on the assumption of linear pathways of propagation in the brain -like the cross-correlation function- have been proposed with limited success. We propose a new analysis method -based on information theory- which does not require this assumption. In our approach, the Average Amount of Mutual Information (AAMI) is computed between pairs of signals, for a range of lag-values, analogous to the cross-correlation function. Estimation of AAMI requires the estimation of the simultaneous probability density function of the signal pairs. For this purpose, an iterative estimation procedure has been developed. Simulations on artificial signals, as well as applications to real EEG signals, show that this procedure is an improvement over the use of the cross-correlation function.