No abstract

The paper describes a connectionnist architecture for an information retrieval system The concept of a "dynamic" thesaurus is considered, in order to improve the construction of a documentary base and to perform associative information retrieval.The authors suggest a set of rules for activating cells in order to start an activatiodpropagation process, on which the associative information retrieval is based while a selforganising mechanism is also activated. These two notions allow to develop automatic query reformulating and dynamic restructuring of the information base.

Purpose -The purpose of this paper is to add precisions to a method, to demonstrate the convergence, to explain time and memory space complexities and new simulated results on a non-linear partial derivative equation system governing corona-electrostatic electric field for granular mixture separation. Design/methodology/approach -The method converts the non-linear partial derivative system into an iterative system of linear equations. Using the well-known finite difference approximation, a numerical solution is computed very quickly. Findings -The paper gives the truncated error and the approximation error to conclude to the convergence. Originality/value -The paper shows the fast numerical solution leads to confidence in the numerical approximations for the comprehension of the phenomenon. Extends the corona-electrostatic electric field for granular mixture separation to new geometries easily.

No abstract

With the development of computer capabilities, memories and network abilities, we need more efficient and robust algorithms to manage databases and to store and retrieve the relevant information for the user. The aim of this work is to automate the construction of a neural network Information Retrieval System (IRS) adapted to a medical image database. The user builds queries and the system must retrieve the relevant documents or images. Queries are groups of keywords or items associated with relevant images. In our approach, the set of queries and the binary relevance judgments on the documents constitute complex learning data associations. There are two phases in the automatic construction of the IRS. The indexing phase builds the learning data base and then a specific learning algorithm builds the neural network. For the system to be able to immediately learn these complex data, we have developed a new specific algorithm. It allows a perfect learning of a binary logical table in a stepwise fashion without forgetting the previously learnt logical combinations. Furthermore, this algorithm works very quickly and leads to a parallel implementation for large databases.