Mercure02: adhoc and routing tasks

1 MercureO2 system MercureO2 is an object oriented information retrieval system based on connectionist approach. It allows the query formulation, the query evaluation based on propagation of the neuron activation and the query modiication based on backpropagation of the user judgements of the document relevance. 1.1 The model The neural information model used in MercureO2 is shown in gure 1. It represents a set of associated neurons. We distinguish in this model :-neuron objects-an input layer representing the user information needs-a term neurons layer-a document neurons layer-an output layer representing the result of the query evaluation-weighted links between the neurons. The MercureO2 neuron object is characterized by : identiier, input value, output value, output threshold value, methods allowing to compute the input values and output signals, to propagate the signal, to backpropagate the signal,... The functions of this neuron are :-Input function : In(N i) = P j=1 Out(N j) w ji-Activation (or state) of the neuron at (k+1)stage is :