http://www.cnr.it/ontology/cnr/individuo/prodotto/ID14334

A Concurrent Neural Classifier for HTML Documents Retrieval (Articolo in rivista)

Type

Label

A Concurrent Neural Classifier for HTML Documents Retrieval (Articolo in rivista) (literal)

Anno

Alternative label

Conti Vincenzo, Pilato Giovanni, Sorbello Filippo, Vassallo Giorgio, Vitabile Salvatore (2003)
A Concurrent Neural Classifier for HTML Documents Retrieval
in Lecture notes in computer science
(literal)

Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#autori

Conti Vincenzo, Pilato Giovanni, Sorbello Filippo, Vassallo Giorgio, Vitabile Salvatore (literal)

Pagina inizio

Pagina fine

Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#numeroVolume

Rivista

Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#descrizioneSinteticaDelProdotto

With the recent increasing of digital libraries it has grown the necessity of designing automatic systems for information retrieval. In this paper a system, based on both neural networks and multi-agent paradigm, for automatic concurrent retrieval of HTML pages, has been presented. The system is composed by three agents: the Query Agent, the Locator Agent and the EaNet Classifier Mobile Agent. The proposed system has been implemented exploiting the features and facilities of the Jade platform, a multi-agent FIPA specification compliant platform. At system start-up, an autonomous process for EaNet neural network training starts. The trained neural network is successively embedded into the EaNet Neural Classifier Mobile Agent. The user interacts, through the Query Agent, with the system in order to retrieve documents satisfying a query and belonging to a given class. The EaNet Neural Classifier Mobile Agent receives the request and clones itself in the document repositories registered in the platform. Each clone interacts with the Locator Agent, present in each repository, in order to find the location of documents to be classified. After the classification task, each clone sends the results to the Query agent that shows them to the user. (literal)

Note

Http://www.cnr.it/ontology/cnr/pubblicazioni.owl#affiliazioni

1- ICAR-CNR; 2- Centro di Ricerche Elettroniche in Sicilia; 3- Dipartimento di ingegneria Informatica, University of Palermo (literal)

Titolo

Abstract

A neural based multi-agent system for automatic HTML pages retrieval is presented. The system is based on the E-alpha-Net architecture, a neural network able to learn the activation function of its hidden units and having good generalization capabilities. The starting hypothesis is that the HTML pages are stored in networked repositories. The system goal is to retrieve documents satisfying a user query and belonging to a given class (i.e. documents containing the word football and talking about Sports). The system has been implemented using the Jade platform and it is composed by three agents: the E-alpha-Net Neural Classifier Agent, the Query Agent, and the Locator Agent. The system is very efficient: the preliminary experimental results show that in the best case a classification error of 9.98% is obtained. (literal)

Prodotto di