A Connexionist Model for Information Retrieval

  • M. Boughanem
  • C. Soulé-Dupuy


This paper describes a connectionist architecture for an information retrieval system based on neural networks. This approach allows to define a “dynamic” thesaurus, in order to improve the construction of a documentary base and to perform associative information retrieval. We suggest a set of rules to activate cells in order to start an activation/propagation process on which the associative information retrieval is based. A learning mechanism is also started. These two notions allow to develop automatic reformulations of queries and dynamic restructuration of the information base.


Information Retrieval Information Retrieval System Document Descriptor Query Reformulation Term Network 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    W. B. Croft, H. R. Turtle. “Efficient probabilistic inference for text retrieval”. Proceedings of RIAO’91 on Intelligent Text and Image Handling. Barcelone (Spain). Apr. 1991.Google Scholar
  2. [2]
    D. Desbois, J. Savoy. “Bayesian inference networks in hypertext”. Proceedings of RIAO’91 on Intelligent Text and Image Handling. Barcelone (Spain). Apr. 1991.Google Scholar
  3. [3]
    G. Salton. “On the use of spreading activation in automatic information retrieval”, communications of the ACM, Vol 31, N°2, Feb. 88.Google Scholar
  4. [4]
    L.B. Doyle. “Information retrieval and processing”. Melville Publishing compagny, Los Angeles. CA, 1975.Google Scholar
  5. [5]
    M. Boughanem, G. Mothe, D. Puget, C. Soulé-dupuy. “Optimization of knowledge representation in information retrieval”. Proceedings of NEURO-NIMES’91. Nimes (France). Nov. 1991.Google Scholar
  6. [6]
    G. Van slype. “Les langages d’indexation: conception, construction et utilisation dans les systèmes documentaire”. Les Editions d’Organisation, collection Systèmes d’Informations et de Documentation, Paris, 1987.Google Scholar
  7. [7]
    G. Salton. “Introduction to Modem Information Retrieval”. McGRAW HILL International Book Company (New York) 1983.Google Scholar
  8. [8]
    R. J. Brachman, D. L. Mcguiness. “Knowledge representation, connectionism and conceptual retrieval”. Proceedings of SIGIR’88, June 13–15 1988.Google Scholar
  9. [9]
    N. Kimoto, T Iwadera. “Construction of Dynamic Thesaurus and its Use for Associated Information”. Proceedings of SIGIR’90, 5–7 September 1990.Google Scholar
  10. [10]
    C. Davalo, P. Nairn. “Les réseau de neurones”. Editions Eyrolles. 1989.Google Scholar
  11. [11]
    M. Wettler R. Rapp. “Parallel associative process in information retrieval”. Parallel Processing in Neural Systems and Computers. Elsevier Science Publishers B. V. (North-Holland), 1990.Google Scholar
  12. [12]
    J. B. Crampes. “Aide à l’interrogation d’un dictionnaire de données”. RAIRO Informatique-Computer Science, Vol. 14 n°1, pp 86-95.Google Scholar
  13. [13]
    C. Soulé-dupuy. “Systèmes de Recherche d’Informations: Mécanismes d’indexation et d’interrogation”. PhD Thesis, Department of Computer Science, University of Toulouse III (France), Feb. 1990.Google Scholar
  14. [14]
    G. Salton. “Automatic Information Organization and Retrieval”. McGRAW-HILL International Book Company (New York), 1968.Google Scholar
  15. [15]
    B. Kosko. “Neural networks and fuzzy systems A dynamical systems approach to machine intelligence”. Prentice-Hall International 1991.Google Scholar
  16. [16]
    M. Tuffery. “Système documentaire, Bases de données textuelle: le projet ETOILE”. PhD Thesis, Department of Computer Science, University of Toulouse III (France), June 1984.Google Scholar
  17. [17]
    D. O. Hebb. “The organization of behaviour” J. Wiley&Sons (New-York). 1949.Google Scholar

Copyright information

© Springer-Verlag/Wien 1992

Authors and Affiliations

  • M. Boughanem
    • 1
  • C. Soulé-Dupuy
    • 1
  1. 1.IRIT/SIG — Université Toulouse IIIToulouse CedexFrance

Personalised recommendations