Avoiding Long and Fruitless Dialogues in Critiquing

  • David McSherry
  • David W. Aha
Conference paper


An important issue in critiquing approaches to product recommendation is how to avoid long and fruitless critiquing dialogues when none of the available products are acceptable to the user. We present a new approach called progressive critiquing in which the non-existence of a product that satisfies all the user’s critiques triggers an explanation alerting the user to the possibility that none of the available products may be acceptable. A recovery mechanism based on implicit relaxation of constraints ensures that progress can again be made if the user is willing to compromise. Our empirical results show that progressive critiquing is most effective when users give priority to critiques on attributes whose values they are least inclined to accept.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bridge, D., Ferguson, A.: An Expressive Query Language for Product Recommender Systems. Artificial Intelligence Review 18 (2002) 269–307CrossRefGoogle Scholar
  2. 2.
    Burke, R.: Interactive Critiquing for Catalog Navigation in E-Commerce. Artificial Intelligence Review, 18 (2002) 245–267CrossRefGoogle Scholar
  3. 3.
    Burke, R., Hammond, K.J., Young, B.: The FindMe Approach to Assisted Browsing. IEEE Expert 12 (1997) 32–40CrossRefGoogle Scholar
  4. 4.
    Hammond, K.J., Burke, R., Schmitt, K.: A Case-Based Approach to Knowledge Navigation. In: Leake, D.B. (ed.) Case-Based Reasoning: Experiences, Lessons & Future Directions. A A AI Press/MIT Press, Menlo Park, CA (1996) 125–136Google Scholar
  5. 5.
    Hurley, G., Wilson, D.C.: DubLet: An Online CBR System for Rental Property Recommendation. In: Aha, D.W., Watson, I. (eds.) Case-Based Reasoning Research and Development. LNAI, Vol. 2080. Springer, Berlin (2001) 660–674CrossRefGoogle Scholar
  6. 6.
    Linden, G., Hanks, S., Lesh, N.: Interactive Assessment of User Preference Models: The Automated Travel Assistant. Proceedings of the 6th International Conference on User Modeling (1997) 67–78Google Scholar
  7. 7.
    McCarthy, K., Reilly, J., McGinty, L., Smyth, B.: Experiments in Dynamic Critiquing. Proceedings of the 10th International Conference on Intelligent User Interfaces (2005) 175–182Google Scholar
  8. 8.
    McSherry, D.: Similarity and Compromise. In: Ashley, K.D., Bridge, D.G. (eds.) Case-Based Reasoning Research and Development. LNAI, Vol. 2689. Springer, Berlin (2003) 291–305CrossRefGoogle Scholar
  9. 9.
    Reilly, J., McCarthy, K., McGinty, L., Smyth, B.: Incremental Critiquing. Knowledge-Based Systems 18 (2005) 143–151CrossRefGoogle Scholar
  10. 10.
    Salamó, M., Reilly, J., McGinty, L., Smyth, B.: Improving Incremental Critiquing. Proceedings of the 16th Irish Conference on Artificial Intelligence and Cognitive Science (2005) 379–388Google Scholar
  11. 11.
    Shimazu, H.: ExpertClerk: A Conversational Case-Based Reasoning Tool for Developing Salesclerk Agents in E-Commerce Webshops. Artificial Intelligence Review 18 (2002) 223–244CrossRefGoogle Scholar
  12. 12.
    McSherry, D.: On the Role of Default Preferences in Compromise-Driven Retrieval. Proceedings of the 10th UK Workshop on Case-Based Reasoning (2005) 11–19Google Scholar
  13. 13.
    McSherry, D.: Incremental Nearest Neighbour with Default Preferences. Proceedings of the 16th Irish Conference on Artificial Intelligence and Cognitive Science (2005) 9–18Google Scholar
  14. 14.
    McGinty, L., Smyth, B.: Comparison-Based Recommendation. In: Craw, S., Preece, A. (eds.) Advances in Case-Based Reasoning. LNAI, Vol. 2416. Springer, Berlin (2002) 575–589CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London Limited 2007

Authors and Affiliations

  • David McSherry
    • 1
  • David W. Aha
    • 2
  1. 1.School of Computing and Information EngineeringUniversity of UlsterColeraineNorthern Ireland
  2. 2.Navy Center for Applied Research in Artificial IntelligenceNaval Research LaboratoryWashington DCUSA

Personalised recommendations