Incremental Iterative Retrieval and Browsingfor Efficient Conversational CBR Systems

  • Authors:

  • Igor Jurisica;Janice Glasgow;John Mylopoulos

  • Affiliations:

  • University of Toronto, Faculty of Information Studies, 140 St. George St., Toronto, ON M5S 3G6, Canada. jurisica@fis.utoronto.ca;Queen's University, Department of Computing and Information Science, Kingston, ON K7L 3N6, Canada. janice@qucis.queensu.ca;University of Toronto, Department of Computer Science, 6 King's College Rd., Toronto, ON M5S 3H5, Canada. jm@ai.utoronto.ca

  • Venue:
  • Applied Intelligence
  • Year:
  • 2000

Quantified Score

Hi-index 0.00

Visualization

Abstract

A case base is a repository of past experiences that can beused for problem solving. Given a new problem, expressed in the formof a query, the case base is browsed in search of “similar” or“relevant” cases. Conversational case-based reasoning(CBR) systemsgenerally support user interaction during case retrieval andadaptation. Here we focus on case retrieval where users initiateproblem solving by entering a partial problem description. During aninteractive CBR session, a user may submit additional queries toprovide a “focus of attention”. These queries may be obtained byrelaxing or restricting the constraints specified for a prior query.Thus, case retrieval involves the iterative evaluation of a series ofqueries against the case base, where each query in the series isobtained by restricting or relaxing the preceding query.This paper considers alternative approaches for implementingiterative browsing in conversational CBR systems. First, we discuss a naivealgorithm, which evaluates each query independent of earlierevaluations. Second, we introduce an incremental algorithm,which reuses the results of past query evaluations to minimizethe computation required for subsequent queries. In particular,the paper proposes an efficient algorithm for case base browsingand retrieval using database techniques for incremental viewmaintenance. In addition, the paper evaluates scalability ofthe proposed algorithm using its performance model.The model is created using algorithmic complexity and experimentalevaluation of the system performance.