Maintaining Unstructured Case Bases

With the dramatic proliferation of case based reasoning sys tems in commercial applications many case bases are now becoming legacy systems They represent a signi cant portion of an organization s assets but they are large and di cult to maintain One of the contribut ing factors is that these case bases are often large and yet unstructured they are represented in natural language text Adding to the complexity is the fact that the case bases are often authored and updated by di er ent people from a variety of knowledge sources making it highly likely for a case base to contain redundant and inconsistent knowledge In this paper we present methods and a system for maintaining large and unstructured case bases We focus on two di cult problems in case base maintenance redundancy and inconsistency detection These two problems are particularly pervasive when one deals with an unstructured case base We will discuss both algorithms and a system for solving these problems As the ability to contain the knowledge acquisition problem is of paramount importance our methods allow one to express relevant domain expertise for detecting both redundancy and inconsistency nat urally and e ortlessly Empirical evaluations of the system prove the e ectiveness of the methods in several large domains