19-04-2014, 03:36 PM
CASE –BASED REASONING SYSTEMS
CASE –BASED REASONING .ppt (Size: 438.5 KB / Downloads: 14)
CASE –BASED DECISION SUPPORT
It assists companies in the following ways:
Help companies learn from previous experiences
Support customers in both pre- and after-sales activities.
CASE-BASED DECISION SUPPORT, CONT’D
It is strategic in achieving collaboration for customer relationships by the following:
Even after the decision maker has left the company, case knowledge is preserved
Employees of the company can share their experience as documented cases
HISTORY OF CBR
In 1977, Roger Schank introduced this concept.
He explored how memory of previous situations help in both problem solving and learning.
EXAMPLE
Doctors may use CBR systems in diagnosing a patient.
When a patient comes in, the doctor records their symptoms.
In a CBR system, past cases will be searched with similar symptoms.
REUSE
Here, we are finding the difference between the new and the old case by the following:
Copying: the solution is simply copied from the old case.
Adapting: transforming or reusing the old solution.
RETAIN
Incorporate the lesson learned from the problem -solving experience into the existing knowledge by:
Extracting: if the problem was solved using an old case, the system can build a new case or generalize an old case.
Indexing: decide what types of indexes to use for future retrieval.
Integrate: modify the indexing of existing cases after the experience.
Advantages
Decrease in response time
Allows users to take preventative measures
Greater quality of solutions
Increased amount of user acceptance
Singles out important parts of the problem
Disadvantages
Requirement for cases
Reasoner might be tempted to use cases blindly
Reasoner might allow cases to bias them
People are not always reminded of the most appropriate case