Sunday, 1 December 2013

Week 13


Quality Assurance: The KM system should meet user expectations. Performance usually depends on the quality of explicit/tacit knowledge stored in the knowledge base. For the expert: quality relates to a reasoning process which produces reliable and accurate solutions within the KM system framework. For the user, quality relates to the system’s ability to work efficiently. For the knowledge developer, quality relates to how well the knowledge source is and how well the user's expectations are codified into the knowledge base.

Knowledge Testing: It is required to control performance, efficiency, and quality of the knowledge base.

Types of testing:

  • Logical Testing:
To make sure that the system produces correct results.
  • User Acceptance Testing:
It follows logical testing and checks the system's behavior in a realistic environment.

 Issues:

  • Subjective nature of knowledge (tacit)
  • Lack of reliable specifications
  • Verifying correctness/consistency
  • Negligence in case of testing
  • Time limitations for knowledge developers to test the system
  • Complexity in case of user interfaces
Two approaches:
  • Verify the knowledge base formation:
    • The structure of the knowledge as it relates to circular or redundant errors is verified.
    • Consistency, correctness and completeness of knowledge base rules are also verified.
  • Verify the knowledge base functionality:
Deals with confidence and reliability of the knowledge base.
Attributes:
    • Circular Errors
    • Completeness
    • Confidence
    • Correctness
    • Consistency
    • Inconsistency
    • Redundancy Errors
    • Reliability
    • Subsumption error
Steps:
  • Selecting a person/team for testing.
  • Deciding on user acceptance test criteria.
  • Developing a set of test cases.
  • Maintaining a log on different versions of the tests and test results.
  • Field-testing the system.

Test Team/Plan:

A testing plan indicates who is to do the testing. Commitment initiates with management support and a test team with a test plan. The team is expected to
  • be independent of the design/codification of the system
  • understand systems technology/knowledge base infrastructure
  • be well versed in the organization's business
  • Deciding on user acceptance test criteria:
    • Accuracy
    • Adaptability
    • Adequacy
    • Appeal
    • Availability
    • Ease of use
    • Performance
    • Face validity
    • Robustness
    • Reliability
    • Operational/Technical Test
  • User Acceptance Test Techniques:
    • Face Validation
    • Developing a set of test cases
    • Subsystem Validation
    • Maintaining a log on different versions of the tests/test results
    • Field testing the system

Kulsum Raza

No comments:

Post a Comment