Challenges during Knowledge Acquisition in KBS (Knowledge Based Systems)

Challenges during Knowledge Acquisition in KBS (Knowledge Based Systems)

 

There are many challenges that Knowledge Engineer face during Knowledge Acquisition process, a crucial process in Knowledge Engineering. To understand these challenges, we first have to understand the stages/phases/steps of Knowledge Acquisition:

Knowledge Acquisition Process

Knowledge Acquisition process is a very crucial process in the development of Knowledge Based Systems (KBS). It involves 5 major staeges/ steps/phases:

• • Identification
    
•  
  • Conceptualization
    
•  
  • Formalization
    
•  
  • Implementation
•  
  • Testing

 

Challenges

Complexity of Knowledge

Knowledge possessed by experts is complex as most of it is tacit, and hence cannot be captured without converting tacit knowledge to explicit knowledge (externalization), which is very difficult. The only way to represent tacit knowledge is by representing only the most relevant and important aspects of a specific expert domain.

 

Experts’ loyalty to their knowledge 

Experts value their knowledge as it is the main source of their command or power in their field. This makes them reluctant to share their knowledge with other experts in their own field as well as knowledge engineers. Some experts may fail to provide full details, so that they may remain competitive in their field, while others may fear being replaced by expert systems.

 

Subjective Knowledge 

When acquiring knowledge from different experts, they may provide highly arguable and subjective knowledge. It is difficult to map this inter-subjectivity, to recognize mutual understanding as well as conflicts, to come up with subjectivity-free domain-specific knowledge.

 

Expert availability 

Experts always have a tight schedule, with their expertise being critically required by their clients, organizations and professional bodies. This makes it hard for knowledge engineers to get valuable time with the experts, which means that the knowledge engineers have to be very accurate when asking questions as well as representing the solutions, to make the most out of that time. Unavailability of an expert may force the knowledge engineer to employ the protocol analysis technique (Being present as the experts think out loud as they perform their tasks, and recording the process for later analysis).

 

Rule development 

Knowledge engineers need to create rules to make it easy to characterize knowledge during knowledge acquisition. This might be difficult especially when the knowledge engineer is drafting a rule base, stable enough to be used to gather feedback in structured knowledge acquisition techniques.

 

Challenges during Knowledge Acquisition in KBS (Knowledge Based Systems)
Artificial Intelligence | thetqweb