Knowledge Structure Manager

What is KSM?

KSM (Knowledge Structure Manager) is a software environment that helps developers and end-users in the development and maintenance of large and complex knowledge-based applications.

It covers different steps of the life-cycle of an application:

ANALYSIS:

• It manages a particular knowledge-level paradigm, based on the knowledge unit concept to be used for modeling expertise. This make easier the knowledge acquisition and the analysis of the application. The developer uses this paradigm to create a high level model to be accepted by the end user before starting the implementation.

• In addition, abstract knowledge models may be defined in such a way that they may be reused to be applied to different kind of applications. With the KSM knowledge modeling approach, the developer can define different categories of domain independent knowledge modules. This may reduce drastically the effort of defining new applications.

DESIGN & IMPLEMENTATION:

• KSM assists the developer to create the executable version of the knowledge model. In order to do so, KSM manages reusable software components which are assembled following the structure of the knowledge-level model.

MAINTENANCE

• Once the application is running, the end-user can use KSM to consult the knowledge-level model to better understand it, and he/she may change the model at this level, without programming, in such a way that KSM automatically translates those changes to the implementation level.

The ultimate goal of KSM is to reduce the gap between the human understanding and the implementation of applications using knowledge models closer the way the end-user understands its application than conventional programming languages. Those models characterize the expertise using an intuitive modular approach with different levels of abstraction to cope with large knowledge bases.

Origin of KSM

KSM was developed at the Department of Artificial Intelligence of the Politechnic University of Madrid. KSM was proposed by Jose Cuena and Martin Molina [Cuena, Molina 1993]

KSM included previous experience of knowledge engineering projects developed by the own research group about bank management decision support, river emergencies by floods, and chemical plants monitoring. Initially, in these projects was necessary to improvise high level knowledge models to cope with the complexity of the existing knowledge.

KSM was influenced by the idea of generic tasks of Chandrasekaran and the knowledge level concept of Newell. It includes also some ideas from other parallel approaches of knowledge engineering methodologies and tools such as KADS, PROTEGE-II and KREST.