Expert System Builder (Standard PC Based Edition)
The standard system, currently ESB4.3, is PC based and provides the means to develop a knowledge based system that can be operated on a Windows based PC (Windows 95, 98, Me, 2000 and XP). The system comprises three programs namely the ESB Question Editor, ESB Knowledge Acquisition Program and ESB User Interface. Each of these programs is described below.

Question Editor
The ESB Question Editor (ESB QE) (images here) is the program used to start developing the system. The ESB QE is used to:

• Builds the bank of questions, and the series of associated responses, upon which the complete system is developed.
• Question importance is assigned.
• Question dependencies are specified such that certain questions are only asked if previous responses dictate that they need to be asked.
• Questions can be expanded upon (i.e. Help added) by linking the question to an HTML page giving extra information (via a URL). This can be on the local machine or on the internet.

Knowledge Acquisition
The Knowledge Acquisition program (ESB KA) (images here) is used to

• Populate the knowledge base of the system with possible solutions (results).
• The optimal user response for each possible solution is entered at this stage. This allows ESB to calculate the probability of this solution being the right one given the users responses.
• Each system solution (or result) can be linked to an HTML page, via a URL, so that a simple click on the result in ESB can take the user to the information they require.
• The system developer would usually be aided by a domain expert during this phase.

User Interface
The User Interface program is the part of ESB that the end user sees and uses. It performs the following:

• Collates the information entered in the above two programs and presents the user with a set of questions to answer.
• Determines which questions to ask based on the users current responses and the systems question dependencies.
• Uses the responses given by the user, the knowledge base of information entered with the knowledge acquisition program and its own 'fuzzy logic' inference engine to determine the solution(s) that best suit the data entered by the user.
• Presents the list of solutions in an ordered list with the most probable at the top. This entries in this list are each assigned a confidence factor (0% to 100%)