What is the business case for building an expert system?
Let’s start with the costs.
Someone has to create the knowlege base for the expert system. Knowledge engineering can be slow, tedious, complicated and expensive.
The expert system will require the technology to serve as the platform. Someone will have to provide the ongoing support for that technology.
The knowledge base will also have to be continually updated based on user feedback and changes in the domain.
In some cases, there will be opposition or reluctance among human subject matter experts to commonditize their expertise in a computer-based system. Other human experts may argue it’s inappropriate to trust a computer to take on this role out of a fear that something might go wrong.
Now let’s look at the benefits.
Volume: Finding an area with a high volume of non-expert users or transactions who will benefit from an expert system will justify the costs. Once the system is created, it can serve many users all at once, 24 hours a day, 7 days a week.
Services for Non-Experts: An expert system can emulate the support and guidance a human expert would provide for non-expert users who require support in that domain. It’s especially effective when there is a scarcity of human experts or resources for non-expert users to pay for human experts in that domain. Users should be able to access the expert system any time from any internet-connected device.
Knowledge Management: Once the expert knowledge is built into the knowledge base, it will become a type of information management and retrieval system for that knowledge. It’s possible to capture knowledge from multiple experts from multiple disciplines in a single knowledge base.
Standardization: The content in a knowledge base can be vetted, reviewed and amended by any number of human subject matter experts. After that happens, the expert system drawing on that knowledge will tell every user the same thing every time in the same order. There will be a record of what it told the user as well.
Data & Analytics: The expert system should have the ability to capture every users’s journey and through the knowledge base. System administrators will see what types of problems people have and in what volume.
If knowledge engineering processes and expert system technologies improve, the costs can go down, while benefits increase.