Understanding Expert Systems

4 min read

This blog reviews the first principles and definitions of the components constituting Expert Systems.

Expert Systems

Part of Reactive AI, Expert Systems execute specific processes in a repeatable way. This execution is based on a configuration, meaning to reproduces the decision a human expert would make following the acquisition of data points from the user.

The combination of rules and logic creates the knowledge of the Expert System.

In business, Expert Systems have the unique advantage of requiring no training but a configuration based on existing documentation as well as 100% accuracy in outcomes delivered which is not a small feat for an AI-based solution.

Components of Expert Systems

Below is a helpful infographic we found from the ETH, the leading technical university of Zurich in Switzerland.

Let's go over each component starting from the end:

• User: the User is the individual using the Expert System. By interacting with the Expert System, the User provides information such as personal information or information related to the product. 
• Fact base: the information collected from the user creates a Fact base, which is used by the rules to make decisions and create new information to be shared with the User
• Rule base: the business rules make the rule base, which defines the perimeter of action or decision of the Expert System
• Programmer: this the individual designing and building the Expert System. One of our roles at Spixii is to make the role of programmer accessible to business analyst with limited or zero programming skills thanks to a no/low code interface of the Spixii platform.
• Expert: the business individuals who create and maintain the rules which make the initial knowledge inputted in the Expert System. Customer service operational experts usually fill this role.
• Inference engine: programming logic applying the rules from the Rule base and the information from the Fact base to derive decisions and new information to provide the User. For example, in the business world, the Inference engine will tell a user if an application for a financial product has been accepted.

Making the human expert better

There is an arrow linking the User and the Expert. For us, this arrow exists to constantly improve the Expert thanks to unique data based on the user's interactions with the Expert System. One of the unique features of the Spixii platform is recording conversational paths, which are the discreet interactions of users with their conversational self-service. Data and insights are crucial sources of information in enriching the knowledge of human experts.

What's next?

For more information on how we use Expert Systems in the Spixii conversational process automation and how it can help your customer service operations to save money, you can download a copy of the most recent Spixii white paper ⬇⬇⬇