Expert systems: a practical response to complexity

On the boundary separating the objectivity of logic and the subjectivity of human experience, expert systems play with algorithms andr put artificial intelligence, fuzzy logic, probabilities and inferences to work. For critical decisions or data evaluation, expert systems even involve humans.

Decision support systems, automatic management systems, BRDM (Business Rules and Decision Management Systems), intelligent agents and many other names depending on the specialization, expert systems help manage complexity and aggregate thousands of complementary expertises into a whole that can be used by a person who has not been exposed to the full range of possible situations and yet has to face them and make the right decisions.

An expert system in response

To give an idea of the complexity and usefulness of an expert system, let's take an example that almost all of us face: email management. For a small volume, we apply simple sorting criteria: personal or business? Urgent or not? Requiring a lot of attention or little? Involving money or not? etc.

But none of these criteria are absolute and each may depend on the context: whether you have a lot of time available or not, whether you're in the office, on the road or at home, whether you're under deadlines, whether the kids are in daycare, how your file system is organized, etc. It won't be the same answer depending on the time of day or the context.

Now try to imagine that not hundreds but tens of thousands of emails are pouring in and the temperature outside or the trend of the stock market indexes affects the queries and expected responses. You begin to get an idea of how one can program and interact with an expert system and also how it can help us optimize and elevate the quality of responses to situations, however varied they may be.

For example, we can program everything we need to do to maintain a certification or to satisfy a specific customer. In education, it may be pedagogical principles coupled with student data in real time or near real time, admission requirements by discipline, etc. In research these are tools valued for their semantic intelligence capabilities.

Data vacuum

The expert system compiles data, rules, and contexts and produces inferences inferred or induced. It thinks in terms of "if... therefore" to which we also add "but if", "a little more" or "a lot more", fast or slow variation, etc.. The expert system can take into account a vast amount of data and thus improve decision making.

New data, new situations are added and it is constantly improving. The most sophisticated ones learn by themselves from the results obtained. By allowing the transmission of knowledge between old and new employees, It also contributes to the stability of companies and the quality of production.

There are dozens of them, of which you will find the directory below, many open source, many business oriented. Some are good tools to start systematizing your processes.

Illustration: Desiree Walstra - ShutterStock

For more

Artificial Intelligence Tutorial - Tutorials Point
http://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_expert_systems.htm

Expert Systems Design Workshop
http://www.infres.enst.fr/~danzart/ACSE/doc/AideSE.html

Expert Systems Directory - Thot Cursus
http://cursus.edu/institutions-formations-ressources/formation/26910