Table of Contents
Knowledge engineering is the process of solving complex problems. The process uses algorithms to mimic the thought process and patterns of a human subject matter expert. He treats questions and problems like a human expert and makes decisions the way human decisions remain made.
Knowledge Engineering is an emerging discipline that tries to emulate the human mind to solve specific problems. We will explain it to you in-depth.
Knowledge Engineering is the discipline that creates Expert Systems. It is, therefore, a fundamental piece within the field of Artificial Intelligence. The specialists from Strategy Big Data, the technological partner of B12, explain what Knowledge Engineering is, what it is related to Artificial Intelligence, and its practical application in the business field.
Knowledge Engineering And Artificial Intelligence
Artificial Intelligence is dedicate to studying and developing technologies that allow machines to think for themselves . It is a simplified definition. Specialists want these machines to be capable of learning, reasoning, evolving and, therefore, self-correcting and perfecting themselves, just as a human being would. It is an incredibly wide field that brings together many disciplines, including Knowledge Engineering.
Definition Of Knowledge Engineering
Knowledge Engineering is the branch of Artificial Intelligence dedicated to studying, designing and developing Expert Systems. We are talking about software created to solve problems by emulating the mind of a person who is an expert in a specific field. For this, they learn business rules and techniques such as decision trees to emulate the knowledge of an expert in a field.
What are Expert Systems
In Artificial Intelligence, Expert Systems are software, computer programs dedicated to solving problems as a human being would. The addition, in this case, is that the human being whose mind it emulated would be an expert in a specific subject. That is, he would have specialized knowledge in it. Hence its name.
Explained: imagine that you have a problem related to a specific field, for example, a math problem. The person you are consulting is a mathematician, an expert who has prior knowledge to apply to reason, hypothesize and calculate, and ultimately solve the problem.
An Expert System does the same thing: the software is programmed with initial specialized knowledge of a subject and uses that knowledge when a problem is present.
Therefore, an Expert System is basically divided into:
- Knowledge base: the initial knowledge with which the Expert System works. Semantic networks, practical cases, rules, predicates. All the initial information on a specific subject.
- Inference engine: the engine that gives you the ability to reason, the one that leads you to select and combine data, contrast the facts, make hypotheses. In short, to use the initial knowledge to find a solution based on the problem that arises.
An addition that the Expert System has is that knowledge is not static. As it finds solutions, it feeds back learns by itself so that the response in the future is more accurate and faster.
The knowledge base is, therefore, constantly updated, and this, in turn, provides more arguments to the inference engine and increases the success rate of the response.
Types Of Expert Systems
There are two fundamental types of Expert Systems, depending on what they are based on to answer the problem.
These Expert Systems apply heuristics. They use a series of initial knowledge to find the solution to the problem.
They are based on previous cases, previous examples that pose a problem already solved similar to the current one. These systems adapt the reasoning then done to the new reality, following similar steps to find the answer.
First Expert Systems
Expert Systems can be used in many fields, from business to medicine. The first Expert System that had a real and fruitful practical application became quite popular software in chemistry and biology.
We are talking about Dendral, from the team led by Edward Feigenbaum from Standford University. It is a software developed in the 1960s that interpreted molecular structures, which speed up experts’ tasks in the abovementioned fields.
In the following decade and with Dendral as a base, Edward Shortliffe and his team, also from Stanford, developed Mycin. Its software was use in medicine to diagnose infectious diseases in the blood. Based on that, Mycin also decided which medication was the most suitable for each case.
Mycin presents a very clear example of what an Expert System is: it had initial information (knowledge base), gathered data on a specific person’s symptoms, and crossed that data with its initial information (using the inference engine) exhibited a result. He then used the conclusions to broaden his knowledge base.
Contemporaneous with Mycin, other relevant Expert Systems emerged and were applied in various fields, such as Prospector (which was use to find mineral deposits), Puff (which studied lung function) or Macysma (which solve computer algebra problems).
How Knowledge Engineering Is Use In A Company
Among other things, Knowledge Engineering can develop Expert Systems that speed up the internal functioning of a company and, therefore, increase productivity. Thus, an Expert System can take care,
For Example, Of
- Internal audits.
- Task management.
- Customer Support.
- Automation of repetitive processes.
Also Read: What is Gymnastics? – types of gymnastic