AI-Portal Artificial Intelligence for Enthusiasts
- Written by Administrator
- Category: Uncategorised
What is knowledge? In our colloquial thinking, it''s something that "we know" , mostly information.
For computers, this might not be as easy as it seems. Knowledge can be split in some categories:
- Common sense knowledge: We know that if we throw a rock on a window, it will break. Can a computer actually know this if we tell him about the window and the rock? Does the computer realize that the rock will eventually fall down to the earth ? In common sense, we have a lot of knowledge that we take for granted, like physics laws, which may not be as intuitive for computers as they are for us. To give a computer a "common sense knowledge" implies a lot of factors that we may or may not be aware of. In computer decision process, minor things can be of extreme importance.
- Logic knowledge: We know that if we add one apple to another apple we obtain two apples. This is quite logical and it''s appliance can be extended for various problems and situations. This is also fair to implement in a computer knowledge database.
Computers cannot have a full set of common sense knowledge. How can we implement a basic system that can be extended so that computer share a bit of our understanding? One idea is too use a rule-based system.
If thrown(rock, window) then break(window)
Actions, predicates, simple logic can be useful in such situations.
Objects can be implemented into classes, actions can move from one state to another.
However, knowledge representation is a very important part in having a basic start for creating an artificial intelligence and designing a system that will respond fast to stimuli and give an adequate response as well.