What is one good for that the other's not in practice? I understand the theory of what they do, but what are their limitations and capabilities in practical use? I'm considering Drools vs a java prolog for a new AI project, but open to other suggestions. What are some popular approaches for inferencing on a complicated relational data set or alternatives?
4 Answers
In the old old old old expert systems days they used to say forward chaining was good for looking around (checking for what could be) while backward chaining was good for confirming (checking if "it" really is).
Think configuration (forward chaining, XCON [1]) and medical diagnosis (MYCIN) [2]
Forward chaining is concerned with the question "what will happen next?", while backward chaining looks at the question "why did this happen?".
An example of forward chaining is predicting whether share market status has an effect on changes in interest rates.
An example of backward chaining is the diagnosing of blood cancer in humans.
Simply put, forward chaining is mainly used for predicting future outcomes while backward chaining is mainly used for analyzing historical data.
Related