Take the same knowledge base as before. Assume now that we start with an
empty STM. We add the fact Purpose = serious
. Does this trigger
any rules? Well, it matches the conditions of 2 and 3. However, there's
nothing to match the rest of those conditions, so neither of the rules
can be triggered. Unlike the last example, forward chaining will not
follow the rules backwards to try satisfying conditions; it just tests
the data already there.
Now add another fact to STM, /verb/Lessons = 178/. This matches the
condition of rules 6 and 7, but neither of them can be triggered since
there's nothing else to match the Lessons
part.
Now add a third fact, Pressups = 5
. This triggers rule 8, which
adds a fourth fact, Fitness = poor
. This in turn triggers rule 6
(both parts of whose conditions now match STM). Rule 6 adds
Rating = beginner
. And that triggers rule 2, which can now add
a value for Resort
.