Advertisement
Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- 4.4.1 Navigation Support
- The navigation menu is used to navigate through the concepts under consideration. To help the student browse the
- materials, BITS marks each concept with an appropriate traffic light. These traffic lights are computed dynamically from the Bayesian network and indicate the student’s knowledge regarding these topics.
- In BITS, each concept is marked as belonging to one of
- the following three categories:
- (i) already known,
- (ii) ready to learn, and
- (iii) not ready to learn.
- A concept is considered already known if the Bayesian network indicates the probability p(concept = known|evidence) is greater than or equal to 0.70, where the evidence is the student’s knowledge on previous concepts obtained indirectly from quiz results or directly by the student replying to a query from BITS. It should be noted
- that the choice of 0.70 to indicate a concept is known is arbitrarily chosen.
- If a concept is marked ready to learn, it means that all prerequisites are known by the student. In terms
- of the Bayesian network, the probability p(concept = known|evidence) is less than 0.70 for the current knowledge concept and all of the parent concepts are already known. Please note that knowing the parent concepts does not mean the child concept is already known. On the contrary, it only means that the child concept is ready
- to learn, provided the current probability p(concept = known|evidence) < 0.7. The second point we need to note
- is that we only need be concerned with whether the parent concepts are already known. According to the conditional
- independencies encoded in the Bayesian network, given the parents of one variable v i , v i is conditionally independent
- of all non-descendants.
- Finally, a concept is labelled not ready to learn if at least one of the prerequisites for this concept is not already
- known by the student; in terms of the Bayesian network, there exists at least one parent concept that is not already
- known.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement