KNOWLEDGE ENGINEERING
The work done on this assignment must be your own work. Think carefully about
any problems you come across, and try to solve them yourself before you ask
anyone else for help. Under no circumstances should you work together with
another student on any code used in this assignment. Any code you reuse from
the CLIPS sample code MUST be referenced.
Assignment
Due: Friday 25th October 2019 11.59 pm
Worth: 10% of total CS.367 marks
Aim of the assignment
This assignment is intended for you to model knowledge using an intermediate
knowledge level representation (decision trees) and gain knowledge of a
declarative rule-based programming language (CLIPS) commonly used for
business intelligence applications. Additionally, this assignment is aimed at
solving problems declaratively rather than algorithmically.
You should create a small decision tree for a simple diagnostic task of your own
choice. Simple computer diagnostics such as: “I can’t connect to the Internet” or
“I can’t print” are very suitable tasks. As a guide your decision tree should
contain approximately about 10 decision nodes. Your decision tree may contain
more nodes but will not earn extra marks
If you cannot develop a decision tree of your own you may use the decision tree
below that diagnoses why a car will not start (note: you will loose 2 marks for
not creating your own decision tree).
Using your decision tree as your knowledge-level model create a simple rulebased
expert system in CLIPS to implement your diagnostic system. CLIPS and
its documentation can be downloaded from the AI Resources section of the 367
Canvas website and from the CLIPS website.
COMPSCI 367 - Assignment 01 (p2 of 3)
Input and output should be via the CLIPS console only (i.e., no marks will be
given for creating an interface in Python or Java).
MAKE SURE YOUR CLIPS CODE WORKS CORRECTLY!
Submit your decision tree (as a pdf) and CLIPS code in a single .zip
file via the Computer Science Assignment Dropbox before the
deadline.
https://adb.auckland.ac.nz/
COMPSCI 367 - Assignment 01 (p3 of 3)
Marking Guidelines (marked out of 10)
1. Specify problem domain (0.5 mark)
Can the marker tell what problem you're solving; without having to step
through your decision tree. This could either be a comment on the decision
tree diagram, or as a comment in the code, or in of banner displayed when
running your code (or preferably all of the above).
2. Created your own decision tree (2 marks) if you did not just use the
example tree given above.
3. Size of decision tree is adequate (2 marks)
The briefing specified about 10 nodes, and gave an example with 8. So
you’ll loose 0.5 marks for every node less than 8 to a maximum of -2
marks.
4. Do something "extra" (0.5 mark)
If your decision tree is actually a graph (but without infinite loops),
If your decision tree doesn’t just have binary choices.
5. Your CLIPS expert system accurately reflects your decision tree diagram (5
marks).
Is every part of the knowledge base reachable? Do the tree diagram and
knowledge-base say (essentially) the same thing?
Failure to reference any sources that helped your code (-4 marks)
Did you just take/use an example CLIPS code, and change it - without saying so.
Comments in the code such: "Adapted from auto.clp", "Parts copied from
stove.clp", "Modified from examples given" would all be sufficient, it just needs to
be clear.
No one should have to tell you to do this; refer to: