Ben Bartlett's Project Log

September 21, 2000
==================

    Alright, so it's high time I got this log up and running. So far, all
I have really done is read several books, two on AI and one on expert
systems in particular. Oh yes, I perhaps ought to mention what it is
exactly I am doing with this project. Let's see...

Project Goals:
1. To examine the theoretical aspects of expert systems, with
   a focus on the mechanics of the system.
2. To write an expert shell.

    So, anyway, as I was writing, I have pretty much just been reading up
until this point. I give an outline on next Thursday, and so I need to
meet with Jim tomorrow.

Addendum:
    Wrote up the outline for my first presentation. I'll have to flesh it
out this weekend.

September 22, 2000
==================
    I met with Jim and had my outline for my presentation approved.

September 23, 2000
==================
    I've been reading Introduction to Expert Systems, by Steve Jackson.

September 24, 2000
==================
    Continued reading Introduction to Expert Systems.

October 4, 2000
===============
    Ok, so I haven't done a very good job of keeping up with this
log. I've continued doing reading, but have decided that if I keep using
the strategy I am using now (just reading everything), I won't build the
knowledge I need to begin writing the expert system fast enough. Thus, I
intend to start a new strategy: since I am planning on writing a
rule-based shell, I am going to read strictly on the rule-based system at
first, doing in-depth reading. Then I am going to do more shallow reading
of the other material in order to write the research paper. I have decided
to use a knowledge base about simple problems with a computer (why is the
printer not working, why can't I load this file, etc, etc). I believe that
this will make a good knowledge base because it fits into the rule-based
system well, and because I know something about the subject.
    Just remembered that I have a book of AI Case Studies in Lisp. Looked
in it, and lo and behold, one of the case studies is that of an expert
system. Now I remember why I keep a library of this sort of thing. This
gives me a good idea of how to go about the practical design of the
system, in addition the the more theoretical sources I was already
using. Now I just need to get my knowledge base in line and I am on my
way.

October 7, 2000
===============
    Today I began reading the case study of a knowledge system implemented
in Lisp. Very insightful. Gives me a good idea where to start the
implementation of my project.

October 8. 2000
===============
    Continued reading the case study. This project will obviously require
quite a bit of effort. I plan to start on the design over midterm break.