Lecture March 19. 2003

Testing our Assumptions about AI:  Part 4 – Expert Systems – 19 March 2003
 


1.  What are they and how do they work? -->What is MYCIN?
2.  How do we function as "experts"?
3.  What are the significant differences between us and "them"?
4.  Other questions Wessell raises: What are the risks and benefits associated with expert systems? (see kit)

****************
1.  Definition of an expert system
 
" a rule-based system that embodies part of the skill of a human expert in a computer program" (Forsyth, 1984; cited in Wessells, 1990).
 
General Characteristics of an Expert System:
 
•           designed to offer advice  about how to perform a PARTICULAR task
 
•           its knowledge is limited to a specific DOMAIN
 
•           most can tell the user how it arrived at a decision
 
•           typically, an expert system is easy to interact with—users can use natural language
 
•           contains rules expressed in "IF-THEN" format
 
(Wessells, in course kit)
 
MYCIN ( an early system that diagnosed and treated infectious diseases) and EON (change in terminology to “knowledge” systems
See http://www.wired.com/wired/archive/6.09/crucialtech.html?pg=7
 
à EON is different from stand-alone expert systems like MYCIN:
à it uses component-based architecture: 
if today’s expert systems are like documents in a file cabinet, if you add something on one page, the others remain unaffected.
 
“ BUT with this new kind of expert system, alter a single fact and it ripples through the system seamlessly, updating all related data.”
à “this means the database gets smarter each time a doctor revises a patient’s file.” 
(from Wired archive) 
 
 
2.  How are we "experts"? How can computers be experts? (by emulating us)
 
1.  A specific goal (problem) sets our mind  in motion.
 
2.  A vast collection of FACTS and RULES wait to be called upon to help reach the goal.
 
3.  PRUNING helps us carry out a quick efficient search for only those rules that pertain to the immediate goal.
 
4.  We make inferences from facts and rules that we know.
 
5.  We use a set of strategies for generating and testing hypotheses concerning the problem at hand.  (Wessells calls this an INFERENCE ENGINE which has two commonly used strategies -- BACKWARD and FORWARD CHAINING -- see kit)
(from Levine, Drang, and Edelson, A Comprehensive Guide to AI and Expert Systems, 1986 and Wessells)
 
++++++++
 
Facts and Rules

 
Levine et al's premise:  "What is generally considered to be "intelligence" can be broken down into a collection of "facts" and a means of utilizing these facts to reach "goals."  (p. 5)
 
(do you agree with this definition of "intelligence"?)
 
FACT 1:    A burning stove is hot.
 
RULE 1:    IF I put my hand on a burning stove,
                                    THEN it will hurt.
 
Inference Mechanism
 
We're told:
 
1.  Jim's parents are John and Mary.
 
2.  Jane's parents are John and Mary.
 
GOAL:  determine the relationship between Jim and Jane.
 
Pruning zeros in on what rule?
 
IF parents are the same, THEN children are siblings.
 
We infer Jim and Jane are siblings.
 
Now we have new knowledge.
 
 
Here's an interesting example of verification through the inference mechanism:
 
" Suppose a murder has been committed:  A person is found locked in an apartment, shot three times.  The medical examiner rules out the possibility of suicide because of the angle of the wounds (PRUNING in action), and the police immediately go to work.   The first thing they consider is who else besides the victim had a key to the apartment.  They question the landlord and several neighbours, who say the murdered person had a friend who frequently came to use the apartment.  Further investigation reveals that the couple had recently been quarrelling.
 
The police now have a suspect.  They are able to INFER from their interview that the victim's friend is probably the murderer (FORWARD CHAINING), using the data from the interview to arrive at the conclusion, but they need some concrete evidence to nail down the case.  Their best chance of nabbing the suspect is to find the murder weapon.  They obtain a search warrant for the friend's apartment and look through the belongings, but to no avail.  Finally, a detective finds a gun in a garbage can in a nearby alley.  A fingerprint check verifies that the gun indeed was handled by the friend, and a ballistics test establishes that it is the murder weapon.  Case solved.
 
By obtaining a new piece of data and seeing if it was consistent with their original conclusion, the police verified the goal of identifying the murderer.  The process of using a conclusion to look for supporting data is known as "BACKWARD CHAINING."  In this case, the conclusion is the suspect and the data is the weapon."
 
(from Levine, Drang, and Edelson, A Comprehensive Guide to AI and Expert Systems, 1986, p. 17)
 
  3.  Differences between Human Beings and "Knowledge" Systems

Human Expert   Knowledge Systems
1.  can be expert more than one fiel one domain
2.  reason using general and from heuristics analogy needs domain
heuristics
3.  learns from experiences restricted to learning from "rules" taught by human
4.  possess common sense/can act
spontaneously
 
no common sense
5.  can be biased  bias of initial rules
6.  can jump to conclusions maintain those con-clusions in face of disconfirming evidence does not jump to conclusions and
7. can avoid/misread some detail does not skip details

 

4. Questions Wessells raises:
 
Who will benefit from these developments in AI?
Who is responsible for problems/errors?
Will expert systems become more credible than humans? 
Will expert systems add to dehumanization?

This page last revised 03/19/03