Page 104 - Code & Click - 7

1. Expert Systems

In these systems, a human expert in the domain of the application creates precise rules that a
computer can follow, step by step, to decide how to respond to a given situation. These rules, known
as algorithms, are often expressed as code in an ‘if-then-else’ format. Expert Systems are also known
as Handcrafted Knowledge Systems.
The Expert System approach is closely aligned to
how human experts make decisions. Furthermore,
humans can easily understand how these systems
make specific decisions. They can easily identify
mistakes or find opportunities to improve the
program and update the code in response.
Symbolic AI works best in constrained environments which do not change much over time, where
the rules are strict, and the variables are unambiguous and quantifiable.

Knowledge Discovery Subject Enrichment

One of the most famous examples of an Expert System is Deep Blue, the IBM-developed, chess-playing AI
that defeated the human world chess champion in 1997. Deep Blue was developed in cooperation between
IBM’s software engineers and several chess grandmasters, who helped translate their human chess expertise
into tens of thousands of computer code rules for playing grandmaster-level chess.

2. Fuzzy Logic System
In the expert system described earlier, each variable or condition
is either True or False. For it to work, the system needs to know
the absolute answer to questions such as whether or not the
patient has a fever.

Fuzzy logic is another approach to expert systems which allow
variables to have a ‘truth value’ that is anywhere between 0
and 1 and captures the extent to which it fits a category.
Fuzzy logic is particularly useful for capturing intuitive Expert System Vs Fuzzy Logic
knowledge, where experts make good decisions in the face of System
wide-ranging and uncertain variables that interact with each
other. In each case, the fuzzy system continually assesses dozens of variables, follows rules designed
by human experts to adjust truth values and uses them to automatically make decisions.

Limitations of Symbolic AI Approach
Symbolic AI systems require human experts to encode their knowledge in a way the computer can
understand. This places significant constraints on their degree of autonomy. While they can perform
tasks automatically, they can only do so in the ways in which they are instructed, and they can
only be improved by direct human intervention. This makes symbolic AI less effective for complex
problems where not only the variables change in real-time, but also the rules.
It is particularly useful in supporting humans working on repetitive problems in well-defined domains
including machine control and decision support systems. The reliable performance of symbolic AI in
these domains has earned it the endearing nickname ‘good old-fashioned AI’.

102