|
Q: Do I need to use Prolog if I want to build an expert system or
a NLP system? |
|
A: No, you can use any language you want. Prolog is just more suitable
for this kind of tasks but you can build a good expert system, or NLP, in any other
language. Prolog just makes it easier. |
|
|
Q: Does it make a big difference that Prolog is declarative and not procedural? |
|
A: Yes, it does. It's a Copernican revolution: instead of telling the computer
how to do something, you tell the computer what something is and let it figure out how
to do it. It completely shifts the focus of what you are doing and how you are writing
the code. |
|
|
Q: Does Prolog have an impact on the quality of the result,
of the system created? |
|
A: The quality of the system in the end depends on the concept, idea and
design you have created. If you have a poor design Prolog won’t make it any better. It’s
just a tool, a good one, to translate AI projects into working systems. It makes the
development simpler and cleaner. |
|
|
Q: If I decide to use Prolog for my AI system, do I have to use it also
for non AI functionalities? |
|
A: No, you can write your code using a traditional, procedural,
language and then use Prolog only for the AI module or the engine that makes
the decision. |
|
|
Q: What Prolog environment or compiler would you advise? |
|
A: I would not advise any particular Prolog environment or compiler.
The choice of the environment/compiler depends on many factors, such as the platform
you want to run your project on, the portability you need, the complexity of the project
etc. However, taking aside the platform issue, I would say that most of the freely available
compilers are suitable for most projects. |
|
|
Q: Are there other languages suitable for AI applications? |
|
A: While there are many environments suitable for AI applications, both
proprietary and open source, in terms of languages the choice is very limited.
The most suitable ones remain Prolog, Lisp and Pop11. |
|
|
Q: Can you use an Expert System for a real, mission critical, application? |
|
A: Yes, there are plenty of expert systems in mission critical environments,
used for taking important decisions. For example in finance, the system can decide whether
to buy or sell a stock in a much faster and reliable way of any trader. Besides, the decision
is purely based on facts and reasoning. In this kind of decisions, emotions and other human
factors are often a recipe for disaster. |
