Archive for December, 2011

Expert Systems Vs Human Expert

What is Expert Systems? Expert systems are designed to solve complex problems by reasoning about knowledge, like an expert, and not by following the procedure of a developer as is the case in conventional programming.

What is Human Expert? An expert / expert (human expert) is an individual who has a superior capability understanding of a problem. For example: a doctor, financial advisor, an expert in car engines, etc..

 Comparison of Human Expert and Expert Systems

  • Time Availability : Human Expert (Working day). Expert Systems (any time)
  • Geographical :  Human Expert (specific location). Expert Systems (wherever)
  • Security: Human Expert (not replaceable). Expert Systems (can be replaced)
  • Perishable : Human Expert (yes). Expert Systems (no)
  • Performance : Human Expert (variable). Expert Systems (consistent)
  • Speed:Human Expert (variable). Expert Systems (consistent and more fast)

The fundamental reason why the ES (Expert Systems) was developed to replace a experts:

  • Can provide expertise at all times and in various locations
  • Automatically doing routine tasks that require an expert.
  • An expert will retire or leave
  • An expert is expensive
  • Expertise is needed also in a hostile environment (hostile environtment)

Comparison of Conventional Systems and Expert System

 The Conventional Systems  andExpert System

Conventional Systems vs Expert System



Definition Of Expert Systems

What is Expert Systems? Expert Systems is a computer program designed to model ability to solve problems like a certain expert (human expert). What is human expert? An expert / expert (human expert) is an individual who
has a superior capability understanding of a problem. For example: a doctor, financial advisor, an expert in car engines, etc..
Ability of expertise:

  • Can identify (recognizing) and formulate the problem
  • Resolving problems quickly and accurately
  • Describe the solution
  • Learning from experience
  • Restructuring of knowledge
  • Determining relevance / relationship
  • Understand the limits

What is expertise? Expertise is Broad understanding of the task or the specific knowledge obtained from training, reading and experience.
Type-type of knowledge possessed of expertise:

  • Theories of problems
  • The rules and procedures that refer to the problem areas
  • The rules (heuristics) to be work will on the situation
  • The global strategy to solve various types of problems
  • Meta-knowledge (knowledge about knowledge)
  • The facts

What Is Knowledege?
Data+ processing  = Information
Information processing + (experience, training, etc.) = knowledge



Examples of Knowledge Representation Logic


example of knowledge representation logic

Logic is a form of knowledge representation of the oldest. This representation  type uses the expressions in formal logic to represent the knowledge base. Basically the logic is the process of forming conclusions and draw an inference based on the fact that already exist. Input from the process logic a premise or the facts admitted the truth so that by doing logical process of reasoning can be formed on an inference or conclusion that true as well.
Examples of the simple fact that we will represent the logic is as follows:

Helder is a dog
Facts in English they will be represented in logic, namely:
dogs (Helder)
We can also represent another logical facts, namely that all
dogs have tails
x: dog (x)
caudate (x)
Then deductively (reasoning starts from the general principles for get a more specific conclusion) of the mechanism of  this logic we can obtain a new representation:
caudate (dog)
By using the mapping function is backward, we can download the generate sentences in English
Helder caudate


Advantages Creating Knowledge Representation

the knowledge representation

advantages make knowledge representation

With representation, many things that we will get in our solve a problem. Below are some advantages that
will we get when we make representations of knowledge, namely:

  • With a good representation, create objects and relationships that are important to is clear.
  •  Representation uncovers constraints (limits) in a problem. We can reveal the influence of an object or relation to the object or other relations.
  • With our representation of objects and relations will get together.
  • We will be able to see everything we want in one time
  • We can remove all the components that are not associated with problems which we solve. Or we can hide some information we do not need for awhile, and when we we need to show again.
  • The representation will make the problems we are finished becomes transparent. We will understand the problems we solve.
  • With our representation will be able to uncover a problem in complete, so that issues can be resolved.
  • The representation will allow problems to be concise. We will concise thinking (representing what we want to represent the efficient).
  • With the representation, it will make our work becomes faster
  • With the representation, making the problems we can solve computerized. With this representation we will be able to perform procedures procedure in solving a problem.

Besides the advantages above the one thing a principle in knowledge representation is If a problem is described by
using appropriate representations, it can be ascertained that these problems can be solved.


Definition Of Knowledge Representation

knowledge representation picture

knowledge representation

Representation is intended to capture the essential properties of problems & make such information accessible to problem-solving procedure. Representation language must be able to make a programmer is able to express the knowledge needed to obtain solutions to problems.

Many ways to represent knowledge (facts) in AI programs.
There are two entities that need attention:
– Fact: the truth. The fact this is what we represent.
– Representation of facts. From this representation, we will be able to manipulate it.

Good representation, should:

  • Express it explicitly
  • Make the problem transparent
  • Complete and efficient
  • Featuring natural boundaries that exist
  • Suppress / eliminate the necessary details
  • Computation can be performed (no limit / konstraint)

In short, knowledge representation are classified into 4 categories:

  • Representations of logic: The representation of this type using the expressions in formal logic to represent knowledge base
  • Procedural Representation: Representation describes knowledge as set of instructions to solve a problem.
  • Network representation: This representation captures the knowledge as a graph where knots describe an object or concept of problems encountered, while edge describes the relationship or association between them
  • Structured Representation: Representation structured network extending to the how to make each knot into a complex data structure