Knowledge Representation
19-05-2021
,Inhoudsopgave
Module 1 .......................................................................................................................................................... 3
Video Lecture 1 Preliminary concepts ................................................................................................................ 3
Quiz lecture 1................................................................................................................................................. 6
Video Lecture 2 Logical entailment .................................................................................................................... 7
Quiz lecture 2............................................................................................................................................... 15
Module 2 ........................................................................................................................................................ 16
Video lecture 3 The Prolog Language ............................................................................................................... 16
Quiz lecture 3............................................................................................................................................... 25
Video Lecture 4 Writing Prolog programs ........................................................................................................ 26
Quiz lecture 4............................................................................................................................................... 33
Video Lecture 5 Working with lists in Prolog .................................................................................................... 34
Quiz lecture 5............................................................................................................................................... 45
Video Lecture 6 Satisfying constraints in Prolog .............................................................................................. 46
Quiz lecture 6............................................................................................................................................... 54
Module 3 ........................................................................................................................................................ 55
Video Lecture 7 Sub-symbolic knowledge representation ................................................................................ 55
Quiz lecture 7............................................................................................................................................... 64
Video lecture 8 Causal cognitive modeling ....................................................................................................... 65
,Module 1
The theoretical concepts elaborate on general concepts such as artificial intelligence,
knowledge, representation, thinking, and reasoning. Many of these concepts inherently
have a philosophical connotation. We will also cover the main difference between
symbolic and sub-symbolic artificial intelligence. Towards the end of the first module, we
will discuss a procedure for thinking, which is the fancy name, we (mathematicians,
logicians and computer scientists) have given to the logical entailment procedure.
Video Lecture 1 Preliminary concepts
Artificial Intelligence
Knowledge, representation and reasoning
are the building blocks of this course.
Artificial intelligence is the study of
intelligence behavior achieved through
computational means
We need to imitate human intelligence in
order to solve complex problems that we
cannot solve.
Thinking is just processing of a big
knowledge base from which we can get new pieces of knowledge, so we can solve a certain a
problem.
Reasoning is the formal manipulation of symbols representing a collection of propositions.
Because the idea here is to produce new knowledge, so new propositions.
Artificial intelligence
Symbolic reasoning sub-symbolic reasoning
Symbolic artificial intelligence is about manipulating symbols in order to produce new
knowledge. This is called the Good Old-Fashioned Artificial Intelligence.
In sub-symbolic artificial intelligence the goal is pretty much the same, but here we are not
just manipulating symbols. We are able to manipulate more complex knowledge structures,
for example numbers or even more complex structures.
Advantages or disadvantages
Symbolic reasoning, we can understand how the manipulation goes, so we are able
to understand how a decision is made.
, This is not so simple when it comes to sub-symbolic reasoning. The main reason is
because those very powerful algorithms behave like black boxes. So they’re able to
solve the problem very efficiently, very accurately, however, we cannot always
understand how this process works.
Gottfried (von) Leibniz was a prominent German Logician
• According to Leibniz’s idea, everything can be modelled using symbols, even numerical
quantities.
• For example, the numerical quantity ‘fourteen’ is an abstract entity that can be
represented as 14, XIV or 1110.
• The point here is that we don’t operate with abstract entities instead we operate with
the symbols we use to encode those entities.
• Therefore, we should be able to manipulate those symbols in order to perform
symbolic reasoning and produce new pieces of knowledge.
By convention, we say that numbers are special symbols because we have different rules
(which are the arithmetic rules) to operate these symbols.
In the case of pure symbols, we have the rules of logic in order to manipulate them and
produce new pieces of knowledge.
Symbolic representation
Overall, we want symbolic representations* able to:
• We should represent the problem domain very accurately. This means we should be
able to express the universe, the environment or that problem domain, using precise
facts.
• We should clearly express how to obtain new knowledge.
Granularity means the level of detail to which we’re going to represent the knowledge.
What is meant by the term Granularity in a Data Warehouse?
• The level of detail available
What Determines Granularity?
- The level of data loaded into the fact table
• Per order numbers
• Daily numbers
• Weekly numbers
Proposition versus sentences
• Propositions are ideas that can be expressed through declarative sentences
19-05-2021
,Inhoudsopgave
Module 1 .......................................................................................................................................................... 3
Video Lecture 1 Preliminary concepts ................................................................................................................ 3
Quiz lecture 1................................................................................................................................................. 6
Video Lecture 2 Logical entailment .................................................................................................................... 7
Quiz lecture 2............................................................................................................................................... 15
Module 2 ........................................................................................................................................................ 16
Video lecture 3 The Prolog Language ............................................................................................................... 16
Quiz lecture 3............................................................................................................................................... 25
Video Lecture 4 Writing Prolog programs ........................................................................................................ 26
Quiz lecture 4............................................................................................................................................... 33
Video Lecture 5 Working with lists in Prolog .................................................................................................... 34
Quiz lecture 5............................................................................................................................................... 45
Video Lecture 6 Satisfying constraints in Prolog .............................................................................................. 46
Quiz lecture 6............................................................................................................................................... 54
Module 3 ........................................................................................................................................................ 55
Video Lecture 7 Sub-symbolic knowledge representation ................................................................................ 55
Quiz lecture 7............................................................................................................................................... 64
Video lecture 8 Causal cognitive modeling ....................................................................................................... 65
,Module 1
The theoretical concepts elaborate on general concepts such as artificial intelligence,
knowledge, representation, thinking, and reasoning. Many of these concepts inherently
have a philosophical connotation. We will also cover the main difference between
symbolic and sub-symbolic artificial intelligence. Towards the end of the first module, we
will discuss a procedure for thinking, which is the fancy name, we (mathematicians,
logicians and computer scientists) have given to the logical entailment procedure.
Video Lecture 1 Preliminary concepts
Artificial Intelligence
Knowledge, representation and reasoning
are the building blocks of this course.
Artificial intelligence is the study of
intelligence behavior achieved through
computational means
We need to imitate human intelligence in
order to solve complex problems that we
cannot solve.
Thinking is just processing of a big
knowledge base from which we can get new pieces of knowledge, so we can solve a certain a
problem.
Reasoning is the formal manipulation of symbols representing a collection of propositions.
Because the idea here is to produce new knowledge, so new propositions.
Artificial intelligence
Symbolic reasoning sub-symbolic reasoning
Symbolic artificial intelligence is about manipulating symbols in order to produce new
knowledge. This is called the Good Old-Fashioned Artificial Intelligence.
In sub-symbolic artificial intelligence the goal is pretty much the same, but here we are not
just manipulating symbols. We are able to manipulate more complex knowledge structures,
for example numbers or even more complex structures.
Advantages or disadvantages
Symbolic reasoning, we can understand how the manipulation goes, so we are able
to understand how a decision is made.
, This is not so simple when it comes to sub-symbolic reasoning. The main reason is
because those very powerful algorithms behave like black boxes. So they’re able to
solve the problem very efficiently, very accurately, however, we cannot always
understand how this process works.
Gottfried (von) Leibniz was a prominent German Logician
• According to Leibniz’s idea, everything can be modelled using symbols, even numerical
quantities.
• For example, the numerical quantity ‘fourteen’ is an abstract entity that can be
represented as 14, XIV or 1110.
• The point here is that we don’t operate with abstract entities instead we operate with
the symbols we use to encode those entities.
• Therefore, we should be able to manipulate those symbols in order to perform
symbolic reasoning and produce new pieces of knowledge.
By convention, we say that numbers are special symbols because we have different rules
(which are the arithmetic rules) to operate these symbols.
In the case of pure symbols, we have the rules of logic in order to manipulate them and
produce new pieces of knowledge.
Symbolic representation
Overall, we want symbolic representations* able to:
• We should represent the problem domain very accurately. This means we should be
able to express the universe, the environment or that problem domain, using precise
facts.
• We should clearly express how to obtain new knowledge.
Granularity means the level of detail to which we’re going to represent the knowledge.
What is meant by the term Granularity in a Data Warehouse?
• The level of detail available
What Determines Granularity?
- The level of data loaded into the fact table
• Per order numbers
• Daily numbers
• Weekly numbers
Proposition versus sentences
• Propositions are ideas that can be expressed through declarative sentences
