My Reflection
Overall Reflection
The first unit started with a seminar introduced by the module's tutor, Dr. Godfried Williams, in which he
introduced the overall
topic and concepts of the topic. The unit also included a lecturecast to introduce related concepts, like RDF,
ontologies, and semantic web.
There were also some readings and a couple of activities, including a collaborative discussion.
Overall, I felt the unit to be a swift introduction to the topic. The eBook seemed promising. Considering that I
have been waiting for
this module since the start of my MSc AI program, I was excited to get started, and I am looking forward to a
rewarding learning experience.
Readings Reflection
The unit initiated our reading in the main eBook for the module, "Knowledge Representation and Reasoning" by
Brachman and Levesque (2009). We read Chapter 1,
which is the introduction. For me, it was a good reading, as it had a bit of philosophy and abstract discussion
about what knowledge, reasoning and representation are,
as well as what facts and logic are, which are all interesting topics for me.
Below are some quotes that I specially like from the reading:
-
"One definition of Artificial Intelligence (AI) is that it is the study of intelligent behavior achieved
through computational means."
-
"This suggests that, among other things, knowledge is a relation between a knower, like John, and a
proposition, that is,
the idea expressed by a simple declarative sentence, like “Mary will come to the party.“
-
"Imagine, for example, playing a game of chess against a complex chess-playing program. In looking at one of
its moves,
we might say to ourselves something like this: “It moved this way because it believed its queen was
vulnerable,
but still wanted to attack the rook.” In terms of how the chess-playing program is actually constructed, we
might have said something
more like, “It moved this way because evaluation procedure P using static evaluation function Q returned a
value of +7 after an
alpha-beta minimax search to depth 4.” The problem is that this second description, although perhaps quite
accurate, is at the wrong
level of detail, and does not help us determine what chess move we should make in response. Much more useful
is to understand the
behavior of the program in terms of the immediate goals being pursued relative to its beliefs, long-term
intentions, and so on.
This is what the philosopher Daniel Dennett calls taking an intentional stance toward the chess-playing
system."
-
"The philosopher Hubert Dreyfus first observed this paradox of “expert systems.” These systems are claimed to
be superior
precisely because they are knowledge-based, that is, they reason over explicitly represented knowledge.
But novices are the ones who think and reason, claims Dreyfus. Experts do not; they learn to recognize and to
react.
The difference between a chess master and a chess novice is that the novice needs to figure out what is
happening and what to do,
but the master just “sees” it. For this reason (among others), Dreyfus believes that the development of
knowledge-based systems is
completely wrongheaded if it is attempting to duplicate human-level intelligent behavior."
Reference
Brachman, R.J. and Levesque, H.J. (2009) Knowledge representation and reasoning. Elsevier.
Artefacts
1. Formative Activities
Activity 1
The activity prompted the students to consider the following subjects and whether they consider themselves as
'knowing'
or 'having information about'. In the table below are my answers for each subject.
| Subject |
My answer |
| A second language in which you are fluent. |
Knowing |
| The content of a television news programme. |
Knowing |
| A close friend. |
Knowing |
| A company’s annual report. |
Knowing |
| Your close friend’s partner whom you have yet to meet. |
Having information about |
| The weather on the other side of the world. |
Having information about |
| The weather where you are now. |
Knowing |
Activity 2
The follow up question on Activity 1 was, "What would you suggest is the primary characteristic that
distinguishes the
‘having information’ situations from the ‘knowing’ situations you categorised in the previous activity?"
My answer to this question is that 'knowing' involves a deeper, more personal connection and understanding of
the subject,
often gained through direct experience or interaction. In contrast, 'having information about' is more detached
and impersonal,
typically acquired through secondary sources without direct engagement.
For example, knowing a close friend involves personal experiences, shared memories, and emotional bonds, whereas
having information
about a friend's partner whom I have yet to meet is based solely on descriptions or second-hand accounts without
any personal interaction.
Similarly, knowing the weather where I am now comes from direct sensory experience, while having information
about the weather
on the other side of the world relies on reports or data without any immediate personal experience.
I can relate to this seeing it is similar to the distinction between sources of information and data in research
and data collection, being
primary and secondary sources. Primary sources would be akin to 'knowing', as they provide direct, firsthand
evidence or experience of a subject,
whereas secondary sources would be more like 'having information about', as they offer interpretations,
analyses, or summaries based on primary sources.
Artefacts
2. Collaborative Learning Discussion 1
Initial Post
In the first collaborative learning activity for this unit, we were asked to discuss the following prompt,
mainly by expressing whether
we agree or disagree with the statement, and to discuss how reasoning is related to knowledge representation.
"Knowledge Representation is a recent phenomenon – it only became a topic of discussion
with the development of computing technology and the need to represent knowledge in computer systems."
Here is my initial post:
The field of Knowledge Representation and Reasoning (KRR) is all about how knowledge can be represented and used
for reasoning. While this emerged as a solidified domain with the emergence of complex computational systems,
the underlying questions of the field are as old as human civilisations and have been discussed since the start
of philosophy (Brachman and Levesque, 2009).
At its core, the question about representation and reasoning requires questioning and understanding what
knowledge is in the first place. This also entails a chain of questions on the meanings of beliefs, ideas, facts
and their relations to the world, existence and being (Bouquet et al., 2003). These entailed questions are
primarily philosophical ones, studied by the dedicated branches of epistemology (the study of knowledge),
ontology (the study of being) and—of course—logic, which are all rooted in the first attempts at
philosophy in ancient civilisations (Truncellito, 2024; Maedche, 2002).
Nevertheless, these sorts of questions acquired a new context in the digital age, under the need to clone our
understanding of the world—that is to say, our knowledge—to machines and computational systems to apply specific
tasks. This is when the specific question about ‘representation’ crystallised, as, for the first time, humans
did not only need to understand their own understanding themselves, but to explain—better to say, transfer—it to
‘things’ that had not developed naturally with natural cognitive capacity. The field of KRR then started to
formulate to look into such urgency and has continued to be pushed by the frequent development and expansion of
computational capabilities.
And why do we need to explain knowledge to computers? It is so that they acquire the ability of ‘reasoning’,
which is again a natural tendency for humans, but a technical capability that needs to be founded. How can we
give computers the gift of reasoning? This again required humans to look into how they reason themselves. In all
cases, it is inevitable to see that no reasoning is achievable without a representation of knowledge first—that
is, the presence of relationships between pieces of information and in a specific context. That is why and how
reasoning is related to knowledge representation (Bouquet et al., 2003). The latter is a priori to the first
(Delgrande et al., 2023). For the sake of reasoning, we represent knowledge.
In conclusion, I see the fundamental questions of KRR are as old and everlasting as the wondering human being.
Yet, the emergence of computational systems induced the urgency of transferring human’s natural capacity to
machines, and then the question of knowledge representation formulated as a dedicated field of research, closely
attached to the fields of artificial intelligence, but also other fields impacted by the effect of computational
systems. The goal of transferring, and representing, knowledge will always be reaching the state of reasoning,
and that is how reasoning is related to and dependent on knowledge representation.
Reference list
Bouquet, P. et al. (2003) ‘Theories and Uses of Context in Knowledge Representation and Reasoning’, Journal of
Pragmatics, 35(3), pp. 455–484. Available at: https://doi.org/10.1016/s0378-2166(02)00145-5.
Brachman, R.J. and Levesque, H.J. (2009) Knowledge representation and reasoning. Elsevier.
Delgrande, J. et al. (2023) ‘Current and Future Challenges in Knowledge Representation and Reasoning’, Dagstuhl
Manifestos, 1(1), pp. 1–58. Available at: https://arxiv.org/pdf/2308.04161 (Accessed: 26 October 2025).
Maedche, A. (2002) ‘Ontology — Definition & Overview’, in Ontology Learning for the Semantic Web. Boston,
MA: Springer, pp. 11–27.
Truncellito, D. (2024) Epistemology, Internet Encyclopedia of Philosophy. Available at: https://iep.utm.edu/epistemo/ (Accessed: 26 October 2025).