Knowledge Representation Questions and Answers

Q: Why do modellers often mistakenly use primitive concepts?

Writing that every mother is a woman with a child states a necessary condition, and a woman with a child will not be classified as a mother unless the condition is also declared sufficient. The modeller then wonders why classification does nothing. Processing cost, syntax support and capacity are not the issue. A failure caused by stating half of what was intended, in a formalism where the halves look similar, is among the commonest errors in ontology engineering.

Q: What is the practical purpose of automatic classification?

In a medical ontology a newly described condition is placed under every category it satisfies automatically, which no manual process could keep correct across tens of thousands of concepts. Design correctness remains a human judgement, axiom counts rise rather than fall, and database conversion is a separate exercise. Automating a maintenance task that grows quadratically with size is what makes ontologies of that scale possible.

Q: What is the acquisition bottleneck for a knowledge representation?

Whatever the formalism, somebody must supply the facts, and eliciting them from experts or extracting them from text has consistently proved the limiting cost. That is why learning from data displaced hand built knowledge bases in many applications. Retrieval, extensibility and readability are secondary. A limitation that lies in the supply of content rather than in the machinery is one that better machinery cannot remove, which is why the field's response was to change method entirely.

Q: How have large language models affected knowledge representation?

Such models encode enormous quantities of world knowledge implicitly and answer questions without any declared ontology, which addresses the acquisition bottleneck directly. What they do not provide is verifiable inference, consistency guarantees or an auditable derivation, which is why hybrid arrangements with explicit knowledge bases are actively pursued. Neither replaces the other, and recognising which properties each supplies is what makes the combination sensible rather than fashionable.

Q: What does an explicit knowledge base provide that a learned model does not?

A logical knowledge base can produce a derivation justifying each conclusion and can be checked mechanically for contradiction, neither of which a learned model offers. Coverage, speed and generalisation are where the learned model is stronger. Setting the two side by side by the properties each supplies, rather than asking which is better, is what makes the case for combining them concrete.

Q: Consider the following statements about knowledge representation. 1. Greater expressiveness generally costs tractability of inference 2. Description logics restrict expressiveness to keep reasoning decidable 3. The closed world assumption is appropriate for web scale knowledge Which of the statements given above is or are correct?

The first two are correct. Web knowledge is manifestly incomplete, so the open world assumption applies and the absence of a statement licenses no conclusion. Adopting a closed world reading there would generate confident falsehoods from missing data. Matching the assumption to how the knowledge was gathered, rather than to what makes reasoning convenient, is the point the three statements together make.

Q: Consider the following pairings in knowledge representation. Reification - treating a proposition as an object Circumscription - minimising abnormality Open world assumption - silence licenses no negative conclusion Subsumption - deciding whether an individual is in a concept How many of the pairs given above are correctly matched?

The first three are correct. Subsumption asks whether one concept includes another, while instance checking asks whether an individual belongs to a concept. The two reasoning services are easily confused because both classify, but one concerns the terminology and the other the data, and they are computed at different times and for different purposes.

Q: Assertion (A): Description logics deliberately restrict what can be expressed. Reason (R): Unrestricted first order entailment is only semidecidable, so reasoning could not be guaranteed to terminate. Select the correct answer from the options below.

Both are true and the second is the reason. Choosing a fragment in which subsumption is decidable, and often polynomial, buys a reasoner that always terminates with an answer, at the cost of statements that cannot be made. Whether the restriction is acceptable depends entirely on the domain, which is why several profiles of differing power exist. Deciding what to surrender in order to obtain a guarantee is the recurring shape of every decision in this topic.

51.

Why do modellers often mistakenly use primitive concepts?

A.The syntax for defined concepts is not widely supported
B.Primitive concepts allow many more individuals to be represented
C.They state necessary conditions and expect automatic classification
D.Primitive concepts are easier for the reasoner to process

Answer: C

Writing that every mother is a woman with a child states a necessary condition, and a woman with a child will not be classified as a mother unless the condition is also declared sufficient. The modeller then wonders why classification does nothing. Processing cost, syntax support and capacity are not the issue. A failure caused by stating half of what was intended, in a formalism where the halves look similar, is among the commonest errors in ontology engineering.

Discuss in Forum

52.

What is the practical purpose of automatic classification?

A.It proves that the ontology has been correctly designed
B.It reduces the number of axioms the ontology requires
C.It converts the whole ontology into a relational database schema
D.Individuals and concepts are placed correctly without manual work

Answer: D

In a medical ontology a newly described condition is placed under every category it satisfies automatically, which no manual process could keep correct across tens of thousands of concepts. Design correctness remains a human judgement, axiom counts rise rather than fall, and database conversion is a separate exercise. Automating a maintenance task that grows quadratically with size is what makes ontologies of that scale possible.

Discuss in Forum

53.

What is the acquisition bottleneck for a knowledge representation?

A.Populating it with knowledge is slow and expensive
B.Retrieving facts from it takes too long at scale
C.The representation cannot be extended once it is designed
D.Only trained logicians can read the resulting sentences

Answer: A

Whatever the formalism, somebody must supply the facts, and eliciting them from experts or extracting them from text has consistently proved the limiting cost. That is why learning from data displaced hand built knowledge bases in many applications. Retrieval, extensibility and readability are secondary. A limitation that lies in the supply of content rather than in the machinery is one that better machinery cannot remove, which is why the field's response was to change method entirely.

Discuss in Forum

54.

How have large language models affected knowledge representation?

A.They can only be used once an ontology has been built
B.They supply broad knowledge without an explicit representation
C.They have entirely replaced the need for any formal reasoning at all
D.They are themselves a form of description logic

Answer: B

Such models encode enormous quantities of world knowledge implicitly and answer questions without any declared ontology, which addresses the acquisition bottleneck directly. What they do not provide is verifiable inference, consistency guarantees or an auditable derivation, which is why hybrid arrangements with explicit knowledge bases are actively pursued. Neither replaces the other, and recognising which properties each supplies is what makes the combination sensible rather than fashionable.

Discuss in Forum

55.

What does an explicit knowledge base provide that a learned model does not?

A.Faster answers to queries about the domain
B.The ability to handle questions never seen before
C.Auditable inference and guaranteed consistency
D.Broader coverage of everyday world knowledge

Answer: C

A logical knowledge base can produce a derivation justifying each conclusion and can be checked mechanically for contradiction, neither of which a learned model offers. Coverage, speed and generalisation are where the learned model is stronger. Setting the two side by side by the properties each supplies, rather than asking which is better, is what makes the case for combining them concrete.

Discuss in Forum

56.

Consider the following statements about knowledge representation.

1. Greater expressiveness generally costs tractability of inference

2. Description logics restrict expressiveness to keep reasoning decidable

3. The closed world assumption is appropriate for web scale knowledge

Which of the statements given above is or are correct?

A.1 and 3 only
B.2 and 3 only
C.1, 2 and 3
D.1 and 2 only

Answer: D

The first two are correct. Web knowledge is manifestly incomplete, so the open world assumption applies and the absence of a statement licenses no conclusion. Adopting a closed world reading there would generate confident falsehoods from missing data. Matching the assumption to how the knowledge was gathered, rather than to what makes reasoning convenient, is the point the three statements together make.

Discuss in Forum

57.

Match each formalism with its characteristic feature.

List I

A. Semantic network

B. Frame

C. Description logic

D. Default logic

List II

P. concepts as nodes and relations as links

Q. slots holding an object's properties

R. decidable subsumption over concept definitions

S. rules applied unless contradicted

Select the correct match.

A.A-P, B-Q, C-R, D-S
B.A-Q, B-P, C-S, D-R
C.A-R, B-S, C-P, D-Q
D.A-P, B-R, C-Q, D-S

Answer: A

The first two are largely notational variants of one another and of a first order fragment, the third adds formal semantics and decidability, and the fourth addresses defeasibility. Each was developed to fix a shortcoming of its predecessors, so the sequence reads as a response to successive difficulties rather than as a list of alternatives.

Discuss in Forum

58.

Consider the following pairings in knowledge representation.

Reification - treating a proposition as an object

Circumscription - minimising abnormality

Open world assumption - silence licenses no negative conclusion

Subsumption - deciding whether an individual is in a concept

How many of the pairs given above are correctly matched?

A.All four
B.Only three
C.Only one
D.Only two

Answer: B

The first three are correct. Subsumption asks whether one concept includes another, while instance checking asks whether an individual belongs to a concept. The two reasoning services are easily confused because both classify, but one concerns the terminology and the other the data, and they are computed at different times and for different purposes.

Discuss in Forum

59.

Three of the following are non-monotonic. Which is the odd one out?

First order resolution

Circumscription

Default logic

Inheritance with exceptions

A.Default logic
B.Inheritance with exceptions
C.First order resolution
D.Circumscription

Answer: C

Circumscription, default logic and defeasible inheritance can all withdraw a conclusion when a fact is added. First order resolution is monotonic: anything derivable from a set of premises remains derivable when more are added. That property, which classical logic takes for granted, is precisely what commonsense reasoning violates, and identifying which formalisms preserve it decides which can model defaults at all.

Discuss in Forum

60.

Assertion (A): Description logics deliberately restrict what can be expressed.

Reason (R): Unrestricted first order entailment is only semidecidable, so reasoning could not be guaranteed to terminate.

Select the correct answer from the options below.

A.Both A and R are true, but R is not the correct explanation of A
B.A is true but R is false
C.A is false but R is true
D.Both A and R are true, and R is the correct explanation of A

Answer: D

Both are true and the second is the reason. Choosing a fragment in which subsumption is decidable, and often polynomial, buys a reasoner that always terminates with an answer, at the cost of statements that cannot be made. Whether the restriction is acceptable depends entirely on the domain, which is why several profiles of differing power exist. Deciding what to surrender in order to obtain a guarantee is the recurring shape of every decision in this topic.

Discuss in Forum

Loading…

Topics

Other Subjects

Tags: Knowledge Representation Questions and Answers || Knowledge Representation MCQ Questions and Answers || Knowledge Representation GK Questions and Answers || Knowledge Representation GK MCQ Questions || Knowledge Representation Multiple Choice Questions and Answers || Knowledge Representation GK || GK on Knowledge Representation || Artificial Intelligence Questions and Answers || Artificial Intelligence MCQ Questions and Answers || Artificial Intelligence GK Questions and Answers || GK on Artificial Intelligence

Knowledge Representation Questions and Answers

Topics

Other Subjects

About Us

Stay informed about all important Dates & Notifications