Artificial Intelligence Knowledge Representation and Reasoning 3

Question 1

A Bayesian network is best used in AI for:

Accepted Answer

Representing probabilistic dependencies among variables and performing uncertainty reasoning

Answer

Sorting large datasets with optimal time complexity

Answer

Generating photorealistic images from text descriptions

Answer

Implementing breadth-first search in state spaces

Question 2

Case-based reasoning (CBR) in AI solves new problems by:

Accepted Answer

Retrieving and adapting solutions from similar past cases stored in memory

Answer

Deriving solutions from first-principles logical axioms

Answer

Randomly sampling candidate solutions and selecting the best one

Answer

Training a neural network on historical data

Question 3

Which query language is used to retrieve and manipulate data stored in RDF (Resource Description Framework) knowledge graphs?

Accepted Answer

SPARQL

Answer

SQL

Answer

XPath

Answer

Cypher

Question 4

Description logics (DLs) are used in AI primarily to:

Accepted Answer

Formally represent and reason about concept hierarchies and object properties in ontologies

Answer

Describe neural network architectures in natural language

Answer

Write narrative descriptions of datasets for documentation

Answer

Specify loss functions for machine learning models

Question 5

In knowledge representation, what is the 'is-a' (or subsumption) relationship used to express?

Accepted Answer

That one class is a subset of or specialization of another class

Answer

That two objects share the same numerical identifier

Answer

That an object possesses a particular attribute value

Answer

That a class is associated with an external database table

Question 6

Which of the following reasoning types starts from observations and infers the most likely explanation for those observations?

Accepted Answer

Abductive reasoning

Answer

Deductive reasoning

Answer

Inductive reasoning

Answer

Analogical reasoning

Question 7

Non-monotonic reasoning in AI is important because it allows the system to:

Accepted Answer

Revise or retract earlier conclusions when new contradictory information arrives

Answer

Guarantee conclusions that can never be withdrawn once drawn

Answer

Increase the monotonic growth of a neural network's accuracy over time

Answer

Process data streams in a strictly sequential order