Expert Systems Questions and Answers

Q: Which inefficiency is characteristic of forward chaining?

Data-driven inference pursues every consequence of the facts, so in a large base most of the work has nothing to do with the question of interest. Redundant questioning and repeated subgoals are backward-chaining faults, and duplicated knowledge bases belong to nothing here. Goal-directed filtering, or a hybrid regime, is the usual remedy when a forward system must answer specific queries. Naming the cost of a choice alongside its benefit is what turns a fact into a design decision. Checking the units of each quantity catches a wrong formula before the arithmetic begins.

Q: What does a goal-driven engine waste effort on most often?

A subgoal needed by several rules is proved afresh each time unless the engine caches results, which is why memoisation matters in goal-directed systems. Irrelevant derivation and exhaustive matching are forward-chaining costs, the latter addressed by Rete. Recognising that the two regimes fail in opposite ways is what guides the choice, and hybrid engines exist precisely to get both benefits.

Q: Which regime do many practical shells actually use?

Real systems commonly chain forward from newly supplied data to derive cheap consequences, then chain backward to test the hypotheses that survive. The mixture keeps both the irrelevant derivation and the repeated questioning within bounds. Committing to one direction is a textbook simplification. Knowing that the choice is a spectrum rather than a dichotomy is the practical point.

Q: Assertion (A): An expert system can explain its reasoning to a user. Reason (R): Its domain knowledge is represented declaratively and the trace of fired rules is available. Select the correct answer from the options below.

Explanation is possible precisely because the knowledge is stored as inspectable rules rather than compiled into procedural code, so the sequence of rules that fired can be replayed as a justification. A conventional program has no comparable record. This is the clearest practical dividend of the separation of knowledge from control, and it was decisive in gaining user acceptance for MYCIN and its successors.

Q: Assertion (A): A large rule base is easy to maintain. Reason (R): Production rules are modular and can be understood independently of one another. Select the correct answer from the options below.

Modularity is real at the level of a single rule but breaks down in aggregate, because rules interact through working memory in ways their authors never considered. XCON's history, with thousands of rules and a permanent maintenance team, is the standard counter-example. So the reason is true as stated and the assertion is false, which is exactly the gap that verification tools and model-based methods were developed to close.

Q: Assertion (A): Certainty factors are a sound calculus of probability. Reason (R): They combine evidence using formulae that assume the pieces of evidence are independent. Select the correct answer from the options below.

The reason states the assumption correctly, and it is that very assumption which makes the assertion false: correlated findings are double counted and confidence is overstated. Heckerman showed certainty factors correspond to probabilities only under conditions real domains violate. The pattern of a true reason establishing the falsity of the assertion is common in this topic and worth reading carefully.

Q: Assertion (A): Model-based diagnosis can identify faults never seen before. Reason (R): It compares predicted behaviour derived from the device structure against what is observed. Select the correct answer from the options below.

Because the diagnosis is computed from a model of correct behaviour rather than from a catalogue of known faults, any discrepancy can be localised whether or not it was anticipated. That is exactly the brittleness a rule base suffers from. The cost is the effort of building and simulating the model, which is why shallow and deep methods are so often layered together in a single system.

Q: Statement I: Working memory persists between consultations. Statement II: The knowledge base persists between consultations. In the light of the above statements, choose the correct answer.

Working memory holds facts specific to the current case and is cleared when the consultation ends, whereas the knowledge base is the system's long-term store and persists. Reversing the two is the standard confusion. The division corresponds to the distinction between what an expert knows and what an expert has just been told, and it is what allows one knowledge base to serve an unlimited number of cases.

Q: Statement I: Forward chaining is data-driven. Statement II: Backward chaining is data-driven. In the light of the above statements, choose the correct answer.

Forward chaining begins with the available facts and derives whatever follows, which is the definition of data-driven, while backward chaining begins with a goal and is therefore goal-driven. The pairing of the two labels with the two directions is asked constantly and is worth fixing firmly. Diagnosis chains backward, monitoring chains forward, and most deployed systems mix the two.

Q: Statement I: An expert system shell contains domain knowledge. Statement II: An expert system shell contains an inference engine. In the light of the above statements, choose the correct answer.

A shell is exactly what remains when the domain knowledge is stripped out: the engine, the explanation facility, the working memory and the authoring tools. EMYCIN was created by performing that removal on MYCIN. Supplying the knowledge is the buyer's job, and it is the expensive part. Getting this the right way round is the whole point of the shell concept.

61.

Which inefficiency is characteristic of forward chaining?

A.Deriving conclusions nobody asked for
B.Asking the user for information that the system could have derived from what it already knows
C.Re-proving the same subgoal
D.Storing a separate copy of the knowledge base for each hypothesis being considered in parallel

Answer: A

Data-driven inference pursues every consequence of the facts, so in a large base most of the work has nothing to do with the question of interest. Redundant questioning and repeated subgoals are backward-chaining faults, and duplicated knowledge bases belong to nothing here. Goal-directed filtering, or a hybrid regime, is the usual remedy when a forward system must answer specific queries. Naming the cost of a choice alongside its benefit is what turns a fact into a design decision. Checking the units of each quantity catches a wrong formula before the arithmetic begins.

Discuss in Forum

62.

What does a goal-driven engine waste effort on most often?

A.Requiring the whole of working memory to be re-examined after every single assertion is made
B.Re-establishing the same subgoal repeatedly
C.Deriving a large number of conclusions that turn out to be irrelevant to the question asked
D.Matching every rule each cycle

Answer: B

A subgoal needed by several rules is proved afresh each time unless the engine caches results, which is why memoisation matters in goal-directed systems. Irrelevant derivation and exhaustive matching are forward-chaining costs, the latter addressed by Rete. Recognising that the two regimes fail in opposite ways is what guides the choice, and hybrid engines exist precisely to get both benefits.

Discuss in Forum

63.

Which regime do many practical shells actually use?

A.Backward chaining only
B.Neither, since a modern shell relies on constraint propagation rather than on chaining at all
C.A mixture of both directions
D.Forward chaining only, since it is the simpler of the two to implement in a production system

Answer: C

Real systems commonly chain forward from newly supplied data to derive cheap consequences, then chain backward to test the hypotheses that survive. The mixture keeps both the irrelevant derivation and the repeated questioning within bounds. Committing to one direction is a textbook simplification. Knowing that the choice is a spectrum rather than a dichotomy is the practical point.

Discuss in Forum

64.

Assertion (A): An expert system can explain its reasoning to a user.

Reason (R): Its domain knowledge is represented declaratively and the trace of fired rules is available.

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

Explanation is possible precisely because the knowledge is stored as inspectable rules rather than compiled into procedural code, so the sequence of rules that fired can be replayed as a justification. A conventional program has no comparable record. This is the clearest practical dividend of the separation of knowledge from control, and it was decisive in gaining user acceptance for MYCIN and its successors.

Discuss in Forum

65.

Assertion (A): A large rule base is easy to maintain.

Reason (R): Production rules are modular and can be understood independently of one another.

Select the correct answer from the options below.

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

Answer: A

Modularity is real at the level of a single rule but breaks down in aggregate, because rules interact through working memory in ways their authors never considered. XCON's history, with thousands of rules and a permanent maintenance team, is the standard counter-example. So the reason is true as stated and the assertion is false, which is exactly the gap that verification tools and model-based methods were developed to close.

Discuss in Forum

66.

Assertion (A): Certainty factors are a sound calculus of probability.

Reason (R): They combine evidence using formulae that assume the pieces of evidence are independent.

Select the correct answer from the options below.

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

Answer: B

The reason states the assumption correctly, and it is that very assumption which makes the assertion false: correlated findings are double counted and confidence is overstated. Heckerman showed certainty factors correspond to probabilities only under conditions real domains violate. The pattern of a true reason establishing the falsity of the assertion is common in this topic and worth reading carefully.

Discuss in Forum

67.

Assertion (A): Model-based diagnosis can identify faults never seen before.

Reason (R): It compares predicted behaviour derived from the device structure against what is observed.

Select the correct answer from the options below.

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

Answer: C

Because the diagnosis is computed from a model of correct behaviour rather than from a catalogue of known faults, any discrepancy can be localised whether or not it was anticipated. That is exactly the brittleness a rule base suffers from. The cost is the effort of building and simulating the model, which is why shallow and deep methods are so often layered together in a single system.

Discuss in Forum

68.

Statement I: Working memory persists between consultations.

Statement II: The knowledge base persists between consultations.

In the light of the above statements, choose the correct answer.

A.Both Statement I and Statement II are correct
B.Both Statement I and Statement II are incorrect
C.Statement I is correct but Statement II is incorrect
D.Statement I is incorrect but Statement II is correct

Answer: D

Working memory holds facts specific to the current case and is cleared when the consultation ends, whereas the knowledge base is the system's long-term store and persists. Reversing the two is the standard confusion. The division corresponds to the distinction between what an expert knows and what an expert has just been told, and it is what allows one knowledge base to serve an unlimited number of cases.

Discuss in Forum

69.

Statement I: Forward chaining is data-driven.

Statement II: Backward chaining is data-driven.

In the light of the above statements, choose the correct answer.

A.Statement I is correct but Statement II is incorrect
B.Both Statement I and Statement II are correct
C.Both Statement I and Statement II are incorrect
D.Statement I is incorrect but Statement II is correct

Answer: A

Forward chaining begins with the available facts and derives whatever follows, which is the definition of data-driven, while backward chaining begins with a goal and is therefore goal-driven. The pairing of the two labels with the two directions is asked constantly and is worth fixing firmly. Diagnosis chains backward, monitoring chains forward, and most deployed systems mix the two.

Discuss in Forum

70.

Statement I: An expert system shell contains domain knowledge.

Statement II: An expert system shell contains an inference engine.

In the light of the above statements, choose the correct answer.

A.Statement I is correct but Statement II is incorrect
B.Statement I is incorrect but Statement II is correct
C.Both Statement I and Statement II are correct
D.Both Statement I and Statement II are incorrect

Answer: B

A shell is exactly what remains when the domain knowledge is stripped out: the engine, the explanation facility, the working memory and the authoring tools. EMYCIN was created by performing that removal on MYCIN. Supplying the knowledge is the buyer's job, and it is the expensive part. Getting this the right way round is the whole point of the shell concept.

Discuss in Forum

Loading…

Topics

Other Subjects

Tags: Expert Systems Questions and Answers || Expert Systems MCQ Questions and Answers || Expert Systems GK Questions and Answers || Expert Systems GK MCQ Questions || Expert Systems Multiple Choice Questions and Answers || Expert Systems GK || GK on Expert Systems || Artificial Intelligence Questions and Answers || Artificial Intelligence MCQ Questions and Answers || Artificial Intelligence GK Questions and Answers || GK on Artificial Intelligence

Expert Systems Questions and Answers

Topics

Other Subjects

About Us

Stay informed about all important Dates & Notifications