SEMESTER I EXAMINATIONS ACADEMIC YEAR 2018/2019
COMP 30240 & COMP 41400 Multi-Agent Systems
Prof. J. Pitt
Prof. P. Cunningham Prof. G.M.P. O¡¯Hare *
Time Allowed: 2 Hours Instructions for Candidates
Answer any two questions. All questions carry equal marks Total marks available 100.
Instructions for Invigilators
ýUCD 2018/19
Page 1 of 6
Q.1. (a)
(b)
Explain the essence of Speech Act Theory and why this has been so influential within the development of Multi-Agent Systems (MAS).
[ 15 Marks ]
Cohen and Perrault postulated a plan-based theory of speech acts such that planning systems could reason about them. Adopting a STRIPS like notation define:
Request (S, H, ¦Á) where:
S represents the Speaker;
H represents the Hearer;
¦Á represents the action being requested;
Clearly express the preconditions and effect of this Request speech act.
[ 15 Marks ]
Critically compare and contrast object oriented programming with that of agent-oriented programming.
[ 20 Marks ]
ýUCD 2018/19
Page 2 of 6
(c)
Q.2. (a)
(b)
(c)
(d)
Compare and contrast the reactive or situated action agent architecture and that of the intentional or deliberative architecture.
Indicate the relative advantages and disadvantages of each.
[ 15 Marks ]
Describe in detail how intentions can form a basis for a practical reasoning system.
[ 15 Marks ] Explain what is meant by the terms Bold agent and Cautious agent.
Characterise those environments that are most suitable to each.
[ 10 Marks ]
Reconsideration of intentions is costly. Assume the existence of a Boolean function deliberate which determines when to reconsider existing intentions.
Briefly comment on issues of performance and optimality associated with the deliberate function and the frequency of its invocation.
[ 10 Marks ]
ýUCD 2018/19
Page 3 of 6
Q.3 (a)
(b)
(c) (d)
Explain in detail what is understood by the term Belief Desire Intention Architecture (BDI).
[ 10 Marks ]
Explain and describe the specialisation of this abstract architecture which is incorporated into the AgentSpeak (L) agent programming
language.
[ 10 Marks ] Describe the core constructs provided within Agentspeak(L).
[ 20 Marks ]
Describe the core Agentspeak(L) interpreter cycle making reference to data structures that would need to be accessed.
[ 10 Marks ]
ýUCD 2018/19
Page 4 of 6
Q.4. (a) (b)
(c)
(d)
(e)
Describe the Prisoner¡¯s Dilemma
ýUCD 2018/19
Page 5 of 6
Explain the game-theoretic notation of a payoff matrix and utilise this to characterise the four possible outcomes associated with the dilemma. Explain each outcome.
[ 10 Marks ]
Utilise this payoff matrix to inform the action that a prisoner ought to adopt and clearly explain why they ought to adopt this strategy.
[ 10 Marks ]
Contrast linear time temporal logic and branching time temporal logics. [ 10 Marks ]
Within the following tree assume that S is some point in time. The arcs within the tree represent possible future worlds. For each world state those concepts that are believed by an agent to be true are depicted by the notation {x,v,g} indicating that x, v and g are all believed to be true within that particular future world state.
For State S express using branching time temporal logical formula(e) those relationship(s) that hold true at the states and paths
contained within the diagram.
[ 10 Marks ]
ýUCD 2018/19 Page 6 of 6
[10 Marks ]