Class LotteryDialogueTest
java.lang.Object
org.tweetyproject.agents.dialogues.examples.LotteryDialogueTest
Shows how a simulation of a multi-agent system can be set up. It defines a
dialogue game between different agents, in particular one based on an action
selection strategy using lotteries.
- Author:
- Matthias Thimm
-
Field Summary
Modifier and TypeFieldDescriptionstatic double
The probability that an attack exists between any two arguments in the argumentation framework.static boolean
A flag indicating whether the framework should enforce a tree shape structure.static int
The size of the argumentation framework being used in the simulation.static int
The number of simulation runs to be executed for each experiment.static long
A fixed random seed to ensure comparability of simulation results.static long
A fixed random seed to ensure comparability of simulation results.static long
A fixed random seed to ensure comparability of simulation results.static long
A fixed random seed to ensure comparability of simulation results.static Semantics
The semantics to be used for evaluating the argumentation framework in the simulation.static int
The timeout duration for each simulation run in seconds. -
Constructor Summary
-
Method Summary
Modifier and TypeMethodDescriptionstatic void
LotteryDialogue test main methodstatic void
runSimulation
(boolean baseline) Runs a simulation of the multi-agent system where agents engage in a dialogue game based on a Dung theory.
-
Field Details
-
frameworkSize
public static int frameworkSizeThe size of the argumentation framework being used in the simulation. Represents the number of arguments in the framework. -
attackProbability
public static double attackProbabilityThe probability that an attack exists between any two arguments in the argumentation framework. The value should be between 0.0 and 1.0, where 0.0 means no attacks, and 1.0 means all possible attacks are present. -
enforceTreeShape
public static boolean enforceTreeShapeA flag indicating whether the framework should enforce a tree shape structure. If true, the argumentation framework will be constructed as a tree (no cycles). -
timeout
public static int timeoutThe timeout duration for each simulation run in seconds. Default is set to 72 hours (259200 seconds). -
numberOfRunsEach
public static int numberOfRunsEachThe number of simulation runs to be executed for each experiment. Default is set to 100 runs. -
semantics
The semantics to be used for evaluating the argumentation framework in the simulation. Default is set to GROUNDED_SEMANTICS. -
RANDOM_SEED1
public static long RANDOM_SEED1A fixed random seed to ensure comparability of simulation results. -
RANDOM_SEED2
public static long RANDOM_SEED2A fixed random seed to ensure comparability of simulation results. -
RANDOM_SEED3
public static long RANDOM_SEED3A fixed random seed to ensure comparability of simulation results. -
RANDOM_SEED4
public static long RANDOM_SEED4A fixed random seed to ensure comparability of simulation results.
-
-
Constructor Details
-
LotteryDialogueTest
public LotteryDialogueTest()Default Constructor
-
-
Method Details
-
runSimulation
Runs a simulation of the multi-agent system where agents engage in a dialogue game based on a Dung theory. The simulation is repeated for a number of runs and the results are aggregated. The simulation runs with a set timeout, and if it exceeds the timeout, it is aborted.- Parameters:
baseline
- a flag to indicate whether the baseline agent (UtilityBasedAgent) or the probabilistic lottery agent (ProbabilisticLotteryAgent) should be used in the simulation.- Throws:
ProtocolTerminatedException
- if the protocol ends prematurely during the simulation.
-
main
LotteryDialogue test main method- Parameters:
args
- the args- Throws:
ProtocolTerminatedException
- error
-