SmartMuseum

Homework 3

Task 1

4 x 4 example (N = 4)

1. Register queens in the game: Queen0, Queen1, Queen2, Queen3
2. Queen0

• filled positions (array[])
  • (2, 0) -> safe
  • Send message to Queen1 (array[(2,0)])

3. Queen1

• ACLMessage (array[(2,0)])
• Strategy (0,1)
• save or not (not on row, column, diagonal)
• Send message to Queen2 (array[(2,0),(0,1)])

4. Queen2

• ACLMessage (array[(2,0),(0,1)])
• find position (0,2) -> not safe
• find position (3,2) -> safe
• Send message to Queen3 (array[(2,0),(0,1),(3,2)])

5. Queen3

• ACLMessage (array[(2,0),(0,1),(3,2)])
• find position (1,3) -> safe
• Final (array[(2,0),(0,1),(3,2),(1,3)])

In any case, where we don't find a safe position

• Message to previous queen to move

Task 2

• ControllerAgent and ControllerAgentGUI
  • Issue clone/move commands
  • Create containers
  • ControllerAgent -> Mobile Agent
    • move -> doMove, beforeMove, afterMove
    • clone -> doClone, beforeClone, afterClone
  • Implement GUI
• Mobile Agents
  • MobileAgent
  • MobileAgentGUI
  • Adjust ArtistManager (Auctioneer)
  • Adjust Curator (Bidder)
  • Implement doMove, beforeMove, ..., auctionLogic
  • Implement GUI
• Main Container: ControllerAgent
• Container 1: ArtistManager (clones 2)
  • One clones moves to container 2 and the other to 3
• Container 2: Curator (clones 2)
• Container 3: Curator (clones 2)
• Conduct auction in both containers
• Clones return to container 1 and decide on best price

Homework 2

Task 1

[Auctioneer] -> Agent

• Parameters
• Initial price
• Artwork name
• Reserve price
• Rate of reduction
• Responsibilities

1. Search for "Bidding" service -> is of bidder Agent
2. ACLMessage INFORM
3. Behaviour -> CFP -> Reserve price
4. Initial price
5. Modified price
6. Behaviour

• Handle propose messages (modify price, rate of reduction)
• Handle not udnerstood messages
• Send ACCEPT_PROPOSAL
• Send REJECT_PROPOSAL

[Bidder] -> Agent

1. Register service in DF
2. Receive INFORM
3. Behaviour

• Handle CFP
• Handle accepted proposal
• Handle rejected proposal
• Handle two strategies

2 scenarios with at least 4 agents

Task 2

Agents = {Profiler, Curator, ArtistManager}

Outcomes

• High quality
• Low quality

Actions

• ArtistManager
  • Sell high
  • Sell low
• Curator
  • Quote based on demand
  • Quote based on interest
• Profiler
  • view
  • not view

Each combination of actions: {ArtistManager, Curator, Profiler} eg.: {Sell high quality, Quote price on demand, view}

• develop preferences
• Payoff matrix

Homework 1

Tour Guide Agent

1. Register "virtual tour" service
2. Add behaviour to listen for requests from profiler agent

• matching interest

3. Add behaviour that builds virtual tour for profiler

• Communicate with curator to get list of artifacts
  • Send and receive Interests: 1 Flower, 2 Portrait, etc.

Curator Agent

1. Retrieve requests from virtual tour guide agent
   • send list of artifacts, e.g. {"Painting 1", "Flowers", "..."}
2. Retrieve requests from profiler agent
   • Detailed: id, name, creator, date, type, ..

Profiler Agent

1. Read arguments from user
   • Interests and other info
2. Subscribe to the "DF" -> to use "virtual_tour" service
3. (One or all - parallel behaviour)

• Search for a virtual tour guide agent (DF)
• You might not find the agent quickly
  • Keep searching
• Send requests to one or more virtual tour guide agent
• Receive replies from one or more virtual tour guide agents
• Send response to the accepted virtual tour guide agent
• Receive the contents of the virtual tour
• Communicate with curator to get details about an artifact
• Receive reply from curator

UT Agent|--register-->| DF ( virtual_tour) |<--subscribe--|Profile|-->A1,A2,A3