UIMP: Sistema Multiagente CBR para Turismo de Salamanca

Development of CBR-BDI Agents: A Tourist Guide Application

http://gsii.usal.es

Juan M. Corchado

Departamento de Informática y Automática

Universidad de Salamanca

- 2 -

Development of CBR-BDI Agents

Index: Introduction

Technology review

Proposal

motivation

goals

agents

agents and cbr systems

aplication

demo

conclusions

wireless implementation

- 3 -

Motivation

The telecommunication industry expects a new expansion with the development of UMTS and third generation phone systems.

The new challenges of this field require new technology that facilitate the construction of more dynamic, intelligent, flexible and open applications, capable of working in a real time environments.


Introduction– motivation– goal

Technical Review– agents– cbr-bdi agents– wireless implem.

Proposal– aplication– demo– conclusion

- 4 -





The development of efficient wireless

distributed systems.

Composed of autonomous elements

with reasoning capabilities.

Multiagent technology

JADE-LEAP

Autonomousagent

CBR-BDI system

- 5 -

Goal– The development of an agent based architecture that

facilitate the construction of:Deliberative agents

– Autonomous– With reasoning capabilities– With communication capabilities– With adaptation capabilities

BDI agents– Believes

– Desires

– Intentions

Case study– tourism guide system– Wireless devices




Proposal– aplication– demo– conclusion Cases, Variables, past

experiences, Expected solutions

Plans

- 6 -

Reactive agents– Autonomous application with response capabilities

Deliberative agents– Autonomous– With reasoning capabilities– With communication capabilities– With adaptation capabilities

BDI Agents: believes, desires and intentions





Agents

PerceptionSensors

ActionsActuators

Environment

- 7 -





BDI agent (beliefs, desires , intentions)Belief: stateDesire: set of <final_state>Intention: sequence of <action>

Case: <Problem, Solution, Result> Problem: initial_stateSolution: sequence of <action, [intermediate_state]>Result: final_state

Agent plans the solution strategy

Agent stars to solve a new problem

New CBR reasoning cycle

Agent achieves its goal

CBR solution achieved

Agent updates knowledge

Case retain - learning

Case retrieval Case reuse Case revise

Agent knowledge base-

Case-base

Believes…Desires…Intentions…

Cases

- 8 -

Operative SystemJAVA j2se, j2me

JADE-LEAP 3.0

Agent (customised)Agent Interaction





- 9 -





JADE - LEAP

- 10 -





- 11 -





User [us001]

User [us002]

User [us003]

User [us00X]

………

Tracker agentPerformer agent [us001]

Performer agent [us002]

Performer agent [us003]

Performer agent [us00X]

Planner agent

Internet worldWireless world

- 12 -





-<<Role>>-K-base

-Update Believes/Intentions-<<Role Dynamic>>

-VCBP

<<agent>> Planner

- Input-Request ACL for service (Give MRS)-(ACL content = {O, R, hi, UsedBel} )

-O = Objetivos-R = Recursos

-hi : int-UsedBel

<<Capability>> K-base

-Description-Given a set of Preferences about a problem P

-this service offers the Most Replanning-able Solution

<<Service>> Give MRS

TypeInform, Failure

P rotocol:Request-Best Solution for a dynamic environment

Agent Comunication LanguageFIPA ACL

OntologyPlanning ontologyContent Language

FIPA SL

-Input-{ S(p) } S1(p),S2(p),S3(p),..Sn(p) : Posible Solutions

<<Capability>> VCBP

Output:Sf(p) : MRS (Most Replanning-able Solution)

Description:This capability provides the most replanning-able

solution to the performer Agent

Output:S1(p),S2(p),S3(p),..Sn(p) : P osible Solutions

Descr iption:This capability provides solutions that fulfill a

set of given preferences

-Input-Inform ACL for Update Believes/Intentions

-(ACL Content =-b1(t) ,b2(t) ,...bn(t) : Believe

-t : time )

<<Capability>> Update Believes/Intentions

Output:bi(t-1) <- bi( t) : Believe

I[bi(t-1) <- bi( t)] : Intentions

Description:This capability Updates believes and

intentions

Planner Agent class diagram in AUML

- 13 -





0 2 4 6 8 10 12

0

1

20

5

10

15

20

25

30

35

40

45

χ 0(α ,β )

χ 0(α ,β )

χ 0(α ,β )χ 0(α ,β )

P a s o 1 : t = t0 P a s o 2 : t = t1

P a s o 3 : t = t2 P a s o 4 : t = t f

step1: t=t0 step2: t=t1

step3: t=t2 step4: t=t3

Glez-Bedia M. y Corchado J. M. (2002) A planning strategy based on variational calculus for deliberative agents. Computing and Information Systems Journal. Vol. 9 No. 2. pp: 2-13. ISSN: 1352-9404

Glez-Bedia M., Corchado J. M., Corchado E. S. ,Fyfe C. (2002) Analytical Model for Constructing Deliberative Agents. Engineering Intelligent Systems. Vol. 10. No 3. pp: 173-185. ISSN 1472-8915.

- 14 -





- 15 -





- 16 -





- 17 -





- 18 -

% Evaluation - degree of satisfaction

Tourists that… 8-10 6-8 4-6 0-4 No answer

Used the help of the agent

14% (55,9%) (4,7%) (2,4%) (0,7%) (36,3%)

Used the help of a tourist guide

23% (62,7%) (19,6%) (8,9%) (1%) (7,8%)

Did not use any of the previous

63% (16,7%) (8,3%) (1,2%) (0,2%) (78,8%)





Tested on 6217 Tourist

- 19 -





The CBR-BDI agents

• facilitate the construction of distributed wireless system for mobile devices and

• may be adapted for different problem domains, within the constrains imposed by the industry.

The developed infrastructure includes tools

• for generating CBR-BDI autonomous agents that can reason, learn and communicate with the users and with other agents,

• a simple communication protocol based on the FIPA ACL standards, and

• a number of established processes that facilitate the analysis and design of a MAS using AUML.

UIMP: Sistema Multiagente CBR para Turismo de Salamanca

Technology

Transcript of UIMP: Sistema Multiagente CBR para Turismo de Salamanca