1. Luís Moniz Pereira CENTRIA, Departamento de Informática Universidade Nova de Lisboa lmp@di.fct.unl.pt Pierangelo Dell’Acqua Dept. of Science and Technology Linköping University pier@itn.liu.se

2. Our agents FWe propose a LP approach to agents that can: 3Reason and react to other agents 3Prefer among possible choices 3Intend to reason and to act 3Update their own knowledge, reactions and goals 3Interact by updating the theory of another agent 3Decide whether to accept an update depending on the requesting agent

3. Framework FThis framework builds on the work: 3 Updating Agents - P. Dell’Acqua & L. M. Pereira MAS’99 3 Updates plus Preferences - J. J. Alferes & L. M. Pereira JELIA’00

4. Updating agents  Updating agent: a rational, reactive agent that can dynamically change its own knowledge and goals: 8makes observations 8reciprocally updates other agents with goals and rules 8thinks a bit (rational) 8selects and executes an action (reactive)

5. Updates plus preferences FA logic programming framework that combines two distinct forms of reasoning: preferring and updating. Updates create new models, while preferences allow us to select among pre-existing models  The priority relation can itself be updated.  A language capable of considering sequences of logic programs that result from the consecutive updates of an initial program, where it is possible to define a priority relation among the rules of all successive programs.

6. Preferring agents Agents can express preferences about their own rules. FPreferring agent: an agent that is able to prefer beliefs and reactions when several alternatives are possible. Preferences are expressed via priority rules. Preferences can be updated, possibly on advice from others.

7. Agent’s language  Atomic formulae: A objective atoms not A default atoms i:C projects updates : - i C FFormulae: A  L 1  L n not A  L 1  L n L 1  L n  Z L i is an update or an atom active rule generalized rules Z j is a project integrity constraint false  L 1  L n  Z 1  Z m

8. Agent’s language i : ( A  L 1  L n ) i : ( L 1  L n  Z ) i : ( ?- L 1  L n )  A project i:C can take one of the forms: i : ( not A  L 1  L n ) goal i : ( false  L 1  L n  Z 1  Z m ) FNote that a program can be updated with another program, i.e., any rule can be updated.

9. Agents’ knowledge states FKnowledge states represent dynamically evolving states of agents’ knowledge. They undergo change due to updates.  Given the current knowledge state P s, its successor knowledge state P s+1 is produced as a result of the occurrence of a set of parallel updates. FUpdate actions do not modify the current or any of the previous knowledge states. They only affect the successor state: the precondition of the action is evaluated in the current state and the postcondition updates the successor state.

10. Projects and updates  A project j:C denotes the intention of some agent i of proposing the updating the theory of agent j with C.  denotes an update proposed by i of the current theory of some agent j with C. wilma:C : - i C: - fred C

11. Priority rules FLet < be a binary predicate symbol whose set of constants includes all the generalized rules: r 1 < r 2 means that the rule r 1 is preferred to the rule r 2. A priority rule is a generalized rule defining <.

12. Prioritized abductive LP FA prioritized LP is a set of generalized rules (possibly, priority rules) and integrity constraints.

13. Agent theory FThe initial theory of an agent  is a pair (P,R): - P is an prioritized LP. - R is a set of active rules. FAn updating program is a finite set of updates. FLet S be a set of natural numbers. We call the elements s  S states. FAn agent  at state s, written  ,s, is a pair (T,U): - T is the initial theory of . - U={U 1,…, U s } is a sequence of updating programs.

14. Multi-agent system FA multi-agent system M={   1,s,…,   n,s } at state s is a set of agents  1,…,  n at state s. FM characterizes a fixed society of evolving agents. FThe declarative semantics of M characterizes the relationship among the agents in M and how the system evolves. FThe declarative semantics is stable models based.

15. Distributed databases and cooperative agents Then p can be characterized by (P,R), where rC  reject(rC)  NrC NrC  t:NrC P = R = rC=residence of Carlo NrC=new residence of Carlo Communication and updates allow to integrate distinct agents. Assume that we want to minimize the administrative procedure required for changing residence. For example, we may notify the new residence once in a public office (p). Then it is the responsibility of that office to inform all the relevant offices.

16. Representation of conflicting information and preferences This example models a situation where an agent, Fabio, receives conflicting advice from two reliable authorities. Let (P,R) be the initial theory of Fabio, where R={} and dont(A)  fa(noA)  not do(A) (r 1 ) do(A)  ma(A)  not dont(A) (r 2 ) false  do(A)  fa(noA) false  dont(A)  ma(A) r 1 < r 2  fr r 2 < r 1  mr P = fa=father advises ma=mother advises fr=father responsability mr=mother responsability Preferences may resolve conflicting information.

17. Representation of conflicting information and preferences Suppose that Fabio wants to live alone, represented as lA. U 1 = His mother advises him to do so, but the father advises not to do so: mother ma(lA), : - father fa(nolA) : - Assuming that there are no rejection clauses, Fabio accepts both updates, and therefore he is still unable to choose either do(lA) or dont(lA) and, as a result, does not perform any action whatsoever.

18. Representation of conflicting information and preferences U 2 = Afterwards, Fabio's parents separate and the judge assigns responsibility over Fabio to the mother: judge mr : - Now the situation changes since the second priority rule gives preference to the mother's wishes, and therefore Fabio can happily conclude ”do live alone”.

19. Updating preferences Within the theory of an agent both rules and preferences can Here internal projects of an agent are used to update its own priority rules. The updating process is triggered by means of external or internal projects. be updated.

20. Updating preferences Let the theory of George be characterized by : workLate  not party (r 1 ) party  not workLate (r 2 ) money  workLate (r 3 ) r 2 < r 1 beautifulWoman  george: wishGoOut wishGoOut  not money  george: getMoney wishGoOut  money  beautifulWoman: inviteOut getMoney  george: r 1 < r 2 getMoney  george: not r 2 < r 1 P = R = partying is prefered to working until late to get money, George must update his priority rules

21. Applications FApplications in which our agent technology can have a significant potential to contribute are internet applications, e.g. - information integration - web-site management

22. Engineering agent societies FWe believe that the theory of our agents is rich and suitable to engineer configurable, dynamic, self- organizing and self-evolving agent societies. FJennings argues that: -open, networked systems are characterized by the fact that there is no simple controlling organization. -the computational model of these systems places several requirements.

23. Engineering agent societies FComputational model’s requirements: 3the individual entities must be active and autonomous; 3the individual entities need to be reactive and proactive; 3the computational entities need to be capable of interacting with entities that were not foreseen at design time; 3any organizational relationships that do exist must be reflected in the behaviour and actions of the agents (i.e., the organizational relationships must be explicitly represented).

24. Engineering agent societies F Castelfranchi claims that: -The most effective solution to the problem of social order in multi-agent systems is social modelling. -It should leave some flexibility and try to deal with emergent and spontaneous form of organizations (that is, decentralized and autonomous social control). Problem: modeling the feedback from the global results to the local/individual layer

25. Introspection and meta-reasoning for social modelling FTo solve this problem we need two ingredients : 3 introspection To dynamically change the organization, structure of the multi-agent system, agents must be aware (even if partially) of the structure and must be able to introspect about it. Introspection By using metareasoning the agent can evaluate it, obtain feedback from it and eventually try to modify it via preferences and updates in a rational way. Meta-reasoning 3metareasoning

26. Future work FThe approach can be extended in several ways: 3Dynamically reconfigurable multi-agent system. 3 Introspective and meta-reasoning abilities. 3 Other rational abilities can be incorporated, e.g., learning. 3 Proof procedure for preference reasoning to be incorporated into the current implementation of updates plus abduction.

27. Conclusions - 1 To have dynamic, flexible agent societies we need to have suitable agent theories, otherwise the structure modeling the agent society will be rigid in the sense that it will not be modifiable by the agents themselves. We believe that our theory of agents is a suitable basis for achieving this aim.

28. Conclusions - 2 FWe have presented an approach to agents that can: 3Reason and react to the environment 3Update their own knowledge 3Interact with each other by action rules that allow to express the intention to update the theory of another agent 3Decide whether to update their own knowledge depending on the requiring agent 3Prefer among possible choices

29. Future work FThe approach can be extended in several ways: 3 Agents’ language can be extended to (fully) generalized LP. 3 Agents may be given the ability to update the updating mechanism itself. 3 Other rational abilities (eg. learning) can be incorporated.