Wyniki wyszukiwania - BazTech

1

A belief revision approach for argumentation-based negotiation agents

Pilotti P., Casali A., Chesñevar C.

International Journal of Applied Mathematics and Computer Science

|

2015

|

Vol. 25, no. 3

455--470

EN

Negotiation is an interaction that happens in multi-agent systems when agents have conflicting objectives and must look for an acceptable agreement. A typical negotiating situation involves two agents that cannot reach their goals by themselves because they do not have some resources they need or they do not know how to use them to reach their goals. Therefore, they must start a negotiation dialogue, taking also into account that they might have incomplete or wrong beliefs about the other agent’s goals and resources. This article presents a negotiating agent model based on argumentation, which is used by the agents to reason on how to exchange resources and knowledge in order to achieve their goals. Agents that negotiate have incomplete beliefs about the others, so that the exchange of arguments gives them information that makes it possible to update their beliefs. In order to formalize their proposals in a negotiation setting, the agents must be able to generate, select and evaluate arguments associated with such offers, updating their mental state accordingly. In our approach, we will focus on an argumentation-based negotiation model between two cooperative agents. The arguments generation and interpretation process is based on belief change operations (expansions, contractions and revisions), and the selection process is a based on a strategy. This approach is presented through a high-level algorithm implemented in logic programming. We show various theoretical properties associated with this approach, which have been formalized and proved using Coq, a formal proof management system. We also illustrate, through a case study, the applicability of our approach in order to solve a slightly modified version of the well-known home improvement agents problem. Moreover, we present various simulations that allow assessing the impact of belief revision on the negotiation process.

2

A Nonmonotonic Soft Concurrent Constraint Language to Model the Negotiation Process

Bistarelli S., Santini F.

Fundamenta Informaticae

|

2011

|

Vol. 111, nr 3

257-279

EN

We present an extension of the Soft Concurrent Constraint language that allows the nonmonotonic evolution of the constraint store. To accomplish this, we introduce some new operations: retract(c) reduces the current store by c, updateX(c) transactionally relaxes all the constraints of the store that deal with the variables in the set X, and then adds a constraint c; nask(c) tests if c is not entailed by the store. The new retraction operators also permit to reason about Belief Revision, i.e. the process of changing beliefs to take into account a new piece of information. We present this framework as a possible solution to the negotiation of resources (e.g. web services and network resource allocation) that need a given Quality of Service (QoS). For this reason we also show the the new operators of the language satisfy the Belief Revision postulates [20], which can be used in the negotiation process. The QoS requirements (expressed as semiring levels) of all the parties should converge on a formal agreement through a negotiation process, which specifies the contract that must be enforced.

3

A Framework for Iterated Belief Revision Using Possibilistic Counterparts to Jeffrey's Rule

Benferhat S., Dubois D., Prade H., Williams M-A.

Fundamenta Informaticae

|

2010

|

Vol. 99, nr 2

147-168

EN

Intelligent agents require methods to revise their epistemic state as they acquire new information. Jeffrey’s rule, which extends conditioning to probabilistic inputs, is appropriate for revising probabilistic epistemic states when new information comes in the form of a partition of events with new probabilities and has priority over prior beliefs. This paper analyses the expressive power of two possibilistic counterparts to Jeffrey's rule for modeling belief revision in intelligent agents. We show that this rule can be used to recover several existing approaches proposed in knowledge base revision, such as adjustment, natural belief revision, drastic belief revision, and the revision of an epistemic state by another epistemic state. In addition, we also show that some recent forms of revision, called improvement operators, can also be recovered in our framework.

4

The Compactness of Belief Revision and Update Operators

Liberatore P., Schaerf M.

Fundamenta Informaticae

|

2004

|

Vol. 62, nr 3,4

377--393

EN

A propositional knowledge base can be seen as a compact representation of a set of models. When a knowledge base T is updated with a formula P, the resulting set of models can be represented in two ways: either by a theory T' that is equivalent to T*P or by the pair ‹T,P›. The second representation can be super-polinomially more compact than the first. In this paper, we prove that the compactness of this representation depends on the specific semantics of *, , Winslett's semantics is more compact than Ginsberg's.