Congress detail

Authors: Pablo Pilotti; Ana Casali; Chesñevar, Carlos Iván.

Resumen: A typical negotiation situation involves two cooperative agents that own objects and have the need of exchanging them in order to achieve their goals. Usually, objects have different characteristics or features that are key to allowing or not an agent to achieve his goal. Most of the existing approaches assume that the agents have complete information about all objects features. In this paper, we will assume instead that agents have only complete information about their objects and its features, but they have beliefs, possibly erroneous or incomplete, about the objects of his counterpart. To agree on which objects are to be exchanged, agents establish a negotiating dialog where each agent offers a proposal for possible exchanges and after the first proposal, may respond with a critique and a counterproposal until the negotiation ends. During the dialog, the agents use arguments to reveal information about the objects they hold and their features. This paper presents an automatic negotiation model based on argumentation for agents with incomplete information. Belief revision techniques are used for interpreting and generating proposals, and Defeasible Logic Programming is used as underlying knowledge representation language.

Meeting type: Conferencia.

Type of job: Artículo Completo.

Production: Incorporating Object Features in Collaborative Argumentation-Based Negotiation Agents.

Scientific meeting: Brazilian Conference on Intelligent Systems (BRACIS).

Meeting place: Sao Carlos, Brasil.

Organizing Institution: SBIA.

It's published?: Yes

Publication place: Sao Pablo

Meeting month: 10

Year: 2014.