Dynamic argumentation in UbiGDSS

Abstract

Supporting and representing the group decision-making process is a complex task that requires very specific aspects. The current existing argumentation models cannot make good use of all the advantages inherent to group decision-making. There is no monitoring of the process or the possibility to provide dynamism to it. These issues can compromise the success of Group Decision Support Systems if those systems are not able to provide freedom and all necessary mechanisms to the decision-maker. We investigate the use of argumentation in a completely new perspective that will allow for a mutual understanding between agents and decision-makers. Besides this, our proposal allows to define an agent not only according to the preferences of the decisionmaker but also according to his interests towards the decision-making process. We show that our definition respects the requirements that are essential for groups to interact without limitations and that can take advantage of those interactions to create valuable knowledge to support more and better.

How cognitive and affective aspects can influence the outcome of the group decision‐making process

“…For this, in Martinho, Carneiro, Marreiros, and Novais, (), we considered five behaviour styles (Dominating, Integrating, Compromising, Obliging, and Avoiding) differentiated by four dimensions (Concern for Self, Concern for Others, Activity, and Resistance to Change). More recently, in (Carneiro et al (), we conducted a survey to find homogeneous operating values so as to numerically define the chosen behavioural styles in each dimension. These operating values are provided in Table and differentiate how agents will behave throughout the decision‐making process.…”

“…Multiagent systems have been frequently used as a tool to support group decision‐making (Aronson, Liang, & Turban, ; Carneiro, Martinho, Marreiros, & Novais, , 2015b; Carneiro, Saraiva, Martinho, Marreiros, & Novais, ; Olfati‐Saber, Fax, & Murray, ). For this type of systems, each decision maker is represented by an agent that tries to negotiate and persuade other agents to accept his opinion (Carneiro et al, ; Palomares & Martínez, ; Santos, Marreiros, Ramos, Neves, & Bulas‐Cruz, ). In our ongoing research, we developed a multiagent system adapted to solve multicriteria problems in ubiquitous scenarios.…”

“…The reasoning layer allows the participant agent to reason about received information, whether it receives a request or another kind of message. In this layer, it is considered four main components: Argumentation system—A dynamic argumentation framework has been proposed in (Carneiro et al, ; Carneiro et al, , ), which provides the system with the features that are necessary for decision makers to enjoy the benefits of group decision‐making. This framework allows agents representing decision makers, to seek a possible solution to a problem (choosing between several alternatives) while considering all the preferences of the decision makers.…”

Arguing with Behavior Influence: A Model for Web-Based Group Decision Support Systems

“…Many approaches have been put forward in literature, where agents are de¯ned with characteristics that set them apart from each other. [95][96][97][98] Also under the topic of group decision-making several works with agents have been proposed, 99,100 some of which used agents as a way to represent decision-makers/experts. 33,59,101 This representation of decision-makers allows the systems to become more intelligent and dynamic, given that they are capable of dealing with aspects of great relevance in face-to-face type meetings.…”

A Multi-agent System Framework for Dialogue Games in the Group Decision-Making Context

“…In order to empirically evaluate the proposed framework, we implemented an argumentation-based dialogue model [10,11] designed to the group decision-making context. In order to make the scenario more complex, the agents were defined with different social aspects: behavior styles, levels of expertise and credibility [12].…”