Mediating Meta-Cognitive Conflicts in Group Planning Situations

AUTOR(ES)
DATA DE PUBLICAÇÃO

2001

RESUMO

The emphasis on building co-operative/collaborative environments has brought out the issue of communication (and consequently group interaction). This, in its turn, brings up the issue of conflicting interactions, inevitable whenever we are solving a problem together. If well employed, conflicts can be a tool for provoking reflection and articulation, and thus act as triggers of cognitive changes. They can also be used to avoid phenomena like groupthink and the abilene paradox and thus help to refine the groups solution to the task. On the other hand, if conflicts escalate to a personal level, they will weaken group cohesiveness, and thus probably hinder further interactions. In this thesis, we present a computational framework for detecting and mediating Meta-Cognitive conflicts (happening in a group planning interaction). The theoretical framework presented here enables a computational system to analyse the ongoing interaction, and to detect and mediate conflicts. In order to do so, we consider our model of strategic changes, a model of the group, and the history of the interaction. The objective is to suggest courses of action that provoke articulation and reflection, and thus lead to a more refined solution. To address the conflict detection issue, we have built a classification of conflicts happening in group planning interactions, that, in its turn, generated a BDI model of conflicting interactions. To address the interaction problem, we have designed a dialogue model that is based on a dialogue game approach coupled with a graphical representation of the domain. In order to diagnose which changes are happening to the group plan we have built a model of strategic operations, which describes what types of changes happen to a plan as well as how changes to one component of the strategy (beliefs, intentions, ordering relations, context beliefs and goals) impact the other ones. Moreover, to decide if, when and how to intervene, we have also considered the individual and group models (that consist of four sets of attitudes: beliefs, intentions, goals, and reflectors and are updated as the interaction evolves). We have also analysed the group social roles (part of the group model) and a model of the task. The framework has been implemented in the MArCo system, a computational environment where geographically separated groups can build common plans. In this particular case, plans are represented as PERT-CPM graphs.

ASSUNTO(S)

inteligência artificial cognitive modelling sistemas adaptativos sistemas tutores inteligentes informática na educação group planning interactions conflict interactions artificial intelligence in education ciencia da computacao

ACESSO AO ARTIGO

Documentos Relacionados