Serious Games for Teaching Conflict Resolution: Modeling Conflict Dynamics - Conflict and Multimodal Communication: Social Research and Machine Intelligence - page 439

the stage : Assuring students that they will be listened to and not judged, and
that all the parties are equally valued, Gathering perspectives : Collecting as many
points of view as possible, Identifying interests : Using communication abilities to
determine the underlying sources of conflict, and to focus on people's interests
rather than positions, Creating options : Using creative-thinking abilities to come
up with imaginative, mutual gain solutions to conflict-related problems, Evaluating
options : Using critical-thinking abilities to apply objective criteria for determining
the suitability of a conflict resolution option, and finally, Generating agreement :
Coordinating integrated deployment between the two opposing parties, across all of
the foundation abilities.
Importantly, Bodine and Crawford's ( 1998 ) process puts young people in the
position of resolving their own conflicts. To enable young people to build up
the skills necessary to resolve conflicts, we therefore opted to design a multiplayer
game for the classroom that would create conflict situations for players to expe-
rience from a first-person perspective. Such a game would provide a bounded,
supervised opportunity for players to experiment with conflict behaviours and their
consequences.

21.2.4 Computational Models of Conflict

While the literature extensively discusses variant theoretical models of conflict
(some of which are mentioned earlier), it is rather sparse with regard to computa-
tional models of conflict. In our previous work (Cheong et al. 2011 ) we proposed a
computational model of conflict for serious games as a generic adaptive framework.
The framework consists of five processes: conflict creation, conflict detection,
player modeling, conflict management , and conflict resolution . Campos et al. ( 2013 )
present a conflict process that focuses on conflict emergence and escalation from an
emotional agent. The model accounts for an agent's decision-making process when
latent conflicts (caused by interference of incompatible goals) exist. The model was
implemented in the FAtiMA emotional agent architecture (Campos et al. 2012 ),
which in turn was used as an architecture for NPCs appearing in the serious game My
Dream Theatre .In My Dream Theatre , the player is given the task of directing a play,
which involves allocating roles to different game NPCs. Each NPC may respond to
a conflict-triggering event differently, depending on its TKI style (Thomas 1992 ).
A number of attempts have been made to model the types of conflicts found
in narrative. Swanson and Jhala ( 2012b ) present a computational model of inter-
personal and group conflict. Their model proposes cause, diagnosis, intervention,
expectation , and relationship as contextual factors that may influence the dynamics
of an ongoing conflict. Via Amazon's Mechanical Turk, Swanson and Jhala ( 2012a )
also collected a corpus of narratives recounting real conflict experiences, for the
purpose of learning about the relationships and interactions between the contexts
and conflict dynamics from the data. From the initial analysis of the corpus, they
identified six core dimensions which can categorize the parties in a dispute: active,
aggressive, interest, yielder, solver , and involvement .