This paper presents an ontology that provides the appropriate educational tools for a player of a serious game about conflict handling. The importance of this ontology lies in the fact that it promotes natural interaction (non-invasive methods) and at the same time makes the game as player-specific as it can be for its educational goal. It is an ontology that can be adapted to different educational theories and serve various educational purposes. In this ontology the Facial Expressiveness of a

more »

... layer at the moment of a game event leading to conflict as well as their Facial Expressive Response Amplification is used to determine the Intensity of the Emotional Stimulus. Once this is accomplished, the Player's Conflict Handling Model is taken into consideration in order to present the appropriate educational tool aiming to guide them towards an integrating way of dealing with conflicts in a social environment. children changing responses to conflict as they grow older [3] . This paper uses a game that encourages an integrating way of dealing with a conflict (Section 2.1). The relations between different concepts in the conflict handling and game playing domain, as well as the player's affective behavior, are represented in a multidisciplinary way, by an ontology. Researchers and developers in the fields of affective computing and conflict management/resolution will benefit from this, since it will offer access to relations and concepts from different sources. Affective analysis of the player is used in the abovementioned game in order to detect their emotional state during conflict. Conflict provokes different reactions to the participants according to their character and expressiveness ("To recognize that we are in conflict is to acknowledge that we have been triggered emotionally" [4]). The digital game industry has lately realized an important shift to Natural Interaction (NI). The keyboard and the mouse are not necessary anymore, non-verbal behavioural cues are the new -naturalmeans of interaction. Research work in the fields of psychology and cognitive science related with non-verbal behaviour and communication stress out the importance of qualitative expressive characteristics of body motion, posture, gestures and, in general, human action during an interaction session [5], [6] . In the ontology presented in this paper, visual information is used, providing important cues about conflict progress and possible subsequent reactions of the player. Within the context of computer and information sciences, an ontology defines a set of representational primitives that model a domain of knowledge or discourse [7] . According to Oberle [31], ontologies can be classified according to purpose, specificity and expressiveness, where purpose differentiates between application ontologies and reference ontologies mostly used for terminological reasons and specificity refers to generic, core and domain ontologies, with the latter being low on generality, but more specific and deeper in terms of describing a particular domain. Gruninger [32] describes the advantages of ontologies in three classes: Communication between systems, between humans, and between humans and systems; computational inference; reuse and organization of knowledge, with most developed ontologies being used to make domain assumptions explicit (70%) and enabling reuse of domain knowledge (56%) [33] . Compared to traditional approaches, ontologies provide two advantages in this framework: they help to semantically aggregate information defined in several separate descriptions and provide domain knowledge to non-experts to utilise and manipulate from their point of view. Thus, ontologies can capture a shared understanding of this domain and at the same time provide a formal and machine manipulable model for it. This is achieved by defining a vocabulary to describe the domain of interest. The constraints that describe the additional knowledge about the domain are also specified. Within this framework, ontologies are not only an efficient method for representing a domain, which in this case is the conflict handling game, but also a method for performing automated reasoning tasks to extract any required implicit knowledge. In the ontology presented in this paper, the Facial Expressiveness of a player at the moment of a game event leading to conflict, as well as their Facial Expressive Response Amplification [16] is used to determine the Intensity of the Emotional Stimulus. Once this is accomplished, the Player's Conflict Handling Model is taken into consideration in order to present the appropriate educational tool or intervention aiming to guide them towards an integrating way of dealing with conflicts in a social environment. In terms of representing the conflict domain, this ontology maps visually manifested affective cues and emotional stimuli from the serious game to conflict handling styles and proposes interventions from the part of the game; as a result, it can be used by game developers to design and implement their own conflict management scenarios or design non-player characters that illustrate prototypical behavior and respond to specific events in the game environment. The ontology can also be extended to utilise information from player models or emerging information about the player (e.g. by questionnaires or interviews before playing the game) which can help define the player's status or conflict handling style before the game experience. Thus, adaptation of the game narrative or procedural generation [34] of conflict scenarios can be employed to present a truly personalized player experience and maximize the serious game's learning potential.