Empathy is central to social interaction, yet how it is externally expressed in virtual reality (VR) communication remains underexplored. In this study, we examined how directionality-aware cues of empathy, such as mimicry, eye contact, and body proximity, relate to cognitive and emotional empathy. We designed high- and low-empathy scenarios and recruited participants with acting experience to ensure clear emotional expressions. Our findings indicate that facial mimicry patterns differ by empathy type: cognitive empathy involves subtle, speech-related muscle movements, whereas emotional empathy is associated with more intense affective expressions. Interestingly, we also found that while facial expressions and lower-body mimicry tend to emerge unconsciously, upper-body mimicry occurs more consciously, suggesting distinct pathways of empathic embodiment. We also observed that vocal intensity mimicry and pitch variability serve as important indicators of empathy, and a consistent hand approach is closely linked to empathy. Additionally, emotional empathy fosters longer eye contact, whereas cognitive empathy stabilizes gaze and head movements. Finally, we constructed machine learning models to predict empathy from these external expressions. Our best classifier achieved an accuracy of 0.756 for cognitive empathy and 0.704 for emotional empathy, indicating the feasibility of objective assessment. These findings provide a deeper understanding of how empathy is manifested in VR communication and support the development of empathy-aware virtual agents and training systems.

In virtual reality, avatars represent ourselves and serve as a means to interact with others and the environment. Thus, understanding the self-recognition process in VR and designing self-avatars that have strong connections with the self is critical in enhancing immersion and efficiency of interaction in VR. In this paper, we investigated the characteristics of the self-recognition process in VR and the crucial factors that enhance the bond between the self and the avatar. In Study 1, we tested whether the advantage in cognitive processing of self-related information, often observed in reality, is also replicated for briefly embodied self-avatars in VR and how it is modulated by way of avatar selection and the degree of embodiment. The results showed that the self-avatar was processed more efficiently than other avatars, despite the constant changes in the appearance and a brief embodiment period. Moreover, the self-advantage effect was more pronounced for personally selected avatars, rather than those assigned by the experimenter. In Study 2, we compared the relative importance of motion and visual congruence in the process of identifying an entity as the self-avatar. The results indicated that motion synchrony between the user and the avatar is relatively more emphasized than the match in visual appearance when identifying oneself in VR. These findings highlight the underlying mechanisms and crucial factors for self-recognition in VR, and provide valuable insights for designing more immersive virtual experiences in various social VR applications.

As globalization has driven and has also been driven by innovation in information technology (IT), an effective utilization of IT has become critical in improving global operations. In this paper, we report an exploratory inquiry into what factors influence the effectiveness of IT on coordinating international operations as perceived by managers at global firms. As an initial study, our research focuses on a specific set of variables which we propose are fundamental elements for determining the IT effectiveness in the global context, i.e. firm’s globalization level, IT application, and IT department dedication. We put forth two propositions: first, a global firm can enhance its IT effectiveness by advancing the IT application level; second, as its globalization level expands, the global firm must put in place a comparable amount of organizational support (measured by IT department dedication) for the global IT system if it wants to heighten IT performance by raising the IT application level. In order to verify the propositions, we collected survey data from 35 Korean firms, each of which operates on a global scale, doing operations in foreign countries through subsidiaries. Our analysis result led us to reach conclusions upholding the propositions.

Spatial relationships have gained increasing attention in spatial databases. Reasoning spatial relationships is very expensive since it requires massive geometric computations. In particular, reading and comparing information to reason directional and topological relationships can be repeated as they are processed separately in previous works. In order to avoid the repetition and improve the performance of reasoning spatial relationships, we propose a novel metadata representation scheme, 2D+ (2 dimensional plus) string which contains direction and topology information between objects in a picture, with a generation method of the string, inference rules to reason directional and topological relationships and their application to reasoning. Our analysis of the reasoning performance using 2D+ string show that it depends on the amount of information stored in 2D+ string, which is less than that of previous works. The simplicity and expressive power are the main advantages of 2D+ string.