Chimeria: Modeling Social Identity and Group Membership for Use in Digital Media such as Videogames and social media


Chimeria is a novel technology that enables the creation of more nuanced user identities in games, social media and related digital media forms. Its numerous applications in the gaming and media industries include: being used to determine dialogue and options in a computer game based upon the dynamic identity categories of characters in games to improve narrative depth, offering automatic customization of avatars based on their identities as determined by user actions in a game or preferences on social media, and providing recommendation of real/virtual products to gamers and social media users based upon the dynamic identities of their characters and/or profiles.

Problem Addressed

Computational modeling of social categories can be found in a wide range of digital media works. In computer role-playing games, racial categorization is often used to visually style a player’s avatar or trigger different canned reactions when conversing with a non-player character. In social media, users can categorize groups based on shared taste or group their acquaintances as “friends,” “family members,” etc. However, in most such systems, category membership is determined in a top-down fashion, with no possibility for hybrid identities, identities that exist at the margins of the groups, or identities that change over time. These deficiencies prevent achieving the nuance of social category membership in the real world. Chimera presents an approach in which a model of social categories can drive customized experiences in games and in social media.


Chimera supports authoring narratives of group membership in any social identity domain through a data-driven approach. It models identities (avatars, characters, profiles, and accounts) in two primary ways: (1) by modeling the underlying structure of social categories and related phenomena with the Chimeria engine, which mathematically models users’ degrees of membership across multiple categories; and (2) by enabling users to build their own creative applications involving social characterization, such as videogames and social media.

Important models are those of category membership and naturalization trajectories. In modeling categories, Chimera develops and implements a notion of abstract and concrete categories that can encompass multiple worldviews. To computationally model category gradience, Chimeria computes a closeness value corresponding to the degree to which an actor deviates from a prototypical member of a category, who is defined via a set of features. The degree of membership fluctuates throughout a narrative by actions and choices made by the user. Attributes are added/removed for discrete features (e.g. acquired skillsets) or modified numerically for continuous features (e.g. height), which creates fluctuating degree of membership and naturalization trajectory for the user.


  • Dynamically models of character identities based on real-time user decisions and preferences
  • Cross-platform implementation can alter customizing characters in games based on social media profiles and vice-versa