Date of Award

5-2026

Degree Type

Dissertation

Degree Name

Ph.D.

Degree Program

Engineering and Applied Science - Computer Science

Department

Computer Science

Major Professor

Ben Samuel

Second Advisor

Shreya Banerjee

Third Advisor

Syed Ahmed

Abstract

Creating believable, socially intelligent virtual characters that are culturally and historically grounded requires more than any single artificial intelligence paradigm can provide. Symbolic social physics systems such as Ensemble offer interpretable, human-authored representations of social norms and relationships but produce linguistically stilted interaction. Statistical approaches---particularly large language models---generate fluent dialogue but lack behavioral coherence and cultural grounding. Virtual reality provides embodied presence that transforms abstract social computation into experiential engagement but contributes no social intelligence of its own. This dissertation argues that the integration of all three technologies---symbolic AI, statistical AI, and virtual reality---is necessary for producing socially coherent, linguistically natural, and experientially compelling interactive social worlds.

To support this argument, the dissertation presents five implemented systems spanning education, entertainment, and historical research. VESPACE reconstructs 18th-century French salon culture for historical pedagogy, integrating Ensemble social physics with VR through an innovative "citation as AI" methodology in which literary historians authored thousands of social rules grounded in textual evidence. Pet Hotel addresses the authoring bottleneck by teaching social physics rule creation through progressive gameplay mechanics, enabling domain novices to learn schema definition and rule authoring through immediate feedback. Starcrossed pioneers hybrid symbolic-statistical integration by conditioning large language model dialogue generation on Ensemble social state, establishing architectural patterns for information flow between symbolic and statistical components. Insimul synthesizes three predecessor narrative AI systems---Ensemble, Talk of the Town, and Kismet---into a unified infrastructure featuring declarative Prolog-based world logic and content, hybrid AI-assisted procedural world generation, and collaborative authoring. LLVR applies this hybrid approach to second-language education in VR, integrating social physics curriculum structure, automatic speech recognition, language model dialogue, and adaptive difficulty adjustment into an immersive language learning environment.

In order to evaluate Insimul, the dissertation employs computational trace analysis, validating system intelligence by determining the accuracy of its ability to procedurally regenerate the entirety of the VESPACE corpus. For LLVR, user study results provide evidence of its linguistic assessment accuracy and automated language learning content calibration and generation abilities, which were subsequently validated by several language learning experts. In sum, the dissertation illustrates a concrete architectural pattern for the development of modular symbolic-statistical AI systems, methodologies for enabling domain expert authoring of social physics content, and demonstrated applications of hybrid AI across multiple domains.

Rights

The University of New Orleans and its agents retain the non-exclusive license to archive and make accessible this dissertation or thesis in whole or in part in all forms of media, now or hereafter known. The author retains all other ownership rights to the copyright of the thesis or dissertation.

Download

Share

COinS