Jhon G. Botello AI, M&S and Digital libraries

AI · Modeling & simulation · Digital Libraries

Computational
Modeler

Hey there! I’m Jhon G. Botello, a Ph.D. candidate in Computer Science at Old Dominion University. I’m advancing AI to make simulations easier to build and digital libraries easier to understand.

Research affiliations Virginia Modeling, Analysis, and Simulation Center
Storymodelers
Web Science & Digital Libraries Research Group

Jhon G. Botello
Jhon G. Botello Ph.D. Candidate · Old Dominion University
Research interests

Generative and Agentic AI Modeling & Simulation Social Systems Digital Libraries

Research projects

01 / Artificial intelligence and M&S Separate model components connected into a working simulation system.

M&S 2.0: The next generation of modeling and simulation.

This project aims to advance AI methods and develop tools to make simulation models easier to build, retrieve, and reuse. It focuses on rapid simulation development, composability, and interoperability.

02 / Digital libraries
Cartoon titled The Day the Music Finally Died, showing three people on a couch puzzling over the meaning of American Pie.
Cartoon by Roz Chast · Source: CartoonStock

Bringing editorial cartoons back to life.

This project aims to develop an agentic AI framework to make editorial cartoons easier to understand for teaching and historical analysis. It connects cartoons with the events, people, and cultural references behind them.

03 / Social systems Communities connected by migration routes.

People, place,
and movement.

Studying historical human mobility and how host communities respond to migration.

Latest publications

Illustrated method: narrative input, LLM processing, and a system dynamics feedback loop

2025Conference paper

Toward Automating System Dynamics Modeling: Evaluating LLMs in the Transition from Narratives to Formal Structures

Botello, J. G., Llinas, B., Padilla, J. J., & Frydenlund, E.

Proceedings of the 2025 Winter Simulation Conference

This paper investigates how large language models (LLMs) can support the transformation of narrative descriptions into system archetypes. We tested prompting strategies including zero-shot and guided reasoning approaches, to assess LLM's ability to identify key variables, causal relationships, and feedback structures.

Artificial intelligenceModeling & simulation
View publication

Latest blog posts