Developing a Reliable and Economical Web Portal for Meals on Wheels; Language Models and Online Bot Revolution

Patek, Maxwell, School of Engineering and Applied Science, University of Virginia
Seabrook, Bryn, EN-Engineering and Society, University of Virginia
Ibrahim, Ahmed, University of Virginia

Meals on Wheels Charlottesville/Albemarle is a local nonprofit organization that uses the internet for action. They provide "nutritious meals to the ill, aging, and convalescing residents of the Charlottesville/Albemarle," leveraging donations and volunteers to maintain a flexible cost structure. Meals on Wheels staff use software to organize the customers, volunteers, routes, and food, and the software is hosted on the online cloud. In order to allow the organization to grow while keeping costs low, the Meals on Wheels Charlottesville/Albemarle organization needs a new online portal. Our UVA capstone team has worked with the Meals on Wheels staff to satisfy the staff needs and to unroll the technical debt accrued by the previous two capstone teams that worked on the application.
The STS research paper explores the problem of online bots on social media and analyzes potential solutions from research scientist's perspective. How can the software engineers and research scientists that support online forums adapt to the proliferation of bots that are indistinguishable from humans without infringing on free speech and privacy? Documentary research provides a foundation for analyzing options for tackling both technical and social problems. Actor-Network analysis is an effective way to synthesize the literature and gain new insight into the intricacies of the problem space. Actor-Network analysis also provides a unique perspective on bots, since it treats both human and non-human actors as identical. Wicked problem framing is effective at extrapolating the analysis to the future of bot development and providing insights on how to adapt int the paradigm shift towards ubiquitous bots that are indistinguishable from humans. It is expected that no solution will satisfy all constraints and objectives. Therefore, a holistic approach is necessary, combining a social shift towards increased awareness from users, transparency from online platforms, regulation of those platforms, and a spectrum of choices available to users that offer a variety of freedom, privacy, and authenticity. This research will provide a groundwork for future technical, social, and regulatory work in solving the problem of online bots. The analysis of the engineering process itself will also contribute to other engineering disciplines since all engineers face similar contradictions between users' values and the health of society in general.

BS (Bachelor of Science)
language models, online bots, social media, Meals on Wheels, Web App, Django, Python, Actor Network

School of Engineering and Applied Science
Bachelor of Science in Computer Science
Technical Advisor: Ahmed Ibrahim
STS Advisor: Bryn Seabrook
Technical Team Members: Michael Benos, Alexander Hicks, Kyle Leisure, Kevin Naddoni, Joshua Santana, Nathanael Strawser

Issued Date: