Enhancing Real Estate Predictive Models with Machine Learning for Growing Student Populations; Understanding the Role of Marketing in Shaping Perceptions of ML Technologies

Machine learning (ML) has become an increasingly powerful tool across industries, reshaping sectors from healthcare to finance, and now real estate. As ML technologies continue to evolve, they are being marketed as transformative solutions that can automate processes, improve decision-making, and unlock efficiencies previously thought unattainable. In real estate, ML is advertised as essential for staying competitive, with phrases like "AI-powered real estate" and "embracing data-driven insights" being used to frame the technology as inevitable. The portrayal of ML as a natural progression suggests that adopting it is no longer optional, and those who do not will be left behind. However, this marketing narrative often overlooks critical issues, such as algorithmic bias, data limitations, and the ethical implications of using machine learning. These concerns are frequently sidelined in favor of buzzwords and simplified promises of enhanced efficiency and personalization. My STS research seeks to critically deconstruct this narrative, examining how the discourse of inevitability surrounding ML in real estate not only oversimplifies its potential but also fails to address the risks it may pose to industry stakeholders and consumers alike. The technical portion of my project involves developing a predictive model to help address the UVA student housing crisis. By building a machine learning model that uses data from multiple sources, I aim to create a more equitable system for allocating housing resources. While my technical project may not directly relate to real estate marketing, both projects engage with similar themes about the use and portrayal of machine learning. The insights gained from my STS research have directly influenced how I plan to market the application once it is developed, ensuring that I present the tool as part of a more nuanced conversation around technology, ethics, and decision-making.

In my technical project, I developed an equitable predictive model designed to aid in the decision making process for the UVA student housing crisis. The goal was to create a system that could help predict housing needs and assist in resource allocation by analyzing multiple datasets, including historical housing data, student demographics, and occupancy trends. The model incorporated both k-means clustering and recurrent neural networks (RNNs) to analyze the data and generate insights. The ML model will be trained from a comprehensive big data set that is comprised of a variety of feature data sets which is then processed before accepted by the ML model. Ultimately, the project aimed to provide a fairer housing allocation process, ensuring that factors like socioeconomic status and student needs were incorporated into housing decisions. Although this project is focused on improving the student housing system, it also serves as a microcosm for the broader implications of machine learning applications—namely, the need for ethical considerations, transparency, and accuracy in designing and using these models.

Although not directly related to my technical project, my STS research focused on the marketing of machine learning (ML) as an inevitable, all-encompassing solution, particularly in industries like real estate. I examined how ML technologies are often presented as universal tools that promise to streamline processes and enhance efficiency. Drawing on Neeley and Luegenbiehl’s work on technological inevitability, I explored how these deterministic narratives downplay the risks associated with ML, such as algorithmic bias and the perpetuation of societal inequalities. The marketing rhetoric surrounding ML creates the illusion that its adoption is a natural, necessary step for industries, leaving little room for critical evaluation or alternative approaches. This narrative fosters optimism, assuming that all technological innovations inherently benefit society, yet it often obscures the complexities and ethical challenges that accompany such technologies. In my research, I also explored how framing ML as an inevitable solution narrows the conversation to efficiency and progress while ignoring potential unintended consequences. This perspective is particularly concerning in high-stakes sectors like real estate, where the risks of bias and exclusion can have serious societal impacts. Although ML offers significant opportunities, it requires careful consideration of its social, ethical, and practical implications. I found that this deterministic narrative, if left unchallenged, limits the scope for critical discussions and ethical scrutiny, which are essential for responsible technological adoption. The insights from my STS research directly influenced how I approach the marketing of my technical project, a predictive model for UVA’s student housing crisis. While my project focuses on improving housing allocation, it also reflects the broader need for transparent, equitable, and responsible use of ML tools. I now recognize the importance of presenting my model not as a perfect solution, but as a tool that must be used thoughtfully and critically, with an awareness of its limitations and potential risks.

School of Engineering and Applied Science

Bachelor of Science in Computer Science

Technical Advisor: Rosanne Vrugtman

STS Advisor: Kathryn A. Neeley