Using AI to Improve Price Transparency in Real Estate Valuation

Abstract

This thesis explores the integration of artificial intelligence (AI) into real estate valuation, focusing on visual property attributes to enhance traditional Hedonic models. By incorporating Vision Language Models (VLMs) and generative AI, the research evaluates the potential of these technologies to assess non-standard variables like aesthetic appeal, condition and cohesiveness of interior and exterior property photos. The study contrasts traditional hedonic regression models, which rely on quantifiable factors such as square footage and location, with a new approach that includes AI-generated scores derived from property photos. The study employs three distinct models: the No_Rubric Model, the Composite Model, and the Verbose Model with the Hedonic model serving as the baseline for evaluating their performance. The results demonstrate that incorporating visual data significantly improves model accuracy, aligning valuations more closely with buyer preferences and sold prices. This shift addresses the industry's need for price transparency and highlights how developers can design properties that better meet market demands.