Data-Driven Insights from Post-Occupancy Evaluations (POEs)

Architectural firms and institutions invest in post-occupancy evaluations (POEs) to understand how spaces function, but unstructured feedback and data sparsity limit actionable insights. Without systematic analysis, it’s difficult to link design decisions to user experience, affecting inclusivity, comfort, and interaction in built environments.

Developed an NLP and statistical analysis framework to extract key themes, sentiment, and user experience trends from POE data. Conducted topic modeling and quantitative analysis of survey responses, integrating textual and structured data to identify factors influencing user satisfaction, collaboration, and inclusivity. Led analysis of building-specific interaction patterns, uncovering disparities across user groups (e.g., space assignment and department affiliation).

Produced an evidence-based evaluation of building design effectiveness, revealing critical design-performance mismatches. Identified high-impact areas for improvement, such as collaboration space usability, comfort-driven design refinements, and interdisciplinary interaction barriers. Findings informed future architectural planning strategies, helping designers optimize spaces for usability, engagement, and inclusivity.