Semantic Interferometry: A Complex-Valued Framework for Quantifying Non-Financial Architectural Value
M.U.J. Peeters (TU Delft - Architecture and the Built Environment)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Standard real estate valuation models (e.g., hedonic regression) rely heavily on quantitative financial metrics, failing to capture the intangible “social value” of the built environment. While Large Language Models (LLMs) can process qualitative descriptions, standard vector space models (Real Hilbert Spaces) typically utilize cosine similarity, which is additive and struggles to model “opposition” or “cancellation” of concepts effectively. This paper proposes a novel framework, Semantic Interferometry, which maps architectural descriptions into a simulated Complex Hilbert Space. By treating “Social Value” and “Exclusionary Value” as opposing phases (angles), we demonstrate how destructive interference can be used to mathematically penalize misalignment. This allows for a “Net Social Value” score where negative traits (e.g., “gated, segregated”) actively cancel out positive traits (e.g., “community, park”), providing a more rigorous, automated method for qualitative building assessment.