Mining Twitter Data to Inform Balcony Design in Residential Architecture: A Data-Driven Approach to Human Experience in Semi-open Spaces

Abstract

This study explores public perceptions and lived experiences of balconies in Iranian residential architecture through the lens of social media. Drawing on a curated dataset of over 10,000 Persian-language tweets posted by Iranian users between 2024 and 2025, this research adopts a data-mining approach to uncover how people emotionally and functionally relate to semi-open domestic spaces—particularly balconies and terraces. Tweets were collected through keyword-based filtering and then systematically preprocessed and classified into one or more of 16 architectural and environmental design categories (e.g., spatial dimensions, materiality, greenery, privacy, safety, and spatial perception) through a multi-stage thematic coding procedure. Sentiment analysis was then applied to determine the emotional tone—positive, negative, neutral, or mixed—of each thematic instance. Findings reveal a substantial gap between architectural intent and users' lived experience of their balconies. Categories related to functional use, privacy, and spatial adequacy showed high negative polarity (77.4% negative), largely reflecting complaints about small size, lack of privacy, and impracticality. In contrast, discourse on greenery and esthetics was markedly more favorable (29.7% positive). These results demonstrate the value of user-generated content as a source of data for human-centered design: architects, landscape designers, and urban planners can draw on lived spatial experiences captured in vernacular digital archives to gain empirical insight into occupants'spatial needs. Accordingly, the study proposes revised evaluation criteria for balcony design grounded in climatic factors and social behaviors.

Received: 1 November 2025 | Revised: 22 July 2026 | Accepted: 28 August 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The supplementary datasets supporting the findings of this study are provided with the submission as Supplementary Dataset S1 (Tweets with Sentiment, bilingual) and Supplementary Dataset S2 (Validation Results, English). Additional tweet-level metadata can be made available by the corresponding author upon reasonable request, subject to ethical and platform constraints.

Author Contribution Statement

Amir Shakibamanesh: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Seyedziya Hosseinzadeh: Methodology, Validation, Formal analysis, Data curation, Writing – review & editing.