YouTube as a social listening tool: Mining housing discourse on YouTube

Abstract

This study examines the emerging field of social listening, focusing on analyzing public discourse around housing using unstructured social media data, specifically on YouTube in South Korea. Housing, a key asset serving both as shelter and investment, has long been a major social concern, particularly in recent discussions on social media. By exploring how housing discussions are shaped and shared on YouTube, the study aims to deepen our understanding of the public's responses to the housing markets and policies. It investigates who provides housing-related discourse on YouTube, what topics are covered, and how audiences respond to this information. The study analyzes 16,387 YouTube videos from 2011 to 2023 using keywords related to housing. A comprehensive methodological framework, including topic modeling, is proposed for systematically mining housingrelated YouTube content, comments, and metadata. Findings show that personal media channels have gained prominence, yet traditional mass media continue to significantly influence housing narratives. The content varies distinctly depending on media type and housing market contexts, while viewer engagement (likes, dislikes, comments) has notably increased. This study provides a practical methodology for utilizing social media data in urban analytics.