Exploring associations between urban soundscape and contextual factors based on a textual data-driven analysis and mapping approach: A case study in Daejeon, Korea

Abstract

In soundscape research, it is crucial to understand how contextual factors are associated with people’s perception of acoustic environments. This study proposes a Textual Data-driven Context Analysis and Mapping Method (TD-CAMM), aiming at comprehensively exploring the intricate relationship between soundscape and contextual factors. Through a case study employing TD-CAMM, we demonstrate its effectiveness in quantitatively analyzing text-based contextual data, facilitating the identification of patterns and interrelations among diverse contextual elements via statistical analyses. Utilizing collected contextual data including the purpose of visit, time of use, sound source, emotion, and perceived affective quality, we conducted a hierarchical cluster analysis to characterize soundscapes of the selected locations, revealing four distinct clusters. Subsequently, association rule analysis was performed to identify unique associations among contextual elements shaping the soundscapes within each cluster. Cluster 1, the 'Nightlife Zone of Noisy Urban Settings' was characterized by 'Restless' and 'Noisy' emotions associated with nighttime activities. Cluster 2, the 'Lively Human-Centered Zone for Day and Evening Activities' reflected vibrant human interactions. Cluster 3, identified as 'Restless and Noisy Urban Traffic' was marked by transportation sounds, emphasizing urban bustle. Lastly, Cluster 4, the 'Tranquil Urban Nature Walk' encompassed serene natural spaces with 'Tranquil' emotions amidst natural sounds. These findings present a comprehensive approach to evaluating urban soundscapes, with potential applications in future urban soundscape research and design.