Identifying contextual factors of soundscape based on mental mapping and caption evaluation methods

Context is one of the important factors in analyzing and understanding soundscapes. Various methods for evaluating soundscapes such as soundwalks, questionnaire surveys, and narrative interviews have been proposed in ISO 12913-2. However, these methods are still limited to providing a holistic approach to identifying contexts in a place. This study aims to propose a systematic approach combining mental mapping and caption evaluation methods to identify various contexts for soundscape evaluations. The experiment consisted of two sessions. In the first session, contexts related to places were evaluated using mental mapping. Participants were asked to draw perceived landscape factors (i.e., paths, edges, and landmarks) and soundscape factors (i.e., negative or positive sounds) on the maps. Throughout this session, geospatial data of landscape and soundscape were visualized using frequency analysis and kernel density estimation. In the second session, contexts associated with people and activities are assessed using the caption evaluation method. Participants were asked to describe soundscape elements with their characteristics, impressions, and judgment. Then, the collected text data on soundscapes were analyzed through text mining techniques.