Many public authorities strive for creating more livable cities in future. To better understand what parameters need to be optimized to achieve this, it is necessary to study how humans perceive urban spaces. A proposed technique to compare different environments is the Individual Vocabulary Profiling (IVP). This study investigates how people evaluate an urban environment auditorily and visually by means of IVP. The IVP method delivers descriptive attributes and individual preference ratings. It showcases the residential area “Driescher Hof” in Aachen, which has been presented by eight video sequences in an online experiment.  Four sequences have been captured at a marketplace, four in a park. Although geographically close, these scenes offer a multitude of influencing factors which are expected to provide different perceptual dimensions. The thesis aims to answer the following questions:

• Which environments are similarly perceived?
• Are some attributes related to one another? In that case, which are the ones that have a similar meaning or refer to the same characteristic?
• Which attributes explain the perception of each environment?

The exploration of the questions employs three methodologies:

1. Individual Vocabulary Profiling (IVP).
2. Multiple Factor Analysis (MFA) to reduce the dimensionality of the raw data and to find the most relevant factors that drive human perception in urban environments.
3. Agglomerative Hierarchical Clustering (AHC) in order to investigate the relationship between attributes and the urban stimuli under investigation.

The results provide an idea on which environments are similarly perceived and explained. The main attributes used that describe the spaces are ‘Loudness’, ‘Annoying’, ‘Broad’, ‘Tranquil’ and ‘Urban’. Some characteristics are closely related and give a description of similar places such as ‘Tranquil’ and ‘Broad’ for the park scenes. However, some drawbacks of the method could be observed. One park scene is defined as ‘Loud’, contrary to the others. When analyzing the case, it could be seen that a specific park location had a continuous drilling noise due to building work and that led to different results, which is considered as an unusual one-time event that leads to noise in the data.

Melden Sie sich hier an um Einladungen zu den Kolloquium-Vorträgen per E-Mail zu erhalten.
Register here to receive the invitations to colloquium talks via e-mail.

Zoom-Meeting-ID: 954 4073 3814
Passwort: 450783