Text mining of audio-visual descriptors for concert halls (en)
* Presenting author
Abstract:
Within the research project “Person-focused Analysis of Architectural Design”, it is a goal to elaborate a method which categorizes architectural design based on vocabulary describing audio-visual perception of built environments. The first step is to understand which dimensions are of importance to assess and verbalize the perception. Such a vocabulary can be collected by applying experimental methods such as individual vocabulary profiling. But in this case, it origins from already existing text-based sources.First, a selection of texts is performed, which is used to develop and apply a text mining methodology to extract audio-visual attributes. The machine learning methodology is explained, and the results are statistically evaluated. It is finally discussed whether text mining helps to develop a reference vocabulary, aimed to generally describe audio-visual perception. Those descriptors are to be compared to objective parameters from the acoustic and visual domain.