Publication Date
Spring 2017
Abstract
Parallel coordinate plotting is a data visualization technique that provides means for exploring multidimensional relational datasets on a two-dimensional display. Each vertical axis represents the range of values for one attribute, and each data tuple appears as a connected path traveling left-to-right across the plot, connecting attribute values for that tuple on the vertical axes. Parallel coordinate plots look like time-domain audio signal waveforms. This study investigates several timbral data sonification algorithms for classification in which audio waveforms derive from the shapes of parallel coordinate tuple plots of data being classified. Listening-response survey results and analyses reveal that mapping parallel coordinates of data tuples to audio waveforms can be accurate for generating sounds that human listeners can use to classify data. This study also investigates using machine learning algorithms to build a machine listener that approximates human survey taker performance in classifying data sounds.
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Recommended Citation
Parson, Dale E.; Malke, Wissam; Langley, Halley; and Hoch, Danielle Emily, "Mapping Data Visualization to Timbral Sonification and Machine Listening" (2017). Computer Science and Information Technology Faculty. 4.
https://research.library.kutztown.edu/cisfaculty/4
Comments
This is an open source, open access white paper reporting research on data sonification and machine listening conducted by Dr. Parson and his student coauthors in 2016 and 2017.