Data as design tool. How understanding data as a user interface can make end-user design more accessible, efficient, effective, and embodied, while challenging machine learning conventions.

Tools

Fiebrink, Rebecca. 2015. 'Data as design tool. How understanding data as a user interface can make end-user design more accessible, efficient, effective, and embodied, while challenging machine learning conventions.'. In: International Conference on Auditory Display (ICAD) 2015. Graz, Austria 8-10 July 2015. [Conference or Workshop Item]

Preview

Text
ICAD Proceedings 15-Keynotes.pdf
Download (169kB) | Preview

Official URL: http://iem.kug.ac.at/icad15/icad15.html

Abstract or Description

We often assume "data" is something that is collected or measured from a passive source. In machine learning, we talk about "ground truth" data, because we assume the data represents something true and real; we aim to analyse and represent data appropriately, so that it will yield a window through which we can better understand some latent property of the world.

In this talk, I will describe an alternative understanding of data, in which data is something that people can actively, subjectively, and playfully manipulate. Applying modelling algorithms to intentionally manipulated data—such as examples of human movements, sounds, or social media feeds— enables everyday people to build new types of real-time interactions, including new musical instruments, sonifications, or games. In these contexts, data becomes an interface through which people communicate embodied practices, design goals, and aesthetic preferences to computers. This interface can allow people to design new real-time systems more efficiently, to explore a design space more fully, and to create systems with a particular “feel,” while also making design accessible for non-programmers.

Item Type:

Conference or Workshop Item (Keynote)