New video on Data Art, here is a loose transcript with the core definition.
Data Art or data-driven art is an artistic practice that relies on the usage of a dataset to convey emotions to the audience. Because it is based on data, the piece has a more objective truth behind its construction and does not solely come from the artist's imagination.
In addition, to be considered as data art, the data used to create the piece should be somewhat understandable from the audience with proper explanations even if the meaning is not apparent at first glance.It contrasts with data visualization which seeks to be an effective representation of data into visual elements and generative art which relies on randomness, often found in nature, to elicit emotions.
Most data artworks are digital visual artifacts: images, videos, interactive pieces but data art itself is not limited to a screen or vision. For instance, a sculpture can also be parametrized by data where some of its properties (height, weight, shape, etc.) represent statistical values from a spreadsheet. Physical representations of data, also known as data physicalization, are often created with 3D printers to have a more accurate mapping.
Data can be encoded into sounds or music as well. Known as data sonification, this practice is notoriously difficult due to the opposition of the rules of music, what humans consider pleasant to the ear, and the numeric values found in datasets. To be most impactful, data sonification is used in conjunction with visual elements that describe the data as we progress through the piece.
Finally, a dataset can also be encoded into the composition of meals, drinks, or even scents showing the versatility of data art. However, due to the poor sensibility of smell and taste in humans, only a few numbers of attributes can be transformed in this manner.