An Outline for Computational Art History

The Hawaiian Star, 5930 front pages, 1893-1912 (Vimeo). Created by UCSD undergraduate student Cyrus Kiani in Manovich's visualization class, 2012.

An Outline for Computational Art History

Computational art history - use of algorithms for the analysis and visualization of patterns in art production, dissemination, reception/interaction, and scholarship. (In other words: use of computers to augment human intellect and intuition.)

Three complementary ways to think about research projects and methods in computational art history:

1 | Which stage(s) in art circulation process do you want to analyze?
- production [example 1, example 2]
- dissemination [example]
- reception (visitors movements in a museum, user tags, professional art criticism publication, user searches, etc.) [example]
- user interaction with digital media [many possibilities for capturing data about the experience; the work is co-created by software and participant]
- institutions / exhibitions / collecting / publications / art sales / art world system [example 1, example 2]
- scholarship

and also:
influence [example]
artists networks [example]
interactions with other cultural areas

2 | What data types you want to analyze?
structured, unstructured
text, images, video, 3D shapes, spaces, networks of relations, geospatial data, movements capture, interactions capture

3 | Which stage(s) in data analysis workflow you will work on?
- acquire and clean data (media agnostic)
- organize data for analysis (media agnostic)
- feature extraction (automatic) (media specific) / adding metadata (media agnostic)
- exploratory visualization (media agnostic)
[however images and video allow for unique visualization techniques: example 1, example 2]
- (optional) data analysis using classical statistics and/or data science methods (media agnostic)

Lev Manovich, february 11, 2014.