All entries for Wednesday 03 November 2021

Communicating computation

How do we communicate our computation? We live in a world where we can use do a lot with a little code. For example, an historian may load a set of manuscripts into the computer, use OCR (Optical Character Recognition) to extract the text, and then use NLP (Natural language processing) to identify known entities or topic modelling to look for word or phrase patterns. Each step requires decisions and interpretation drawing on the researcher's expertise and findings may need to be presented within an interpretive framework.

Combining narrative and computation is not a new idea. Knuth (1984) is credited with putting forward the literate computing paradigm. The approach focuses on developing a narrative or documentation in which code is included. In this approach an historian might detail, alongside the code for loading the manuscripts, why those manuscripts were included in the analysis and reference the source. Since 1984 there have been considerable developments in notebook technology (for an accessible overview see the slides for this talk).

Digital Humanities based training and examples are very accessible. The Programming Historian has several tutorials. Of these, Introduction to Jupyter Notebooks is a great place to start. A good source for example is the GitHub code repository because it is widely used, and you can view the text, code, and output from a Jupyter notebook. Some examples include list of Jupyter notebooks within the Digital Humanities and the Women in the Workplace project (see fig. 1), and a search of Twitter for Digital Humanities Jupyter shows some of the ways Jupyter notebooks are being used.

Notebooks, often including code written in the Python programming language, offer a medium for communicating how we compute. In this post, we briefly considered why we might use a notebook, how the idea has grown and what resources are available to learn how to use notebooks. The examples above show how notebooks are used in both teaching and research to great effect.

Our team are looking into the use of notebooks within Digital Humanities. Both of our new SRSEs have used notebooks for teaching and research. We are looking forward to supporting the use of notebooks going forward.

Fig 1. Screenshot of notebook taken from Sammantha Garcia's Women in the Workplace project showing both code and narrative cells in a Jupyter Notebook. The notebook is available at WomenInTheWorkplace/Jupyter Notebook.ipynb at main · sammanthagarcia/WomenInTheWorkplace (github.com)

James Tripp : 03 Nov 2021 09:54 |  Comments (0) |  Close comments |  Report a problem