This winter, I undertook a 3-month placement with the Knowledge Media Institute (KMi), which is The Open University’s Research and Development lab. My role there was as a website designer and programmer. My task? To develop a method of publishing structured data (specifically, linked data) in a user-friendly website format. I used my own research data, on the subject of portraiture and the portrait-sitting, as a case study for this work. The questions that I asked myself were (1) ‘How can I present this data in a way that is accessible to, and engaging for, art historians and the broader public?’ and (2) ‘How can I document this process in such a way that it can be used by other researchers with similar aims?’
To answer my first question, I thought about other digital resources that art historians use, for example, museums' databases, and I took advice from my placement supervisors, based on their experiences of projects including the UK Reading Experience Database, the Listening Experience Database, and SPICE (Social cohesion, Participation, and Inclusion through Cultural Engagement). This led me to design eleven web page templates and a number of website 'landing pages'. The purpose of the templates was to display information about different types of entity (for example, ‘art object’, ‘event’) in my database, while that of the landing pages was to contextualise my portrait-sitting data, facilitate browsing, and signpost additional resources.
It nevertheless remained to navigate between my research data, which took the form of densely populated spreadsheets, and the user-friendly web pages that I had designed. My supervisors and I identified the Jinja templating engine as a solution to this problem. The application that I built with Jinja is essentially a set of instructions, which determines which template should be rendered and with which data it should be populated, based on a URL path. Occasionally, these mappings between spreadsheets, application, templates and web pages highlighted errors or areas for improvement in my data, and in this sense, the placement project involved not only presenting and communicating my research data, as I had anticipated, but also, to some extent, evaluating it.
The immediate outcomes of the placement (which I have numbered to correspond with the questions above) are (1) the portrait-sitting database website and (2) the spreadsheets, templates and application that I used to produce it. However, my improved knowledge of programming languages, including SPARQL and Python, and the experience that I have gained, as a humanities student, of working in a computer sciences environment, might also be considered project ‘outcomes’. It is on this note that I wish to thank KMi and the DTP for enabling me to take part in the placement, and for their support while doing so.
