As is now customary, Friday was presentation day at CulturePlex, and since Dr. Sancho Caparrini is visiting us for a while, he kindly offered to share some pointers on methodology for doing Digital Humanities.
First, he said, we need to choose big questions – because “only big questions produce big ideas”. And no, big ideas cannot be attained without a solid methodology.
Once the “big question” has been asked, the following step is to delimit the “objects” to be studied and abstract the features that are relevant to the problem in question. Then, the relationships between them must be determined and all the information will need to be stored in either, relational or graph databases. Maintaining an unbiased position is key throughout this entire process.
Thirdly, says Fernando, we must query the data; and because SQL for Relational Databases can be quite challenging, Graph Queries are usually favoured.
Once the data has been queried, the answer will need to be visualized and subsequently it will need to be analyzed. And there are a few different ways for an answer to be visualized: a network or graph, a standard chart or using common network measures. Fernando also encourages new, more creative ways of visualizing answers; such as clustering (by similarity), semantic maps (for descriptors, ontology, etc.) or semantic landscapes (using brightness as height).
And lastly, comes the most important part which is interpreting the answer. This is the last and most important step and it is where the role of the humanist comes into place. It is the humanities experts, who based on their academic background will be able to draw significant conclusions from the available visualizations and analysis. It is important to ensure that the drawn conclusions do in fact, answer the initial big question.
Thank you Fernando for your presentation and for clearly showing us the importance of the humanist in the field of digital humanities.
After all, the take home message was: you are not (only) an artist … thus, providing nice visualizations in not enough.