I’ve made the focus of my talk Digital Humanities projects involving small and unusual data. What constitutes small and unusual will mean different things to different people, so keep in mind that I’ll always be speaking in relative terms. My working definition of small and unusual data will be texts and languages that are typically not used for developing and testing the tools, methods, and techniques used for Big Data analysis. I’ll be using “Big Data” as my straw man, even though most data sets in the Humanities are much smaller than those for whom the term is typically used in other fields. But I want to distinguish the types of data I will be discussing the from large corpora of hundreds or thousands of novels in Modern English which are the basis of important Digital Humanities work. I’ll also be primarily concerned with the application of machine-learning, statistical, and quantitative approaches to the analysis of unstructured texts, which forms one part of what might be called the core of activity in the Digital Humanities. But the issues I’ll be addressing overlap significantly with other DH activities such as the application of linked data and digital editing.
