My primary research interests lie in text and data mining historic literature, and then adding back in the results as semantic enhancements to the texts for the benefit of future researchers.
To expand on that statement, as a researcher with the Open University since 2009, my current research centres on information extraction, especially extracting implicit information. Rather than just treating text as tokens, there are typographic cues to exploit, such as change in typeface or alignment, that conventional NLP techniques discard. I am also interested in how we record the results of this information extraction as metadata, or as semantic enhancement to a document, in an effective, accessible manner, while addressing complications due to conflicting user requirements across the disciplines that want to use the common text source.
In semi-retirement now, I am free to pursue a variety of interests. Increasingly that means I am working in the Digital Humanities.
Collaborating with Francesca Benatti, the OU's Digital Humanities scholar, we are analysing early 19th Century periodicals. She provides the literary insights while I provide the computational and statistical ones. We have presented the literary aspects of our work at Romantic Voices, Oxford, 2016, and the technical aspects at the Digital Humanities Congress, Sheffield, 2016. We have been awarded the Research Society for Victorian Periodicals' Curran Field Development Grant for 2017 to pursue our work on stylometrics. We are working on the Edinburgh Review and Quarterly Review 1814-1820.
In 2016 I applied some of the knowledge acquired during my PhD, as well as my more recent NLP skills, in a project assessing post-graduate student forum posts following the introduction of a variant of authentic learning into their courses. There is the prospect of continuing this work later this year, 2017, through to 2018.
By way of interesting and stimulating variety in my work, in June and July 2017 I will return to a project on which I was a consultant back in 2011, evaluating landmark recognition skills among Yachtmaster students.
I remain a member of the Open University's Digital Humanities Steering Group.