Frederico Nanni was not always a text miner. He actually started out as a historian and then switched to digital humanities. During his PhD, he developed a method to detect interdisciplinary research, based on scientific abstracts. Now, he finds text mining fascinating and thinks more historians should learn how to do it.
“My goal was to detect interdisciplinary research in a large dataset of scientific publications. So I wanted to create a method that was able to detect them, automatically, by only looking at the abstract. An abstract is like a short description of a scientific work. So we developed this method and we tested it, and the results are pretty nice.The results don’t follow our initial intuition of how to detect interdisciplinarity. At the end we ended up mapping cue words that were connected to a general idea of interdisciplinarity.
I’m originally an historian. I did a master on contemporary history. And then I switched to digital humanities during my PhD. And now I’m studying and doing interdisciplinary research, combining born-digital sources, so materials that are available online only, and analyzing them with text mining methods.
I think it’s like really cool, and really important. And a fascinating way of approaching this research topic. And I think that other people should definitely use text mining methods in their research. Test them, present the results, and evaluate the result. And especially historians. I mean, in the future we will have more and more born-digital and digitised sources. So I think it’s fundamental for us to learn how to do text mining. And learn how to evaluate our approaches, and use these methods in our research.”