How to connect the dots to a research narrative?

2 minute read

I did not write a blog post for a long time. The last time I did was 13 years ago, when I went to the US for a couple of months and wanted to share my experiences with friends and family.

Now my intention is different. So far, I wrote roughly 100 papers together with various people. These are from some topics that I worked on for quite a while (biomedical information extraction, sentiment analysis, emotion analysis). Others emerged more recently (fact checking, argument mining), and some topics developed in a more general way, because they became relevant to me, independent of the concrete phenomenon (performance prediction, constraint decoding).

In any case, it’s really hard to keep track of these things and how they belong together. Sometimes, it’s really clear, however. In my recent work on appraisals and emotions, we wrote a set of papers that built clearly on top of each other. We could also have written just one pretty long paper (even longer than this one), but for some reasons we decided to write multiple papers, where each focused on quite specific research questions. That’s probably also because that’s how our research culture in computer science and natural language processing is - we write comparably short papers, but many of them.

Now this comes with a disadvantage: readers might really have a hard time to understand how these papers build on top of each other (and I might forget over time). That issue is also reinforced by the aspect that the papers are reviewed anonymously, and we cite ourselves in a way that it’s not clear that we are just the same people. My goal with these blogs now is to provide overviews of the narrative behind a sequence of research projects and papers.

I am however not yet fully convinced that blogs are the right medium to do so. They are not citable, formally, and therefore do not count towards the measures that are so important for researchers (publication count, citation count, h value, impact factors). Not that I believe that these numbers are actually important, they influence the chances to get funding for the next project. Perhaps it might be better to write an actual journal article, or just put such paper on ArXiv. I am not sure yet. I just asked via twitter what others think, let’s see what comes out of that poll.

Currently, I think I’ll start with blog posts and develop them into something “better” like a proper published paper or at least an ArXiv paper. If you have an opinion, please let me know!

Let’s see how this goes. I might also just delete this post and pretend I never had these thoughts, but I hope you’ll soon read here about some research.