Welcome to the forum, Gabriel, and thanks for sharing the information about your workshop. I’m quite interested in the intersection of technology and meetings, though my interest concerns more the use of technology during the actual meeting process (a simple example is what I call Session Docs). But if you see your work on natural language processing and text mining as geared towards automatically producing minutes, then the distinction becomes blurred…
Anyway, I am also gradually working my way into the world of text mining, though it will always be applied (i.e. using the methods and tools that you guys develop). Structural Topic Modelling for the analysis of open survey questions is currently on my agenda. But when I have time, I’ll have to look at your work on “extractive summarization” and how it relates to topic modelling when it comes to mapping a large corpus of text. Intuitively, my understanding is that summarization is about producing a succinct version of a single document, while topic modelling extracts topics across documents and then allows me to cluster those documents based on their similarity. But I may be babbling nonsense…