Using InfraNodus you can measure the urgency of any discourse by analyzing its structure.
Below we prepared a short tutorial that shows you how you can do that using the example of Twitter data during the last events: Coronavirus and BLM protests.
You will see that during the times of mobilization and crisis the network structure of a discourse tends to be focused on only a few interconnected nodes. When things settle down, the discourse structure has a more pronounced community structure and more diverse topics available.
This insight can be very useful when analyzing external data to detect "events" as well as when you are working on your own texts. Please, also read our article on Mind Viral Immunity for another discussion of this topic.
For our analysis we took the most popular Tweets during the time of Coronavirus and BLM protests.
Looking at the network structure of the discourse we realized that at the times when a topic is trending, the network structure of the discourse tends to be more homogeneous and focused on only one or two topics (e.g. "coronavirus", "police violence", "racism").
When things are more relaxed, the structure of the discourse gets more diverse and there are more distinct topics that show up in the discourse.
This insight can be used for analyzing the public discourse and especially to detect events in any field. As soon as the discourse's network structure starts to change dramatically — even if content stays the same — something may be going on.
It can also be used for text generation and speech writing: you could set the text structure objectives depending on the effect you want to have. If it's about giving an overview of a topic, then the network structure should be more diverse. If it's about focusing on one topic and mobilizing people for a collective action, then it could be more homogeneous.
An example below shows the difference between The Bible (Genesis) and poetry (poems by Ben Lerner): the both texts are visualized as a graph and sonification process was applied, so you can hear the structural differences of the graphs: