We will be using this article by Andrew Sullivan from the New York Magazine: A Glimpse at the Intersectional Left’s Political Endgame
Suppose we want to get a good overview of what this article is about.
Step 2: The Analytics panel on the right will show you the main topics on the granular level of words and how they are related.
You can see that this article is talking about sex and gender as well as racial policies and racism.
Step 3: You can merge the nodes that belong to the same topical cluster by selecting them and clicking the Link button at the top right. You can read more on merging the nodes in our help center.
As you do that, the latent topics will emerge, which you can also unify into a group. In the end, you will get something like this:
A precise visual representation of the main topics present within this article.
Step 4: To simplify, we can filter the less connected nodes out of the picture using the Settings menu on the right. As a result, we get:
This article is about the use of sex and gender as well as race and ethnic difference for political campaigns and class division.
We clearly see the point of view that it's trying to carry across. We also have a quick visual summary that will always give us a good relational view of the content.
Step 5: We can then remove these topics to see the latent parts of this article's discourse with the concrete places where they appear:
Step 6: Finally, you can also use the GPT-3 based AI Summarize Graph feature to generate a text summary based on the most important topics / keywords found in text.
Try this on https://infranodus.com