Convert Mind Maps, PDFs, Websites, Texts into InfraNodus Syntax

By default, InfraNodus visualizes connections between the lemmas in your text. The lemmas are the nodes; their co-occurrences are the relations. As a result, you have a text network representation and can use advanced network science to identify the main concepts, topical clusters, and gaps in a discourse.

For instance, if we take this text from the Polysingularity blog on polarization, we will get a graph like this:

As you can see, the resulting graph is very granular, and this allows us to build very precise topical clusters, because they contain not only automatically extracted entities but also all other concepts that are used to describe them (e.g., adjectives for sentiment, auxiliary words for meaning clarification, etc). That's why we use this mode as the default one. It also suits much better for AI-generated interpretations, summaries, and research questions, as we take every little detail into account. 

...

However, there may be instances where you want to have sparser graphs and just get a general overview of relations between the main entities identified in a text. Sparser graphs are also more human-readable and carry less cognitive load.

For this, you can use the built-in entity detection in InfraNodus or our [[Wiki Links]] Chat GPT converter. What they do is extract the main entities from the text and highlight them with [[wiki link]] syntax. Therefore, only the nodes marked with the [[wiki links]] will be visualized, and the paragraphs where they appear describe the relations between them. 

The built-in InfraNodus entity detection is more detailed. For the same text, InfraNodus will, in this case, put all the main concepts found inside in [[wiki links]], so the syntax will be slightly changed, and only the concepts inside [[wiki links]] will be visualized then.

So, for instance, a sentence like "Or we can look at the stats and see that in many cases substance abuse is correlated with childhood abuse. In fact, an adult who has had adverse experiences as a child is 200% to 400% more likely to abuse alcohol later in life [1]." 

will be converted to:

Or we can look at the stats and see that in many cases [[substance abuse]] is correlated with [[childhood abuse]]. In fact, an adult who has had adverse experiences as a [[child]] is 200% to 400% more likely to abuse [[alcohol]] later in life [1].

InfraNodus identified four main entities in this text and highlighted them with the [[wiki links]], so instead of all the lemmas in these two sentences, it will only show these 4 (see the yellow cluster at the bottom left on the graph below):

The resulting graph is more sparse, and it may be easier to read it as a mind map of the main concepts. You can still apply network analysis to it, which reveals that concepts such as "complex network" and "words and beliefs" hold significant importance in this text. 

Compare it to the previous granular word analysis where it was more focused on "people" "taking" "sides." Interestingly, the first, granular, analysis relates more to how we emotionally would perceive this text, while the second, entity-based, examination reveals more scientific underpinnings of the original text. 

You can also use our Wiki Links Custom GPT for extracting the [[wiki links]] from text. As it's a different model extracting the entities here and also there is some paraphrasing going on, the resulting sentences will be rewritten. This is a more "aggressive" model, but it can work great for getting a "second opinion" on a document or a text from the AI, rather than using algorithmic entity extraction. 

As you can see, in this case, the questions of social division and war are quite prominent. The original sentence on childhood abuse was rewritten by the GPT as:

Substance abuse, for instance, correlates with [[childhood abuse]], indicating that solving root [[causes]] like [[adverse experiences]] can reduce later issues.

In this case, what the AI does is summarize repetitive points, so you get the gist of the document better. However, the original use of repetitions may have been necessary to add sentiment to the text and to really emphasize a certain point, so this analysis should be used with care and only to get a sort of "objective" look into the content.