Using InfraNodus, you can generate ideas using network visualization and built-in GPT AI tools from any text document. InfraNodus allows you to read the text in a nonlinear way, offering a way to "touch" the text using the interactive interface and to try out multiple perspectives.
As a result, you can get a better overview of the main topics inside the document, reveal the nuance that would not normally be available, and find the structural gaps in the content, which can help you generate new relevant ideas in relation to the topic.
You can use the following workflow, based on the cognitive variability framework implemented in InfraNodus:
1) Import an existing text or data (e.g. research paper titles)
2) Visualize and explore the text as a graph. This will help you see patterns and interesting connections.
3) Select some of the ideas that interest you, use the AI to generate connections between them
4) Use GPT AI to generate High-Level ideas (names for the topical clusters), use the graph to explore them and zoom in
5) Zoom out, look at the ideas at the periphery, try to develop them further (using research questions or AI)
6) Find the structural gaps in the graph, use the AI to bridge those gaps with research questions
7) Remove the most used concepts from the graph to reveal what's hiding behind them
Below we will look at this workflow step by step:
1. Import an Existing Text or Data
You can copy and paste a text, upload a TXT or a PDF file.
You can also import data from a scientific database, for instance, the search results for a certain topic, like "neuroplasticity" from Google Scholar or other scientific databases (PLOS, Arxiv, PubMed are supported).
In our case, we will use a research paper on
2. Visualize the Text to Reveal the Patterns
Once you import a text, you will visualize it as a graph:
The words in the documents are represented as the nodes and their co-occurrences are the connections. The words that tend to co-occur in the same context are shown near each other on the graph and have the same color (forming the topical clusters shown in the Analytics panel on the right). The most influential ideas are bigger.
Based on this representation, you can see the main terms and topical patterns in this document.
2. Explore the Graph to Reveal the Main Ideas and their Patterns
The graph shows the main ideas and patterns between them. For instance, we can see that this document is talking about the
• "production" of a "metaphor" as well as "language processing"
— all these terms are next to each other and have the same color, belonging to the same topical cluster. Among the other topics are:
• task response analysis
• left brain activation
• study fmr
Actionable insight: As you can see, just from a few seconds of looking at the graph we can see what this text is about, how various concepts are used in this text, and in what context.
We understand that the text is about the production of metaphor and language processing and how it's related to the left brain activation. We can also see how the research was performed.
3. Select Interesting Ideas, Generate an AI Summary
The next step is to select interesting ideas. For instance, we click on "left", "brain" and "activation" and then we can see in which context those ideas are used in the document, getting a better understanding of what the text is saying about these particular topics:
If you like, you can generate a summary of these ideas using the AI module (e.g. summarizing all the statements that contain these ideas):
There are two ways to do that:
1) Summarize filtered statements, containing the terms you selected on the graph (Statements Menu > AI: Summarize Visible Statements button)
2) Generate a summary based on the keywords selected (Analytics > Relations panel > AI: Generate from Related)
The first way will provide a summary of the actual statements, which can be interesting for understanding a particular text.
The second way will provide a shorter summary of the main ideas in those statements, which may be interesting for generating new research ideas.
Actionable insight: The summary generated shows the complex nature of metaphor generation, which involves various parts of the brain (mainly on the left side) responsible for different tasks: from recognizing the words and their meanings to the executive control function of the left dorsomedial prefrontal context. Stimulation of those areas may produce more "creative" metaphors.
4. Use GPT to Reveal High-Level Ideas
You can use the built-in GPT AI to reveal high-level ideas in order to better understand the general discourse inside the document:
Based on this representation, you can zoom out, and see the main topics that are contained in this document. While we know most of them, it's quite interesting to find the one that is talking about "semantic memory". We could zoom into this topic, following the same process as described in the previous state, in order to better understand what it's talking about.