Overview and Summarize any Text, Book, Article or a Research Context
InfraNodus can be used to get an overview of any discourse, book, or research articles to understand the main topics within and how they are connected. You can then use this insight to create a highly pertinent summary of a text to better understand it and to develop it further.
Our approach is based on a combination of network science, data visualization, natural language processing (NLP), and AI.
1. Overview: Main Topics and Concepts
At first, your text will be visualized as a network graph and you will see the main keywords, the topics, and how they are related. Each topic is categorized using GPT AI to help you interpret the results. An example is shown on the visualization of the Ray Dalio's book "Principles" below:
InfraNodus will also reveal the insights about the structure of your discourse, showing where it belongs on the spectrum between highly biased and highly diverse. In the example above, it is "Focused" (shown at the lower part of the graph), which is a good indicator for non-fiction non-scientific works.
2. Summarize and Paraphrase
Text network representation offers an insight into the discourse structure, which can be used to retrieve excerpts from text and to generate a highly relevant summary that touches upon the most important topics. Alternatively, one can also paraphrase the document using the semantic structure identified with network analysis and visualized as a graph:
3. Exploring Hidden Relations
After getting an insight into the text's structure, it is also possible to explore relations between concrete concepts and discover in which context they appear in a text. This can be useful for discourse comparison (e.g. the use of the term "colonialism" in two different texts) or for estimating the local context around a certain topic (e.g. the concepts used with the notion of "race" in a text):
In our example, we learn that the author talks about the relation between the high debt level / increasing wealth gap and military conflicts.
4. Insights and New Ideas
Finally, you can also use the built-in AI to detect the structural gaps in this text: things that the author is talking about, but rarely in the same context. Bridging those gaps can often lead to the new ideas. We can use the built-in GPT AI to generate prompts that will link parts of the discourse we study in a highly relevant way.
5. Zooming Out: Discourse Overview
Suppose, we now want to overview scientific literature for these topics. We can use the built-in Google Scholar app and import the search results for the search query "economic cycles debt economy war". We can then compare the two discourses and see how well Ray Dalio's ideas fit into the general scientific discourse.
As we can see, there's much more emphasis on "business" and on Latin American economies in the general scientific discourse. We can also compare the two contexts together: the scientific discourse on this topic to Ray Dalio's book. As we can see, Ray Dalio, on the other hand, is putting a bigger emphasis on the power structure implications of the rising debt and wealth divide and how it's connected to monetary policy.
Try It Yourself
You can try this approach using InfraNodus with your own texts, books, scientific articles.
Sign Up
Custom Service: Text Corpus Analysis
Our methodology can be used to analyze large corpus of text-based data. Please, contact us to learn more.
Contact Us