Text Analysis and GPT AI App for Personal Diary and Journaling Ideation
Posted Sunday, April 9, 2023 by Dmitry Paranyushkin
In this article, we will demonstrate how InfraNodus journaling app can be used for generating ideas and behavioral changes based on the personal diary and journal data. This approach can be very useful for introspection and for detecting recurring patterns of behavior. Its advantage is a combined use of text mining, data visualization, and GPT AI to derive insights from personal journals and diaries that are not available otherwise.
1. Visualizing a Diary or a Personal Journal
The first step is to get the data necessary for analysis. There are multiple ways of keeping a diary and many tools that are especially made for it. No matter what you use, you can copy and paste your journal to InfraNodus or use it for day-to-day diary writing.
Although the simplest way to keep a diary is to record events in chronological order, there are also other approaches, such as gratitude journaling, where you record things you are grateful for. We recommended, however, keeping several categories for a personal diary, separating the positive and negative experiences, and, as an option, adding a separate category for shortcomings or opportunities. This will help make the analysis more precise, revealing the gaps in positive experiences and noticing patterns in the negative ones, for instance.
Once the diary is visualized you will get an image similar to the one shown above. On the left are the statements in the chronological order, on the right is the graph that shows the main concepts used in the diary and the most prominent patterns of relations between them.
2. Finding Recurrent Patterns in Positive and Negative Experiences
When the diary content is visualized as a network graph, we can use advanced text mining methods and graph science to detect the main topical clusters and top concepts that are used in the journal.
The main topical clusters (shown with a distinct color) will reveal recurrent patterns of meaning circulation, while the most influential concepts (shown bigger on the graph) will reveal the important crossroads: terms that tend to connect different topics together. We will also use the built-in GPT AI feature of InfraNodus that interprets the graph clusters and shows the high-level ideas they represent. All this information is available in the Analytics > Topics panel.
In the journal visualization above, we see the main topics that come up in the "positive" part of the diary. If you don't separate your journal into positive and negative parts, you can also use the built-in BERT AI sentiment analysis module (available in the Analytics panel), which will tag your diary statements for you. Then you can filter the statements by the positive / negative tag and recalculate the text metrics for each type of journal entry.
For instance, in the example above, we can see that the most prominent topics are:
- • financial infrastructure, money matters, and “practical stuff”
- • spiritual and body movement practice
- • relaxation and rest
- • personal affairs
- • nature and exercises
- • caring about life
- • social interactions
- • fulfillment from work
This insight can help us see what is important for our positive experience. We can also see if there are any discrepancies between our expectations and logs. For instance, it may seem like the financial and practical matters take too much attention and perhaps their level in the realm of positive experiences could be reduced. On the other hand, if we see that nature and spiritual practices are important, we may want to increase their prevalence.
A similar approach can also be used for negative experiences: visualize the main topics, reveal recurring patterns, see how prominent some elements are, think of ways to reduce their presence in daily experience.
3. Using GPT AI and Data Visualization for Self-Transformation
A special feature of this approach is that text network visualization is used as a sort of mirror for self-reflection. We look at a representation of our thoughts, reveal the most prominent patterns, think of the different ways to amplify or reduce them.
A network representation, however, can achieve much more. Its main advantage is that it shows relations between the concepts and also — more importantly — the gaps between them. Those gaps are very interesting, because they represent the parts of our experience that belong to a positive realm, but are not yet connected. If we think of a way to link them, we may be able to produce a synergetic effect.
InfraNodus has a built-in Text Diagnostics Workflow, which you can follow to explore your ideas and thoughts. It can also be applied to personal journals and diaries. Both the Text Diagnostics Workflow and the Analytics Panel have a Structural Gap Insight feature, which detects the structural gaps in the discourse and highlights them on the graph, as shown in the image below. You can then think of a possible connection yourself or use the built-in GPT AI to generate a question or an idea for you to help you bridge that blind hole in your personal experience.
In the example above, we can see the gap is identified between the two topics:
- • Work Fulfilment
- • Social Encounter
If we think of a connection ourselves, we could imagine that, for example, working together with other people could help improve social life and also lead to more work fulfillment. On the other hand, GPT AI proposes we think about increasing the variability of our social circle as a way to enhance our capacity to produce work-related insights. Both can be interesting ideas and can be directly implemented into everyday life.
To learn more about this workflow, please, check a more extended tutorial on personal diary analysis on Nodus Labs. To try it on your own diaries and journals, please, sign up for a free account on InfraNodus.
Try It Yourself
You can analyze your own dreams with InfraNodus using this approach. Just sign up for an account or log in if you already have one:
Sign Up Log In