Text classifier into Goal/Idea/Plan from Introduction/Body/Conclusion.
When we read text, we often expect to understand the problem, and solution suggested, and conflate that into one text, like described here in this video, which makes text more consumable and practical right away, because there is a natural introduction, body and conclusion.
So, in order to fill up Infinity with topics, we could simply implement a parser, that, given text, is able to identify the "problem specification", "solution suggested", and "implementation steps" from a traditional "introduction-body-conclusion" type of publications.
For making reading more straightforward, we could even join them up, providing a single document with a line at the left in the color, indicating the type of content.
For example:
换一种说法: [https://wiki.mindey.com/shared/shots/1017b5309975b3acc49290e02-on-importing-web-content.mp4](https://wiki.mindey.com/shared/shots/1017b5309975b3acc49290e02-on-importing-web- content.mp4)。
进一步的想法: [https://wiki.mindey.com/shared/shots/609ea329e0f97fa49e87c96e6-text-classifier-further-thoughts.mp4](https://wiki.mindey.com/shared/shots/609ea329e0f97fa49e87c96e6-text-classifier-further- Thoughts.mp4)。
In other words: https://wiki.mindey.com/shared/shots/1017b5309975b3acc49290e02-on-importing-web-content.mp4.
Further thoughts: https://wiki.mindey.com/shared/shots/609ea329e0f97fa49e87c96e6-text-classifier-further-thoughts.mp4.