I need this hybrid application ( iOS, android ) application to be able to take a picture of an ISBN number of a textbook,ebook,pdf,etc.. and grasp all of the content within it, and load it onto the app. Once the content is loaded, I want to be able to highlight a piece of text, page, chapter, unit, or even whole book, and automatically summarize all content within it , and potentially get me flashcards out of it stored in a pocket of the app.
Basically this is the editorial judgement of any ML problem, where engineering partially helps in selecting a pool of content to look at and some sample queries (in your case notes) and they do a match up, whether the query fits the content.
If we get that right for a narrow set of topics and notes, we can think of how to reproduce the same recipe to other subtopics.
Consigning content has had many solutions, some unsupervised, and some deeper but supervised, but is a second step I guess. Its a harder problem than finding content based on...