Smart Note-Taking System

I want to create an intelligent note-taking system in lecture halls. The goal of the system is to generate a summary of the lecture based on the course material in PDF format, the explanations provided by the professor during the session, and instructions (course objectives/expectations/…) all in order to make the summary more personalized and not just a simple summary. To achieve this technically, we can first do speech to text transcription into a txt document, then apply parsing, segmentation, and embedding simultaneously to it. The same process is applied to the course material in PDF format. Afterwards, we apply cosine similarity between the two results to group everything that is of the same context. The result is then provided to GPT-4, and it takes care of the synthesis part.
I don’t know if it is the right approach, if so can you provide me with some guidance and some improvement points.

Hey,
You can find it on new phones like the Samsung Galaxy S24. I’ve used it a few times and it’s pretty handy.
Are you looking for something like that? Or are you thinking of making your own?

  • Vishnu

Hey,
I am thinking of making my own.
Are you talking about Note Assist that exists in Samsung Galaxy S24 ?