BookMapp is a tool that allows researchers and students to explore large corpora of books and texts semantically and non-linearly by linking ontologies and themes. We use GPT-3 language models to analyze texts and extract themes.
The problem
You have a topic that you want to delve into. You have a set of large books.
You want find whether the answer is in the book!
Even if you find the answer, what background information do you need? Which book has it?
What relationships exist between the treatment of the topic in those books?
Just to respond to a few technical points that were raised by members of this esteemed community:
The essence of BookMapp is NOT search or query. Also, the real value comes when you have more than one book uploaded in the project.
The initial query is just to bootstrap the process. The principal value proposition is to identify similarities across the uploaded set of multiple books. Internally, we are creating an n-dimensional (where n is the number of books you uploaded) graph of connected nodes across those n dimensions. If we have, say, 5 books; we create a tensor of Rank 5, with each pair of dimensions is --roughly speaking-- a matrix of relatedness. Each of the elements of the tensor is connected using a weight or relatedness to every other node. We are doing a multi-dimensional graph traversal on each click to “More like this”.
Looking forward to more feedback, which we will surely try to answer in this forum.