Inspiration
We want to leverage our experience in natural language processing and information retrieval and make tools to help everyone learn and better understand how abstract concepts mature decay, flow, and decay throughout prominent literary movements. Literature provides us with an enormous set of data within which we may expect to see a constant transmission and development of ideas--the literary output of a widely-literate society encodes much of that societies cultural values and mores. We chose to build a corpus of literature from the Gothic and Romantic movements since it includes many documents that are easily available and within the public domain, and it also marks one of the first times in English literature that the voices of women comprise a significant proportion of published thought.
What it does
Lets users make free text queries that return inferences made over an expertly-curated corpus of gothic and romantic literature. Our tool visualizes how prevalence of the queried concepts evolve throughout the movement and which documents best represent them.
How we built it
Hooking up a basic flask backend and a not-so-basic react frontend to lots of weird logic. Very fancy word embeddings with a hardcore preprocessing module. Thousands of pages of manually curated and selected literature.
Challenges we ran into
GitHub does not like us, and Pickle was out to get us.
Accomplishments that we're proud of
We had very high aspirations, and successfully achieved many of our goals and ended up with an impressive (and hopefully educational) end product!
What we learned
We learned more about text processing, forms of natural language processing, and system design.
What's next for The Necromancer
That's a secret :)
Log in or sign up for Devpost to join the conversation.