This project creates abstract paragraphs to combine two fields of the user, one that they are interested in, and one that they are eager to learn.

Install requirements
pip install -r requirements.txt
(I might of forgotten something, if you crash add it or raise an issue)

Download data at https://webhose.io/free-datasets/popular-blog-posts/

Preprocess Data cd data python parse.py path_to_data

Train it python train.py -rnn_size 512 --data data --seq_length 512 I don't know how the best parameters, but these worked decently well

Run server on localhost:8000 python server.py

POST /search/
create user

PARAMS:
"interest"                      : String required
"learn"                         : String required

RETURNS:
"all_info"                      : JSON

based on https://github.com/nawb/chat-rnn-tensorflow