Usage:
uvicorn main:app --reload --port=8123 [debug]
uvicorn main:app --port=8123 [prod]
# to get response
curl --location --request GET 'http://127.0.0.1:8123/papers' \
--header 'Content-Type: text/plain' \
--data-raw '{
"n_return": 5
}'
curl --location --request GET 'http://127.0.0.1:8123/papermatch' \
--header 'Content-Type: text/plain' \
--data-raw '{
"user_string": "BERT",
"n_return": 5
}'
Arxiv-Sanity can show recent papers, this is for NBX dashboard and implments the following apis
:
API for getting latest papers
Performs tf-ids + fuzzy NN matching to get papers from the user query.
Returns a json-able dictionary for the author searched.
Args:
author (str): string shared by the user
n_return (int): number of latest papers to return.
Returns a json-able dictionary for papers in a certain category.
NOTE: The input will be the complete category string as given in the ./taxonomy
file. Eg. send {"user_string": "Artificial Intelligence"}
and not cs.AI
. Provide the category as an option to the user.
There are two files:
main.py
: The file that has the above mentioned APIsprepare.py
: prepares the dataset, expected to be run as a CRON job sometime post midnight PST, that's when the papers hit arxiv servers.