This is the official repository of the web frontend for the EMNLP 2020 long paper STORIUM: A Dataset and Evaluation Platform for Machine-in-the-Loop Story Generation.
What is woolgatherer? It's a webservice that partakes in woolgathering, silly! What exactly does that mean? The term woolgathering comes from people who would literally gather tufts of wool that would get snagged as animals would brush against a fence, tree, or other obstacle. By the mid-16th century the term woolgathering took on the modern day connotation: a mind wandering aimlessly, i.e. indulging in fanciful daydreaming. Thus it makes an apt name for a web service designed to generate interesting topical suggestions for story generation.
There are are currently three environments: dev, stage, & prod. Run make build-dev
or make build-prod
to create the docker containers to deploy the
service.
Afterward you can run deploy-dev
or make deploy-prod
to start the service.
You can start a shell in the container using:
docker-compose -f build/dev/docker-compose.yml run backend sh
This needs to happen before deploying the service, so that you can create and
tables or do any migrations that might be needed. Within the shell you can run
gw-createdb
create the database (if it does not already exist). If you
include the flag -t
, like gw-createdb -t
it will create the tables as well.
This is fine to do in dev, but you should not do this in production. In
production you should run alembic
migrations.
@inproceedings{akoury2020storium,
Author = {Nader Akoury, Shufan Wang, Josh Whiting, Stephen Hood, Nanyun Peng and Mohit Iyyer},
Booktitle = {Empirical Methods for Natural Language Processing},
Year = "2020",
Title = {{STORIUM}: {A} {D}ataset and {E}valuation {P}latform for {S}tory {G}eneration}
}