For Cal Poly senior project, advised by Professor Rodrigo Canaan.

Building off of code and ideas from this paper.

Any documentation, code, or comments referring to "old BabelNet" are dealing with code from the above-mentioned paper, parts of which are still kept in this repo for reference.

Follow these instructions to set up the codebase locally.

Run your favorite version of the git clone command on this repo. I prefer:

git clone [email protected]:gwholland3/BabelNet-Codenames-AI.git

Some of the required data was too large to store on GitHub.

First, download the required data. The downloaded folder should be called cached_babelnet_subgraphs/. Once the folder is downloaded, move it into the babelnet_bots/data/ folder that's in your local copy of the repo.

Optionally, if you will be running the old BabelNet code, download the required data for that as well. The downloaded file should be called CodenamesData.zip. Once the file is downloaded, unzip it. Inside the CodenamesData/ folder, there are two items of interest. Move the babelnet_v6/ folder and the word_to_dict2vec_embeddings file into the babelnet_bots/data/ folder that's in your local copy of the repo. Then rename the babelnet_v6/ folder to old_cached_babelnet_data/. Feel free to delete the rest of the downloaded data, as it is not used.

This code was developed and run on Python 3.10.10, but most likely any version of Python 3.10 will do. Make sure you have an appropriate version installed locally.

I recommend doing this in a fresh Python virtual environment. Cd into the repo and run:

pip3 install -r requirements.txt

If you will be using any of the online BabelNet-scraping code or the old BabelNet code, then you will need an API key, as ordained by BabelNet themselves. Follow their instructions to register a free account and get your own API key (click the "KEY & LIMITS" tab and read the "How do I obtain a BabelNet API key?" section). Your key should look like a long hexadecimal number separated by dashes. Once you have it, copy the API key and run the following command from the root of the repo (making sure to replace {PASTE_API_KEY_HERE} with your API key):

echo -n "{PASTE_API_KEY_HERE}" > babelnet_bots/bn_api_key.txt

If you will be using any of the local indices BabelNet-scraping code, then you will need to fix a small bug in the Docker image provided by BabelNet that is used to read BabelNet data from the local indices. After installing the babelnet Python package, run the following commands from the root of the repo:

docker run -d --name babelnet-rpc babelscape/babelnet-rpc:latest
docker cp babelnet_bots/local_index.py babelnet-rpc:/root/miniconda3/envs/pybabelnet/lib/python3.8/site-packages/babelnet/indices/local_index.py
docker commit babelnet-rpc modified-babelnet-rpc
docker stop babelnet-rpc
docker rm babelnet-rpc

This will create a new docker image called modified-babelnet-rpc containing a version of the local_index.py file that fixes the bug. To use this modified image instead of the original one, add a --print flag to the end of the normal babelnet-rpc start ... command that the BabelNet instructions tell you to run. Then, take the resulting docker command you receive and change the final command-line argument from babelscape/babelnet-rpc:latest to modified-babelnet-rpc:latest before running it.

To play a sample game, run:

python3 codenames_game.py

It is preconfigured to use the BabelNet spymaster bot, but both the spymaster and field operative roles can be set to any bot that conforms to the interface specified in codenames_bots.py by changing the lines of code initializing spymaster_bot and field_operative_bot in codenames_game.py (where None signifies a human player instead of a bot).