- codenames.json: three annotated games of codenames
- semantic_models.py: includes classes for Word2Vec and GloVe embeddings
- baseline_model.py: codemaster and guesser model adapted from
- Kim, A., Ruzmaykin, M., Truong, A., & Summerville, A. (2019, October). Cooperation and codenames: Understanding natural language processing via codenames. In Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (Vol. 15, No. 1, pp. 160-166).
- Download the GloVe dataset here: http://nlp.stanford.edu/data/glove.42B.300d.zip
- Unzip and put it in the codenames directory
- For faster performance, only use top 50 000 words from GloVe dataset (it's important that the finished file is named glove_short.txt)
head -n 50000 glove.42B.300d.txt > glove_short.txt
- Get file with 10,000 most common English words here: https://github.com/first20hours/google-10000-english/blob/master/google-10000-english.txt and put it in the codenames directory
- Install python packages:
- numpy
- run baseline_model.py
- codenamesexp.json: collected at ZiS games night 04/28/2021
- data_Polygon.json: collected from https://www.youtube.com/watch?v=9WipnEbVWOY&list=LL&index=8&t=2303s
- data_MCDM.json: collected from https://www.youtube.com/watch?v=7pvYL60aQZ8&list=LL&index=4 and https://www.youtube.com/watch?v=7SB7HAL7-ZU&list=LL&index=3