Predict chess player strength from game commentary
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
README.md
ranking.py
svm.py
util.py

README.md

Chess ratings

This code uses an SVM to predict the strength of chess players. There are three experiments implemented: classification, ranking, and regression.

You will need the following Python packages:

  • NLTK
  • sklearn
  • pgn

The chess commentary data is available from here:

http://www.ims.uni-stuttgart.de/data/chess

To start the experiment simply run:

python svm.py

The results of this experiment are described in this paper:

@InProceedings{sentimentrelevance,
    author =       {Christian Scheible and Hinrich Sch\"utze},
    title =        {Picking the Amateur's Mind - Predicting Chess Player Strength from Game Annotations },
    booktitle =    "Proceedings of Coling 2014",
    year =         2014
}