The abovementioned task addresses Multilingual Web Person Name Disambiguation (M-WePNaD).
See http://nlp.uned.es/IberEval-2017/index.php/Tasks/M-WePNaD for the task page.
In this repository you will find a Jupyter Notebook with some sample code to get you started. All we do so far is just take a first look at some of the training queries.
You can use conda or pip. To see which Python versions of various packages
were used, check out the file ./environment.yml in this repository.
You can also install the packages numpy, pandas, scikit-learn, ntlk and
jupyter by hand and hope the versions will not be too different.
One of the packages installed with conda was Jupyter Notebook. Run
it from the directory in which you have put this repository (so one
directory level above this repository, the command is jupyter notebook).
Next to this repository, the Notebook assumes you will have a directory
data, with a directory structure like this:
data/training_data/GoldStandardTraining.txt
data/training_data/MWePNaDTraining
Get these two items from the organisers of the task.
That should get you up and running.
If you have pressing questions about the task, ask them at the mailing list (Google Groups, group name: m-wepnad)!