It uses the mean of 4 distance measures (perplexity, kullback-leibler, ranking based, and average cosine) to compute the distance between languages. In the corpus folder, you'll find 40 files representing 40 European languages.
- Linux, Bash, Perl
- Perl module: Math::KullbackLeibler::Discrete Alberto Simões.
It takes as input the name of one of the files in corpus/all, compute all the distances against the rest of files and gives as result the final mean. For instance, if you wish to compare the Basque lanaguage against the rest of languages, you need to run this:
sh ald.sh basque.txt
The final results are in the file mean/mean.txt.
The input language texts are in directory corpus/all