Summary
First version of the cross-lingual component, focused on shallow language-independent representation based on canonical correlation analysis.
More information & hyperlinks