This dataset focuses on multi-lingual dependency parsing. They provide dependency parsing data for more than 10 languages, such as Spanish, Dutch, Swedish, etc.