You could use word2vec [0] to process bodies of text in those languages, then use your C++ to build cool visualizations. The geometrical shapes that the processing finds in the languages might give you interesting insights.
This sounds interesting. I have built some tools in the past to parse for specific occurrences (determined by the text around the search criteria) to determine what might be most important to know from a given text and produced a Summary of the text.