So why start a duplicate place for knowledge? Seems much simpler to integrate a translation system into the canonical English database. Existing entries can be translated into other languages and then improved upon. Hell, you could automate the entire first step and then crowd source the improvements.