model,11-1-P05-1048,ak We directly investigate a subject of much recent debate : do <term> word sense disambigation models </term> help <term> statistical machine translation quality </term> ?
hide detail