Languages and the Computing Profession

Translating natural language is too important and complex for computing professionals to tackle alone. This essay considers the two popular approaches to machine translation of text - knowledge-based " (rule- driven) and statistical describes their weaknesses argues for the development of an unnatural language as an intermediary to make machine translation of a wealth of languages practical and reviews the implications for speech translation. An overall theme is the need for computing professionals to work in support of the appropriate professionals and the dangers of computing professionals tying to do it by themselves."