On Automatic Composition of Stereotypic Documents in Foreign Languages

Abstract

This paper describes an interactive system that composes high quality stereotypic documents. The language for the interaction is totally independent from the target language, in which the documents are written; that is, a user can produce documents in a foreign language by interacting with the system in his language without any knowledge of the foreign language. It is also possible to produce documents in several languages simultaneously. The idea is that the system first builds, by interaction, a semantic content, which contains enough information to produce the documents. Then the system composes the document by looking at the specification file, which specifies the stereotypic document of a particular language. A new type of document or a new target language can be added to the system by simply creating a new specification file without altering the program itself. A successful pilot system has been implemented at Computer Science Department, Carnegie-Mellon University.