TITLE: Controlled Language Authoring Technology (CLAT) Authoring Aids for Multilingual Technical Information

AUTHORS(s) & AFFILIATION(s): Johann Haller, Jörg Schütz: Institute for Applied Information Science (IAI) Martin-Luther-Str.14 D-66111 Saarbrücken, Germany

KEYWORDS: Controlled language, terminology, multilingual technical information

PRESENTER / CONTACT PERSON: Johann Haller

CONTACT EMAIL: hans@iai.uni-sb.de

ABSTRACT:

CLA Technology is a tool which deploys human language technology (HLT) in SGML/XML-based information management environments to achieve increased delivery precision, clarity and quality of the information products.

CLA Technology helps a technical writer in authoring information objects that are grammatically correct, contain consistent terminology and follow the stylistic guidelines both in general as well as in the special environment of a company. A German version is running successfully at BMW Munich and at Heidelberg Printing Machines, test versions have been distributed to a couple of other companies; an English version is tested at SUN Microsystems in Palo Alto, California. Other tools that are based on the linguistic engine of CLAT are translation aids, term mining and managing, summarization and indexing for information gisting and retrieval

They are preceded of, and based on an exhaustive linguistic analysis of the text. The text is divided into sentences and words, and a morphological and syntactic analysis is performed. This analysis shows best results in German, followed by English and French. Other languages have been treated at least in a prototypical way; Swedish is also applied on an industrial site. In the paper, we will describe the steps of the linguistic analysis, the application of the results to terminology extraction and to grammar and style control in technical texts.

Last modified October 28, 2001 by Scott Tilley.