Using a language independent domain model for multilingual information extraction

The volume of electronic text in different languages, particularly on the W orld Wide Web, is growing significantly, and the problem of users who are r estricted in the number of languages they read obtaining information from t his text is becoming more widespread. This article investigates some of the issues involved in achieving multilingual information extraction (IE), des cribes the approach adopted in the M-LaSIE-II IE system, which addresses th ese problems, and presents the results of evaluating the approach against a small pal allel corpus ofEnglish/French newswire texts. The approach is ba sed on the assumption that it is possible to construct a language independe nt representation of concepts relevant to the domain, at least for the smal l well-defined domains typical of IE tasks, allowing multilingual IE to be successfully carried out without requiring full machine translation.