Standardizing the management and the representation of multilingual data: the MultiLingual Information Framework

The extremely fast evolution of the technological development in the sector of Communication and Information Technologies, and in particular, in the field of natural language processing, makes particularly acute the question of standardization. The issues related to this standardization are of industrial, economic and cultural nature. This article presents a methodology of standardization, in order to harmonize the management and the representation of multilingual data. Indeed, the control of the interoperability between the industrial standards currently used for localization (XLIFF)(1), translation memory (TMX)(2), or with some recent initiatives such as the internationalization tag set (ITS)(3), constitutes a major objective for a coherent and global management of these data. MLIF (Multi Lingual Information Framework)(4) is based on a methodology of standardization resulting from the ISO (sub-committees TC37/SC3 "Computer Applications for Terminology" and SC4 "Language Resources Management"). MLIF should be considered as a unified conceptual representation of multilingual content. MLIF does not have the role to substitute or to compete with any existing standard. MLIF is being designed with the objective of providing a common conceptual model and a platform allowing interoperability among several translation and localization standards, and by extension, their committed tools. The asset of MLIF is the interoperability which allows experts to gather, under the same conceptual unit, various tools and representations related to multilingual data. In addition, MLIF will also make it possible to evaluate and to compare these multilingual resources and tools.