Eurotra Continues

After the inevitable hesitations, the Commission of the European Communities passed the new EUROTRA budget. 10 million ECUs have been earmarked for this phase of the program, which covers the period 1991 to 1992, and is formally known as the Second Framework. It is a transitional phase, which marks a shift towards a more “open architecture” in the Community’s efforts to cultivate Europe’s language technology. The Second Framework also marks a de-emphasis on pure MT research and a correspondingly greater interest in other language technology applications. The first in a series of occasional articles on EUROTRA.

High expectations surrounded the setting up of EUROTRA in the late 1970s. Emboldened by modest success with SYSTRAN, European language technologists embarked upon this ambitious project with the hope of creating a state-of-the-art MT system for the seven — now nine — official languages of the European Community with enthusiastic support from Community officials. As time passed, expectations became tempered; illusory ideals became replaced by more pragmatic goals. Luxembourg conceded that “Fully Automatic High Quality Translation” was not a reasonably attainable goal, and the true character of EUROTRA was eventually acknowledged to be in fact pre-competitive research rather than prototype development.
    This general shift in objectives will be completed in the Third Framework (1992 to 1994). EUROTRA Project Director Serge Perschke says the Third Framework, with a budget of 22.5 million ECUs, will culminate EUROTRA’s transformation into the Linguistic Resource Engineering project. Its broad target will be cultivate a body of linguistic tools and resources which can in turn form a base for further development of applications. The new credo? REUSABILITY. In concert with a number of NLP projects — reborn cost-effective — the Third Framework will concentrate on development mono and bilingual lexica, together with sets of grammars, generic enough to be used for a variety of specific NLP purposes. Perschke envisages such applications as advanced document production aids, natural language interfaces to databases, and document indexing and abstraction systems. Another major area for future development will be “intelligent” educational applications, such as computer aided language learning programs.
    As the books close on EUROTRA, almost certainly the most ambitious MT project to date, its importance as a catalyst is finally being confirmed. EUROTRA has made a crucial long-term impact on language industries in European member states, in particular among the southern countries of Greece, Italy, Spain, and Portugal. EUROTRA has stimulated them to develop the necessary infrastructure for their own language industries. This will prove immensely valuable in years to come.