Research on machine translation at the University of Saarbrücken

Research in the field of automatic analysis of language and machine translation has a long tradition at the University of Saarbrücken. In the late 1950s, a first attempt was made at the Institute of Applied Mathematics to develop a system for the automatic translation of Latin sentences (taken from a secondary school textbook) into German. In the early 1960s, a small group of researchers and students at the Institute of Applied Mathematics and the Institute of German Language and Literature, headed by Professor Hans Eggers, began to develop algorithms for the automatic syntactic analysis of a corpus of German texts, taken from newspapers and scientific textbooks (the RDE/FAZ-corpus). In the late 1960s and the early 1970s, this research group was asked by the Deutsche Forschungsgemeinschaft (DFG) to develop an automatic translation system from Russian into German on the basis of a Russian-English version of SYSTRAN. The idea of adapting this SYSTRAN version to German as a new target language was soon abandoned, and it was decided to develop an independent Russian-German MT system. This led to the foundation in 1972 of the ‘Sonderforschungsbereich 100, Elektronische Sprachforschung’ with the aim of developing the ‘Saarbrücker ÜbersetzungsSYstem’ (SUSY). In the following years, the Russian-German version of SUSY was taken as a basis for the integration of further language pairs, first French-German and in 1978 the English-German component. There were also attempts made at adapting SUSY to translation from Esperanto into German, plus German into English and French and to implement prototypes for the language pairs Danish-German and Dutch-German. The languages which, in principle, can be treated by SUSY are shown in Figure 1.