Towards a Sanskrit Wordparser
Andy Verboom · Literary and Linguistic Computing · 1988
When studying languages with the help of computers there are two levels to be distinguished, the sentence and the word level Since Sanskrit has a highly detailed morphological system a wordparser, (working on word level), will yield a lot of information The fact that in Sanskrit there is the Sandhi-phenomenon asks for a Sandhi-preprocessor The programming-language BOBRA has extensive pat-tern-recognition-facilities and a very compact notation, which makes it quite fit for the implementation of the formalisms of the Sanskrit wordparser and the Sandhi-preprocessor Several modules of the wordparser have been completed, some of these, the Sandhi-program and BOBRA will be dealt with in some detail in this article Conclusion the developement of the prototype of a Sanskrit wordparser and of the necessary preprocessors as the San-dhi-program is well on its way, the intermediate results being quite satisfactory The formalisms are easily and satisfactorily implemented with the programming-language BOBRA