Automatic hyphenation of Dutch words based on linguistic rules.
A.M. Nunn · 1999
This paper describes a method (reverse engineering) to improve the quality of hyphens in a dictionary database. Hyphens are recomputed with spellingbased linguistic rules. Since the input of the hyphenation program is supplied with high-quality lexicographic information including morphological makeup, good results can be obtained with a simple algorithm without compound analysis. These results could not have been achieved with earlier hyphenation programs based on word lists. The current method also has advantages over earlier hyphenation programs based on phonological syllable structure. Traditionally, the compiling of dictionaries has been the work of lexicographers who laboriously and conscientiously record words with their meaning, usage and formal features such as spelling with hyphens, inection and pronunciation. Particularly with respect to formal features, however, computers have two advantages over human editors: when provided with a correct algorithm, they can calculate the...