Boguraev, B., E. Briscoe, J. Carroll and A. Copestake (1992) `Database
models for computational lexicography'. In Proceedings of EURALEX-90,
Barcelona: Biblograf/VOX. 59-78.
Also IBM T. J. Watson Research Report RC 17120, 1991.
This paper examines, from the perspective of computational lexicography/lexicology, the range of options for representation and storage of lexical data. The discussion is limited to the particular context of extraction of such data, insofar as it is of utility to natural language processing, from machine-readable dictionary sources. We thus ignore, for instance, considerations of building on-line dictionaries for human use, or designing software systems for compiling dictionary entries for inclusion in human dictionaries. Instead, we concentrate on the predominant paradigm for acquisition of lexical information and look at the suitability of existing data models for generic tasks like identification, discovery, extraction and representation of lexical properties of words on the basis of studying large dictionaries available in electronic form.