Data Model for a Lexical Resource Based on Lexical Functions

The purpose of this paper is to describe a data model that has been used in the construction of a lexical database for Spanish. This lexical resource takes the Meaning-Text Theory (MTT) as the theoretical basis, especially lexical functions (LF), which associate a lexical unit (LU) to values that express a specific meaning related to that LF for that LU. We have developed a database, called BaDELE3000, for the 3,000 most frequently used nouns in Iberian Spanish. We followed a systematic process for the design, so that the lexical data are well structured and separated from the applications that might use them. This way, the features of the data model and the subsequent database make them useful for different purposes such as word sense disambiguation, machine translation and text generation.