Spanish Lexical Acquisition via Morpho-Semantic Constructive Derivational Morphology
Margarita Cote González · Institutional Repositories DataBase (IRDB) · 1996
This paper describes an algorithm for Spanish derivational morphology whose output is generalizable to two different lexicon acquisition situations.One is the process of automatic lexicon acquisition via the use of Morpho-Semantic Lexical Rules (MSLRs), (Viegas, Gonzalez, & Longwell 1996) usable in semantically based Natural Language Processing (Nirenburg, et al 1996) in order to considerably reduce acquisition time per entry in a large-scale semi-automatic acquisition environment (Viegas, Onyshkevych, Raskin, & Nirenburg 1966b).The other is its application in the language classroom to fascilitate vocabulary acquisition for second language learning students.It is an ancillary tool to be used in reading or writing tasks.The constructive approach used in the design of this tool addresses the overlap between the morphological and semantic criteria while implicitly addressing morpho-phonological phenomena.The objective is accomplished by means of three major strategies: pre-categorized base stems, inherited stem changes, and eight separately identified patterns of attachment which produce noun, adjective, verb and adverb subsets.These mechanisms produce words depending on whether the word structure is right-headed or left-headed, based on the type of affix producing Morpho-Semantic Lexical Rules (MSLRs) used and on the sequence in which affix attachments succeed.The set of MSLRs, approximately 151, produces words with their respective semantic component for the morpheme or sequence of morphemes, thus defining each derivational paradigm.Their product includes not only simple and complex affixation but also word compounding, allowing also overgeneration of derived forms for broader coverage of written and oral word forms.Unnecessary overgeneration of output paradigms can be automatically checked against the contents of machine readable dictionaries and against electronic texts.The tool includes a Graphical User Interface (GUI) in order to facilitate the vocabulary acquisition process for the L2 learner of Spanish.1 The GEN_WORD Objective A dual purpose guided the design process while taking advantage of a simple yet sophisticated and productive features of language in general.• To automate the production and acquisition, via Morpho-Semantic Lexical Rules, of a large lexicon to be used by a machine translation system, using a computational semantic approach.• To automatically produce data to be accessed by means of a GUI in order to facilitate the vocabulary aquisition process for Spanish as a Second Language learners.