Abstract - In this work we shall present two software systems for automatic text analysis developed at the Institute of Computational Linguistics (ILC) in Pisa. Although originally designed for modern Spanish texts, both can be applied, with appropriate modifications, to Old Spanish. The first one, a mainframe application called Morfsin©, was released in its final version in 1986. The second, developed for the IBMcompatible PC platform, stems from the earlier work on Morfsin and is known by the acronym AyDA (Analyzer and Automatic Disambiguator). AyDA represents a useful tool for linguistics specialists in that it reduces the need to manually tag words morphologically. Moreover, it can provide quite a high degree of automatic disambiguation of functional homographs in a text. The final outcome depends on the number of words recognized, which depends, in turn, on the lexis included in the lookup dictionaries and the type of text being analyzed - modern or old, literary prose or specialized language, etc. This paper provides a description of the basic structure of the two systems, with particular emphasis on AyDA, and reports the results obtained in applying AyDA to texts in Old Spanish.
Morfsin and AyDA: two systems for analyzing modern and old Spanish
Saba A;Cappelli G
2003
Abstract
Abstract - In this work we shall present two software systems for automatic text analysis developed at the Institute of Computational Linguistics (ILC) in Pisa. Although originally designed for modern Spanish texts, both can be applied, with appropriate modifications, to Old Spanish. The first one, a mainframe application called Morfsin©, was released in its final version in 1986. The second, developed for the IBMcompatible PC platform, stems from the earlier work on Morfsin and is known by the acronym AyDA (Analyzer and Automatic Disambiguator). AyDA represents a useful tool for linguistics specialists in that it reduces the need to manually tag words morphologically. Moreover, it can provide quite a high degree of automatic disambiguation of functional homographs in a text. The final outcome depends on the number of words recognized, which depends, in turn, on the lexis included in the lookup dictionaries and the type of text being analyzed - modern or old, literary prose or specialized language, etc. This paper provides a description of the basic structure of the two systems, with particular emphasis on AyDA, and reports the results obtained in applying AyDA to texts in Old Spanish.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
