Descrição de compostos toponímicos e formalização para processamento de linguagem natural

Abstract

One challenge comes from the development of computer science is the communication of man with the machine through natural language. The developments seen in recent years has boosted the development of research that requires knowledge from several areas, including engineering and linguistics in the field of artificial intelligence, specifically in the automatic processing of natural languages. The researchers in this area have the common goal to optimize the human access to machines. In the process of understanding the structures that make natural languages and complex set of rules that govern them is essential, therefore, these formalized and encoded information can be entered in informat systems to facilitate the access of human beings to the computer. The objective of this research is the development of an electronic dictionary for the Portuguese of Brazil, specifically the description of toponymic compounds the state of Espírito Santo (Grande Vitória) frame: noun - of - noun, noun - noun and noun - adjective, from the theoretical and methodological principles of the Lexicon-Grammar developed by Maurice Gross (1975). An electronic dictionary of words to place names is a tool that will enable improvements in software that deals with natural language processing, such as programs, information retrieval, where the dictionaries help identify the function of the lexical items selected, taking into to the morphosyntactic and semantic properties of the toponymic compounds.