Análise do programa bag of tools na sua funcionalidade sobre conceitos matemáticos, linguísticos e computacionais

Data
2019-03-11
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal Rural do Semi-Árido

Resumo

The Bag of Tools is a program developed by the study group in computational linguistics (GELC) of the Federal Rural University of the Semi-Arid (UFERSA), which contains systems for natural language processing. In this work, I present the systems that integrate this program, wordlist: organizes the added text as a list of words; syllabic separator: system that separates words in syllables from the Brazilian language; grapheme to phoneme: phonic graphic converter that transcribes the orthographic forms into a phonetic or phonological form; and, the morphosyntactic labeler: decomposes the text into lexical items and assigns a label to each word in a text. For the development of this software, we used Foma, a library and a multi-language programming language that speeds up the creation of algorithms for the natural language processing, due to the support of regular expressions, the agility in the manipulation of characters and the ease in the use of its syntax. In order to distribute the software and facilitate its use, the program was implemented in the Java programming language, which allowing the addition of new functions to the software and the development of a graphical interface, which made the usability of the algorithm more intuitive. After the development, from the results obtained and analyzed, we conclude that the Bag of tools system presents a good performance as a whole, because in each one of its functions more than 90% of hits were obtained in its performance, but with some errors.


Descrição
Monografia
Citação
Nogueira (2019) (NOGUEIRA, 2019)