INTERPRETACIÓN DE SENTENCIAS MATEMÁTICAS EXPRESADAS EN LENGUAJE NATURAL UTILIZANDO MÉTODOS ESTOCÁSTICOS

SÁNCHEZ JARDINES, ANDER (2019) INTERPRETACIÓN DE SENTENCIAS MATEMÁTICAS EXPRESADAS EN LENGUAJE NATURAL UTILIZANDO MÉTODOS ESTOCÁSTICOS. Maestría thesis, Universidad Autónoma de Chihuahua.

[img]
Vista Previa
Text
ANDER SÁNCHEZ.pdf

Download (2217Kb) | Vista Previa

Resumen

The purpose of this research is to interpret, in arithmetic language, mathematical sentences extracted from natural language. Mathematical sentences are compiled from oral speeches and written documents, which are stored in a linguistic Corpus. These sentences refer to the four basic arithmetic operations, addition, subtraction, multiplication, and division. For the interpretation of arithmetic operations and their correct classification, probabilistic techniques such as hidden Markov models (HMM) and N-grams were used. Before training the HMMs, the linguistic Corpus was cleaned using Levenshtein's distance. After training the models, tests with HMM were evaluated using the metrics of, F-score, Precision, Recall, and Accuracy. In the case of the N-grams, the bigrams and trigrams are determined, this in order that, given a new mathematical sentence, the N-grams help to evaluate this sentence probabilistically to determine if there are sentences in the linguistic Corpus that are similar to her and thus make the process of interpretation of the sentence lighter. The stages of development of the interpreter consist of: lexical analysis, syntactic analysis, semantic analysis, and intermediate code generation, with a visual interface that allows the user to check the result of the mathematical operation from the sentence entered

Tipo de Documento: Tesis (Maestría)
Palabras Clave: HMM (Hidden Markov Models), N-Grams Entropy, Interpret, Lexical Analysis, Syntactic Analysis, Semantic Analysis, Corpus
Clasificación temática: Q Science > QA Mathematics
Usuario Remitente: Admin Administrador del Respositorio
Depositado: 28 Jul 2020 17:38
Ultima Modificación: 28 Jul 2020 17:38
URI: http://repositorio.uach.mx/id/eprint/263

Actions (login required)

Ver Objeto Ver Objeto

Universidad Autonoma de Chihuahua