Text Classification for literature search of research study designs

Information search is used every day in many fields of the human labors. The information found is needed as evidence for making decisions. In healthcare, information has also become a necessity to find the best path to invest into a health program. Some organizations like the Health Technology Asses...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autor Principal: Galarza Quishpe, Eddie Daniel
Otros Autores: Verspoor, Karin
Formato: Tesis de Maestría
Lenguaje:eng
Publicado: Australia / Universidad de Melbourne 2016
Materias:
Acceso en línea:http://repositorio.educacionsuperior.gob.ec/handle/28000/2478
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:localhost:28000-2478
recordtype dspace
spelling oai:localhost:28000-24782017-02-20T17:16:04Z Text Classification for literature search of research study designs Galarza Quishpe, Eddie Daniel Verspoor, Karin INFORM?TICA APLICACI?N INFORM?TICA PROCESAMIENTO DE DATOS PROCESAMIENTO DE PALABRAS INVESTIGACI?N CLASIFICACI?N DIGITAL Information search is used every day in many fields of the human labors. The information found is needed as evidence for making decisions. In healthcare, information has also become a necessity to find the best path to invest into a health program. Some organizations like the Health Technology Assessment have developed alternatives to find this information provided by life sciences and biomedical records. A manual technique could give the support for finding concrete information like ?health economic evaluations? by looking at the terms contained in the records. However, this technique does not provide high levels of accuracy. In this research it is proposed an alternative to contrast manual techniques by using an automatic machine learning approach. Classification is the method from machine learning used to achieve the information search task. It has been proven that it can give acceptable levels of accuracy as well as providing proof that terms analysis can be used to determine the category of a record. Moreover, the use of a good classification algorithm and different combinations of features could give an efficient model that contrasts a manually developed technique. 2016-10-14T17:43:04Z 2016-10-14T17:43:04Z 2015-12-17 masterThesis Galarza Quishpe, Eddie Daniel. (2015). Text Classification for literature search of research study designs. (Trabajo de titulaci?n de M?ster en Tecnolog?as de la Informaci?n). Universidad de Melbourne. Australia. 36 p. http://repositorio.educacionsuperior.gob.ec/handle/28000/2478 eng 36 p. Australia / Universidad de Melbourne
institution SENESCYT
collection Repositorio SENESCYT
biblioteca Biblioteca Senescyt
language eng
format Tesis de Maestría
topic INFORM?TICA
APLICACI?N INFORM?TICA
PROCESAMIENTO DE DATOS
PROCESAMIENTO DE PALABRAS
INVESTIGACI?N
CLASIFICACI?N DIGITAL
spellingShingle INFORM?TICA
APLICACI?N INFORM?TICA
PROCESAMIENTO DE DATOS
PROCESAMIENTO DE PALABRAS
INVESTIGACI?N
CLASIFICACI?N DIGITAL
Galarza Quishpe, Eddie Daniel
Text Classification for literature search of research study designs
description Information search is used every day in many fields of the human labors. The information found is needed as evidence for making decisions. In healthcare, information has also become a necessity to find the best path to invest into a health program. Some organizations like the Health Technology Assessment have developed alternatives to find this information provided by life sciences and biomedical records. A manual technique could give the support for finding concrete information like ?health economic evaluations? by looking at the terms contained in the records. However, this technique does not provide high levels of accuracy. In this research it is proposed an alternative to contrast manual techniques by using an automatic machine learning approach. Classification is the method from machine learning used to achieve the information search task. It has been proven that it can give acceptable levels of accuracy as well as providing proof that terms analysis can be used to determine the category of a record. Moreover, the use of a good classification algorithm and different combinations of features could give an efficient model that contrasts a manually developed technique.
author2 Verspoor, Karin
author_facet Verspoor, Karin
Galarza Quishpe, Eddie Daniel
author Galarza Quishpe, Eddie Daniel
author_sort Galarza Quishpe, Eddie Daniel
title Text Classification for literature search of research study designs
title_short Text Classification for literature search of research study designs
title_full Text Classification for literature search of research study designs
title_fullStr Text Classification for literature search of research study designs
title_full_unstemmed Text Classification for literature search of research study designs
title_sort text classification for literature search of research study designs
publisher Australia / Universidad de Melbourne
publishDate 2016
url http://repositorio.educacionsuperior.gob.ec/handle/28000/2478
_version_ 1634995147091476481
score 11,871979