Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 2 de 2
Filtrar
Adicionar filtros








Intervalo de ano
1.
Japanese Journal of Drug Informatics ; : 111-119, 2018.
Artigo em Japonês | WPRIM | ID: wpr-688350

RESUMO

Objective:The topic model is a well-known method used in the field of natural language processing (NLP)that defines adocument as constructed of topics that combine specific t erms. This method is used to model topic co-occurrencemathematically. In this study,we extracted topics from featu re vectors of explicit documents called medical package insertsby using cluster analysis. Methods:We counted the terms(nouns)recognized by the morphological analysis engine MeCab and created a documentterm matrix. A value of“tf・idf”was calculated in this matrix for term weighting to avoid the effect of term frequency. We reduced the dimensionality of the matrix using singular v alue decomposition,which removed unnecessary data,and weextracted feature vectors attributed to each medical package insert. The distance between feature vectors was calculatedusing cosine distance,and cluster analysis was performed based on the distance between the vectors.Results:Cluster analysis on our document-term matrix show ed that medical package inserts of drugs that have the sameefficacy or active ingredient were included in the same cl uster. Moreover, using term weighting and dimensionalityreduction,we could extract topics from medical package inserts.Conclusion:We obtained a foothold to apply our findings t o the recommendation of similar drugs. Cluster analysis ofmedical package inserts using NLP can contribute to the pro per application of drugs. In addition,our study revealed thesimilarities of drugs and suggested possibilities for new applications from several points of view.

2.
Subj. procesos cogn ; 14(2): 333-349, dic. 2010.
Artigo em Espanhol | LILACS | ID: lil-576368

RESUMO

El objetivo de este trabajo es identificar un método de evaluación automática de resúmenes realizados a partir de textos de tipo narrativo y expositivo en español. Para llevar a cabo esta tarea se correlaciona la evaluación realizada por tres docentes a 373 resúmenes con los resultados entregados por el análisis semántico latente. Los puntajes asignados por el análisis semántico latente se obtienen utilizando tres métodos 1) Comparación de los resúmenes con el texto fuente, 2) Comparación de los resúmenes con un resumen consensuado 3) Comparación de los resúmenes con tres resúmenes construidos por tres evaluadores. Entre los resultados más relevantes se destacan: a) una alta correlación entre la evaluación realizada por los evaluadores ( 0,63); b) una alta correlación entre los métodos computacionales utilizados ( 0,62) y c) una correlación promedio positiva media-alta entre las evaluaciones realizadas por los docentes y el análisis semántico latente en el segundo y tercer método ( 0,53 en ambos casos y tipos de textos). Ambos métodos presentaron mayor correlación promedio con los evaluadores cuando los textos evaluados eran predominantemente narrativos ( 0,59 y 0,45 respectivamente).


The objective of this study is to identify a method for the automatic evaluation of the summaries developed from narrative and expository Spanish texts. In order to fulfill this task evaluation of 373 summaries carried out by three teachers is correlated with the results delivered by latent semantic analysis. Scores assigned by the latent semanticanalysis are obtained through three methods: 1) Comparison of the summaries with the source text, 2) Comparison of the summaries with a consensuated one, 3) Comparison of the summaries with three summaries developed by three evaluators. The mostrelevant results include: a) a high correlation between assessments by the evaluators (:0.63), b) a high correlation between the computational methods used (:0.62) and c) a positive medium-high average correlation between assessments undertaken bythe teachers and the latent semantic analysis in the second and third method (;0.53 in both cases and types of texts). Both methods presented greater average correlation with testers when the texts evaluated were predominantly narratives (;0.59 and 0.45 respectively).


Assuntos
Narração , Processamento de Texto , Psicologia , Resumos
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA