Mi SciELO
Servicios Personalizados
Articulo
Indicadores
- Citado por SciELO
Links relacionados
- Similares en SciELO
Compartir
Revista Cubana de Ciencias Informáticas
versión On-line ISSN 2227-1899
Resumen
FONSECA BRUZON, Adrian; LOPEZ LOPEZ, Aurelio y MEDINA PAGOLA, José E.. Preliminary assessment of Random Indexing variants for Text Categorization in Online Learning Context. Rev cuba cienc informat [online]. 2016, vol.10, n.1, pp. 162-171. ISSN 2227-1899.
ABSTRACT Random Indexing is a recent technique for dimensionality reduction that allows to obtain a word space model from a set of contexts. This technique is less computationally expensive in comparison with others like LSI, PLSI or LDA. These characteristics turn it an attractive prospect to be used in text categorization. In this work, we compare several variants reported in the Random Indexing literature applied to text categorization task. Experiments conducted in a subcollection of the dataset Reuter-21578 show that Random Indexing produces promising results, identifying some versions without actual advantage for the task at hand.
Palabras clave : random indexing; text categorization; dimensionality reduction.