Towards a frequency normalization of CREA and CORDE corpora

CORDE (Corpus Diacrónico del Español) and CREA (Corpus de Referencia del Español Actual) are two of the largest and most frequently used databases in the study of the Spanish language. However, they have some limitations in terms of size, sample unit and representativeness that may influence the res...

Full description

Saved in:
Bibliographic Details
Published inRevista signos Vol. 48; no. 89; p. 307
Main Authors Claudio Molina Salinas, Gerardo Sierra Martínez
Format Journal Article
LanguageSpanish
Published Valparaíso Dr. Giovanni Parodi 01.12.2015
Subjects
Corpus linguistics
Linguistics
Spanish language
Studies
Online AccessGet full text

Cover

More Information
Summary:CORDE (Corpus Diacrónico del Español) and CREA (Corpus de Referencia del Español Actual) are two of the largest and most frequently used databases in the study of the Spanish language. However, they have some limitations in terms of size, sample unit and representativeness that may influence the results of studies and descriptions of linguistic phenomena. In this paper we identify these limitations and propose a method for the normalization of document frequencies by computing moving averages. We show how this method allows for a more realistic interpretation of corpus data and, thus, a more effective use of these resources.
ISSN:0035-0451
0718-0934
DOI:10.4067/S0718-09342015000300002