Tweet Shock: Text Mining Application for Sentiment Analysis
Keywords:
Sentiment Analysis, Text Mining, Twitter, R Studio, WEKAAbstract
This article presents modest results of an analysis process of political sentiment that was carried out, applying text mining on the Twitter accounts of representatives of the "opposition" and "official" representatives of the Madurista government in Venezuela, in the context of the 2015 Venezuelan parliamentary elections. This study was carried out with the purpose of identifying if both political groups represent two clearly differentiated classes in their linguistic discourse and in which case establish and describe their feelings, interests, concerns and preferences. To carry out this work, an adaptation of the CRISP-DM methodology was applied to complex text data known as the text mining process. For the study, the R, R Studio, TwitteR and WEKA tools were used. The results observed in the form of classification models, groups and association rules, involve the terms frequently used by the tweeters analyzed, show that there are important differences in the discursive content, very modestly the themes and feelings that characterize the two are shown groups vying for power in Venezuela. This study and its results demonstrate the feasibility and usefulness of applying text mining on social networks to perform sentiment analysis in the political arena.
Contacto del autor
Correo electrónico: liborjas@ucab.edu.ve / LivaCaro7@gmail.com
Downloads
References
A. Bifef, F. E. (2010). “Sentiment Knowledge Discovery in Twitter Streaming Data”. University of Waikato. Nueva Zelanda: Hamilton
Gentry, J. (2015, julio 29 ). Package ‘twitteR’. repositorio CRA
Gundecha, P. y. (2012). “Mining Social Media: A Brief Introduction”. Revista Informs, disponible en http://dx.doi.org/10.1287/educ.1120.0105.
K. Sylwester, M. P. (2015). “Twitter Language Use Reflects Psychological, Differences between Democrats and Republicans” . Revista PLOS One
S. Moro, M. R. (Marzo, 2014). Using Data Mining for Bank direct marketing: An application of the CRISP-DM Methodology. Repositorium, http://repositorium.sdum.uminho.pt/bitstream/1822/14838/1/MoroCortezLaureano
Wikimedia. (2014, Marzo). A diagram showing the relationship between the different phases of CRISP-DM and illustrates the recursive nature of a data mining project. Retrieved from http://commons.wikimedia.org/wiki/File:CRISP-DM_Process_Diagram.png
Zhao, Y. (2015). R and Datamining: Examples and Case Studies. In Y. Zhao, “Text Mining” (p. capítulo 10 ). http://www.RDataMining.com.
