Joint Sentiment Topic Model for objective text clustering¹

Issue title: Special Section: Similarity, correlation and association measures - dedicated to the memory of Lotfi Zadeh

Guest editors: Ildar Batyrshin, Valerie Cross, Vladik Kreinovich and Maria Rifqi

Article type: Research Article

Authors: Sánchez, Octavio^{; *} | Sierra, Gerardo

Correspondence: [*] Corresponding author. Octavio Sánchez, Grupo de Ingeniería Lingüística, Instituto de Ingenieráa, Universidad Nacional Autónoma de México, México. E-mail: oct_sanc@unam.mx.

Note: [1] This work was possible thanks to the funding of Conacyt fellowship 387405, CoNaCyT project number 002225, DGAPA-PAPIIT projects IN403016 and IA400117.

Abstract: This article introduces a different method for text representation in order to perform clustering over different articles which, arguably, has no subjective information with similar topic-sentiment use of language. Using the joint sentiment/topic model, the text is vectorized in a low dimensional space. These vectors were then used as distance measurement for clustering texts. While comparing this unusual method with a traditional bag ofwords representation an improvement in the performance of the algorithms was observed. The authors think this method of representation might have implications for future studies of the computational interpretation of texts.

Keywords: Text representation, Joint Sentiment Topic Modeling, text clustering

DOI: 10.3233/JIFS-18530

Journal: Journal of Intelligent & Fuzzy Systems, vol. 36, no. 4, pp. 3119-3128, 2019