INTRODUCING PROTOTYPE FOR CLASSIFYING AND QUANTIFYING EMOTIONS IN SOCIAL NETWORK SITES

Authors

  • M. N. F. Jamaluddin Faculty of Computer and Mathematical Sciences, UniversitiTeknologi MARA, Perlis
  • S. Z. Z. Abidin AdvancedAnalytics Engineering Center (AAEC)
  • N. Omar Faculty of Computer and Mathematical Sciences, UniversitiTeknologi MARA, Selangor
  • S. S. M. Fauzi Faculty of Computer and Mathematical Sciences, UniversitiTeknologi MARA, Perlis
  • R. A. J. M. Gining Faculty of Computer and Mathematical Sciences, UniversitiTeknologi MARA, Perlis

DOI:

https://doi.org/10.4314/jfas.v10i2s.73

Keywords:

emotion classification, emotion quantification, latent semantic analysis, information retrieval, textual modality

Abstract

Usage of social network sites for dissemination of information has become common nowadays. Easy publication of opinions has led to various kind of emotionsexpressed in online discussions. In this paper, we introduce a computerized prototype which is able to classify type of emotion and quantify its degree. The prototype developed using JAVA programming language, features automatic comments extraction from social network sites and produces statistical reports on the emotions expressed. It is implemented with Latent Semantic Analysis (LSA) technique which is known to have capability of extraction and comparison of underlying semantic structure between passages. This initial work shows that the prototype can be further enhanced for better accuracy. Understanding public emotions and sentiments may alert the authority and government for ensuring safety among citizens.

Downloads

Published

2018-05-31

Issue

Section

Research Articles