Spectra-based sentiment analysis

Main Article Content

José Roberto Amazonas
Alexexandre Barbosa de Lima
Marcos Roberto Pereira Barretto

Abstract

This paper introduces a method to identify the emotions conveyed by voice signals based on spectral analysis. The adopted tool is the periodogram associated to the measurement of the cumulative energy. It is shown that the variation of the emotional state produces a shift of the spectrum that can be taken as a emotional blueprint of the speaker and can be used to track his/her mood along the time.

Article Details

How to Cite
Amazonas, J. R., de Lima, A. B., & Pereira Barretto, M. R. (2018). Spectra-based sentiment analysis. INFOCOMP Journal of Computer Science, 17(2), 01–06. Retrieved from https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/571
Section
Articles