Home // eKNOW 2023, The Fifteenth International Conference on Information, Process, and Knowledge Management // View article

Tools Based on Word Embedding to Make Easy the Analysis of Emotions in Spanish Text

Authors:
Jorge Silva Pedreros
Alejandra Segura-Navarrete
Christian Vidal-Castro
Claudia Martinez Araneda

Keywords: word embedding; analysis of emotions; fastText.

Abstract:
In the context of the analysis of emotions of text obtained from the web and social networks, this article aims to describe the development process of a tool to support this analysis based on the embedding of words. The main motivation has to do with the need to offer a tool for text preprocessing and emotion analysis for the Spanish language integrated into a single web application. The web application for the analysis of emotions, called fastText-embedding-viewer integrates three modules: The first corresponds to an Application Programming Interfaces (API) called corpus-preproc for text preprocessing that includes a collection of functionality, such as whitespace normalization, modifier and mark removal, lowercase folding, punctuation trimming around words and words without alphabetic characters, and content extraction main HTML and Cascading Style Sheets (CSS); the second corresponds to a novel module that allows obtaining similar words and analogies based on word embedding; and the third enables the analysis of emotions via a supervised machine learning model created using fastText for an input sentence. Promising results were obtained in the preliminary evaluation of the web application

Pages: 20 to 27

Copyright: Copyright (c) IARIA, 2023

Publication date: April 24, 2023

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-68558-082-7

Location: Venice, Italy

Dates: from April 24, 2023 to April 28, 2023