16–17 Dec 2021 ONLINE
Évora, Portugal
Europe/Lisbon timezone

Unsupervised machine learning with Python and social discourse analysis: methodological notes on the study of sentiments of xenophobia in Brazil

Not scheduled
15m
Évora, Portugal

Évora, Portugal

Contributed Talk Using Python and Julia in Digital Humanities

Speaker

Vitor Ribeiro (ESCOLA SUPERIOR DE PROPAGANDA E MARKETING - ESPM)

Description

This abstract aims at exploring the methodological aspects of a research by the authors titled "Evolution of migration governance in Brazil: theoretical challenges, normative change and social echoes of xenophobic feelings", from early 2020. The empirical dimension of the research took advantage of algorithmic rountines of Python language code to leverage the retroactive collection of textual contents contained in Twitter repository. A total number of 2404 tweets were treated as secondary netnographic data. The selection criteria for the retrieval were both the time frame (january 2016 to january 2020), and the text expression "out Venezuelans" ("fora venezuelanos" in Portuguese). Although the sampling were not statistically representative of the broad public opinion in Brazil, the choice helped to convey the existence of a diffuse sense of xenophobia in Brazil. The analysis utilized techniques of natural language processing in Python and unsupervised machine learning algorithm word2vec, which provides a means to computationally learn semantical associations between words within a corpus of text by making use of vectorized representation of those words. As the research findings show, semantic context learnt through the parsing of text into the word2vec algorithm made possible to discern typologies of discourses either pro or against the presence of immigrants in Brazil.

MIKOLOV, Tomas et al. Distributed representations of words and phrases and their compositionality. In: INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING SYSTEMS, 26., 2013, Sydney. NIPS’13: Proceedings of the 26th International Conference on Neural Information Processing Systems - Volume 2. Nova York: Curran Associates Inc., 2013. p. 3111-3119. Available at: https://arxiv.org/pdf/1310.4546.pdf. Access on 16, October, 2021.

MIKOLOV, T.; CHEN, K.; CORRADO, G; DEAN, J. Efficient estimation of word representations in vector space. arXiv preprint arXiv:1301.3781, 2013. Available at: https://arxiv.org/pdf/1301.3781.pdf. Access on 04, May 2021.

OLIVEIRA, Antônio Tadeu Ribeiro de. Nova lei de migração: avanços, desafios e ameaças. Revista Brasileira de Estudos de População, Belo Horizonte, v. 34, n. 1, p. 171-179, jan./abr. 2017. Available at: http://www.scielo.br/pdf/rbepop/v34n1/0102-3098-rbepop-34-01-00171.pdf. Access on 16, July 2021.

SCHULTZ, Émilien; BROOKER, Phillip D.. Programming with Python for social scientists, Los Angeles, CA, Sage, 2020, 328 p. Reseaux, n. 1, p. 290-293, 2021.

UEBEL, R. R. G.; BRÍGIDO, E. V.; RIBEIRO, V. E. A. . Evolução da governança migratória no Brasil: desafios teóricos, mudanças normativas e ecos sociais de sentimentos xenofóbicos. Ideias, Campinas, SP, v. 11, p. e020009, 2020. DOI: 10.20396/ideias.v11i0.8658545. Available at: https://periodicos.sbu.unicamp.br/ojs/index.php/ideias/article/view/8658545. Acccess on: 22, November 2021.

Primary author

Vitor Ribeiro (ESCOLA SUPERIOR DE PROPAGANDA E MARKETING - ESPM)

Co-authors

Eveline Brígido (ESCOLA SUPERIOR DE PROPAGANDA E MARKETING) Roberto Uebel (ESCOLA SUPERIOR DE PROPAGANDA E MARKETING)

Presentation materials

There are no materials yet.