Data mining in tweets for analyzing dairy consumption in Brazil

Authors

  • Thallys da Silva Nogueira Universidade Federal de Juiz de Fora, Brasil https://orcid.org/0000-0002-8499-0181
  • Anna Letícia Franco Monteiro Universidade Federal de Juiz de Fora, Brasil
  • Darlan Henrique da Costa Silva Universidade Federal de Juiz de Fora, Brasil
  • Kennya Beatriz Siqueira Empresa Brasileira de Pesquisa Agropecuária, Brasil
  • Priscila Vanessa Zabala Capriles Goliatt Universidade Federal de Juiz de Fora, Brasil

DOI:

https://doi.org/10.18540/jcecvl8iss10pp14863-01a

Keywords:

Consumer. Milk and derivatives. Artificial intelligence. Social networks. Market research.

Abstract

Brazilians' daily routine was affected by the COVID-19 epidemic in a number of ways, with food being one of them. This study used data from the social network Twitter and the tool Observatório do Consumidor to examine the consumption of dairy products in Brazil in recent years. Natural language processing techniques were applied to the data to determine the verbs relating to consumption and their respective frequencies over time in order to respond to the queries "Which are the most consumed dairy products in Brazil?" and "How was this consumption over time?". It was found that the five dairy products with the largest number of consumption-related verb references were ice cream, condensed milk, cheese, dulce de leche, and milk, making them the most popular choices during the study period. However, it was noted that since 2020, dairy consumption has been declining. These findings demonstrate that it is feasible to swiftly, dynamically, and affordably assess food consumption through social networks.

Downloads

Download data is not yet available.

Published

2022-12-01

How to Cite

Nogueira, T. da S., Monteiro, A. L. F., Silva, D. H. da C., Siqueira, K. B., & Goliatt, P. V. Z. C. (2022). Data mining in tweets for analyzing dairy consumption in Brazil. The Journal of Engineering and Exact Sciences, 8(10), 14863–01a. https://doi.org/10.18540/jcecvl8iss10pp14863-01a

Issue

Section

General Articles