Comportamento do consumidor brasileiro de alimentos na pandemia de covid-19 : um estudo via mineração de texto

Abstract

The COVID-19 pandemic, starting in Brazil in February 2020, caused one of the biggest social and economic crises ever seen in the last decade. To contain the spread of the virus, isolation, and quarantine measures were announced the closing of the commerce, which provoked one of the biggest crises in the food sector. Restaurants, bars, and snack bars saw their businesses close, and some went failed. In this context, food suffered inflation of 14.1\% in 2020, with some products such as meat, which doubled in price. With the rise in prices, there were changes in the food consumer. Therefore, this work aims to: Provide information on food consumers one year after the COVID-19 pandemic in Brazil. For this, this work will be divided into two articles where the first seeks to verify changes in food purchases and places of purchase, in addition to verifying whether there is an association between the terms of the pandemic and food in publications on Twitter. In the second article, we seek to demonstrate the correlation between the frequency of publications on pandemic food and establishments on Twitter with the actual data of pandemic in Brazil (cases, deaths, and vaccinated), in the period of the first and second waves of COVID- 19. In the first article, the research results by Crisis + food on Twitter showed that consumers are anxious and are using coffee in this period. They were confirming the association between pandemic and food on Twitter. Isolation has also led consumers in their vast majority to adopt orders for delivery, reducing the frequency of purchase from twice a week to weekly. The foods most bought by the participants were fruits, and rice and the least purchased were Meat and Soft Drinks. Candy, chocolates, and sandwiches have been associated with stress relief. Therefore, the impacts generated by COVID-19 continue to affect the eating habits of Brazilians, with the financial implications being the leading cause of this dietary change. In the second article, it was verified that the rate of COVID-19 cases in Brazil was significantly correlated with the terms "pandemic" (0.95), "restaurants" (0.80), and "chicken" (0.85). In addition to the frequency of publication of Tweets being influenced by controversial news published during this period. Thus, it concluded that one year after the beginning of the pandemic in Brazil, its impacts are still felt by Brazilians, mainly in the food field where Twitter is an environment correlated with real-life data, even when analyzed over long periods.