Understanding Consumer's Post COVID-19 Perceptions Towards Processed Food Using Natural Language Processing

The emergence of the COVID-19 crises created humongous panic among the general public and left a huge impact on people's cultural and behavioral aspects. In this study, we analyze how COVID-19 has impacted the perception of consumers towards processed foods. For this study, we have collected our data in two periods. In period one, we had collected our data (tweets about processed foods) from July 2019 to December 2019 (Just before the first case of COVID-19 was detected). In period two, we have collected our data from July 2020 to December 2020 (During the mid of the COVID-19 crises). In total, we have used 124,000 tweets about processed foods for this study. We have conducted our study in two parts. In part one, by comparing the polarity score of the tweets about processed food across two periods (Before COVID19 and after COVID-19), we determine consumers' attitude towards processed food across two periods and how it differs in the post-COVID-19 era. In part two, we analyze the aspects that determine consumers' attitudes towards processed foods in the postCOVID-19 era using the newly introduced Structural Topic Modeling (STM) technique. Our results indicate that the attitude of consumers towards processed foods is more polarized during the post-COVID-19 era. Freshness and Service delivery was the key aspects voiced out by consumers about processed foods in the post-COVID-19 era