A Study on Diet Perceptions and Trends Before and After COVID-19 Using Big Data Analysis

This study aimed to explore consumer perceptions and trends by deriving major keywords for diet before (2019) and after (2021) COVID-19 based on big data. To achieve this, data were collected by social media and text mining. In addition, semantic network, network visualization, and sentiment analyses were conducted. Text mining analysis showed keywords with the highest frequencies before and after COVID-19 were exercise, diet, start, health, and effect. TF-IDF (term frequency-inverse document frequency) analysis showed keywords with high scores before COVID-19 were lunch boxes, diary, dinner, efficacy, food, and summer and that dance, sales, and products appeared only before COVID-19. On the other hand, TF-IDF after COVID-19 showed recommendations, snacks, delicious, food, salad, care, making, and weight were used more frequently and that new words such as shake, cooking, and protein featured more after COVID-19. After COVID-19, consumers showed more interest in snacks and food that can be made and eaten rather than dietary supplements. Semantic network analysis showed that before COVID-19, preparation, management, health, and summer had high frequencies, whereas after COVID-19, keywords such as maintenance, control, weight, management, and weight loss appeared. These findings show that the purpose of diet before COVID-19 was to improve health and summer preparation, but that after COVID-19, consumer focus changed to weight loss and maintenance. The findings of this study provide fundamental data for in-depth marketing and consumer research studies for diet-related industries and academia.