Exploring the Sentiment of Latvian Twitter Food Posts in Various Weather Conditions

Abstract

Food choice is a complex phenomenon influenced by factors such as taste, environment, culture, weather and many others. Although people spend most of their lives indoors, weather conditions remain influential, both in shaping seasonal food cultures in particular geographical areas and in influencing individual choices. With the recent increase in the availability of datasets on food and its perception as reflected in Twitter and historical weather data, we seek to explore food-related tweets in different weather conditions. In this paper, we examine a Latvian food tweet dataset covering the last decade in conjunction with a weather observation dataset consisting of average temperature, precipitation and other phenomena. We find out which weather conditions lead to specific food information sharing; we automatically classify tweet sentiment and discuss how it changes depending on the weather. We also explore the dynamics of sentiment related to meat and meat consumption on Twitter over a ten-year period. The rationale for focusing on tweeters’ sentiments about different meat-containing foods is due to the emergence of new discourses related to food consumption - the meat industry’s impact on planetary health, the levels of biodiversity, pollution and CO2 that influence and shape climate change, as well as the planet’s ecosystems as a whole.