In this project, we extracted tweets from January to April 2020 related to COVID in Australia, which resulted in 56,000 tweets. One of the key objectives was to discover the concerns of Australians and how the concerns have propagated over time. Topic modelling and Natural Language Processing techniques were used to extract this information from the tweet contents. It was observed from a global concerns level, the topics gradually transformed into more personal concerns affecting day-to-day life. Furthermore, emotions were analysed to gauge the emotional pulse of Australians over this time period. Using Emotion extraction techniques, it was observed that starting from high-intensity fear and sad, emotions have gradually settled as more positive emotions such as trust and anticipation were observed after a couple of months. A profiling exercise was carried out to profile citizens based on the concerns expressed, which demonstrated different groups of citizens with different concerns.

COVID19 Funnel Analytics