IBM Research and Johns Hopkins University have introduced a Vira chatbot that answers questions about COVID-19 vaccines in order to combat the anti-VAX movement in the United States. The data used to train the neural network includes more than 150 studies of the effectiveness and safety of vaccines.
The chatbot is available at the link VaxChat.org. Vira (Vaccine Information Resource Assistant) classifies user questions by 300 topics related to vaccination. All chatbot responses are based on evidence-based medicine. The developers plan to collect query statistics and use it to expand the topics on which Vira can support a conversation. After each response, Vira prompts the user to evaluate how useful the response was.
The chatbot, accessible via PCs and mobile devices, was designed specifically for use by young people in US communities who may find it difficult to find reliable information about vaccines. A recent study by the US Centers for Disease Control and Prevention, conducted among people aged 18 to 39 years, showed that Americans younger than 24 years old, compared with other age groups, are either less likely to be vaccinated, or do not plan to be vaccinated at all. Other groups that have expressed indecision about vaccination include people of color, people without education, without insurance and with lower income levels.
A special feature of Vira is the presence of an emotional coloring in its responses. For example, if you ask her about the potential side effects of vaccines, Vira will answer: “Although you may experience some short-term side effects when vaccinating against COVID-19, such as fatigue and nausea, this is a small price to pay to protect your loved ones.”
