ABSTRACT
COVID-19 caused more than 5 million deaths in the world. After lot of efforts and hard work of many scientists, few vaccines are discovered and are approved for use. It is necessary to understand and to evaluate systematically with the potential side effects due to the vaccine itself. This work proposed a sequence-to-sequence learning (Seq2Seq) model to predict the adverse effects due to COVID-19 vaccine. Seq2Seq model is used to convert sequences of one domain to another domain. In this work, a structured data such as Vaccine Adverse Event Reposting System (VAERS) data are used to predict the adverse side effects of COVID-19 vaccination. The data formulated for Seq2Seq model architecture and trying to predict the adverse side effects of vaccination with age and gender attribute as input and obtained the result of 88% as average accuracy using long short-term memory-based (LSTM) deep learning model in adverse effect prediction. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.