COVID-19 Early Symptom Prediction Using Blockchain and Machine Learning

Tadepalli, S. K.; Thulasiram, R. K.

The COVID-19 outbreak has resulted in unprecedented and difficult times for world’s population. Social distancing and self-isolation have become very important to reduce the spread. This called upon the creation of numerous applications that have used proprietary models for symptoms-tracking and contact-tracing around the world to mitigate the spread. In most of the applications data collected is stored in a centralized database without verification and hence, the data is not reliable. In this study, a decentralized application for COVID-19 symptoms tracking using Blockchain is proposed in order to enhance reliable data collection for training Machine Learning (ML) models. The Blockchain integration in this application will help in collecting COVID-19 symptoms data from the patients with trust. In addition to this, the data would be first verified by an entity of the decentralized network (e.g. a COVID-19 testing lab). Then, with the consent of the patient, this data is provided to the centralized system for retraining the ML. In short, the main advantage of this architecture is that the data from the users is collected and checked by a laboratory first and then provided to the ML model. The process helps in identifying the incorrect ML prediction and further train the ML model with reliable data for accurate prediction. Moreover, the trust of the users is earned as the data transfer happens with their consent and, besides, all transactions are recorded on the Blockchain, which is possible with the help of the Distributed Ledger Technology (DLT). © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.