TRUSTd: Framework for Trustworthy Data in IOMT for the Judgment of Drug to Drug Interactions

TRUSTd: Framework for Trustworthy Data in IOMT for the Judgment of Drug to Drug Interactions

연구 분야: Verification

학회: International Conference on Computing, Intelligence and Data Analytics

In the Internet of Things today, data integrity is a matter of concern. This paper presents a TRUST framework that aims to make IoT systems more secure. Contrasting to the previous methods that concentrate only on the accuracy of sensors, the proposed framework judges the trustworthiness of the data shared by these devices. Analysis of time-series sensor data, we are able to detect patterns like auto-correlation and cross-correlation thereby creating the basis for machine learning models that are trained to distinguish whether the data is reliable or not. The proposed framework is illustrated through a scenario involving the design of a health-focused application in which patient data from devices like smartwatches is carefully checked for reliability. This supports healthcare providers to take rightful decisions regarding a treatment plan that may involve possible drug reactions. Finally, also to deal with sparse data, we utilized RWI which is an algorithm that improves the quality of the time-series data for the model testing. TRUST framework along with RWI-enhanced dataset with this configuration, gives a trusted environment for data integrity in IoT situations. The study not only identifies critical obstacles in the IoT environment but highlights the role of reliability and trust measurement processes as well.