AI-enhanced Metric Package for Assessing Reliability in Service Composition for Drug Discovery and Development

Abstract

Artificial intelligence (AI) has become more prevalent in many spheres of life, but especially in the pharmaceutical business. In the pharmaceutical sector, service composition is a key component for combining various services into a single service. We conducted how AI is being used in several areas of the pharmaceutical sector, such as clinical trials, drug repurposing, drug discovery, and development. This service increases in complexity and decreases human workload while also accomplishing targets quickly and leaving workers unable to finish their tasks if any errors happened during execution. Reliability is a key factor in successfully ensuring each service and handling this failure. We have determined from the many study studies that these aspects have not been fully measured. Our goal is to suggest a collection of metrics for gauging reliability in service composition. Their element has been recognized, described by a taxonomy, and measurements have been developed for each feature. In order to recover from failure, experiments are designed and the replication decision model, an existing recovery decision model, is also utilized. Colored Petri Net (CPN) is used to evaluate the metrics, and the related simulation results are also provided.