Our analysis of these studies motivated us to incorporate the
following components in our proposed framework: the development
and evaluation of a scalable and semantic-rich relation
extraction method for adverse drug event extraction and a robust
report source classification method to identify adverse drug events
based on actual patient experience.