Steven Bedrick

A Multi-Lingual Web Service for Drug Side-Effect Data

Steven D Bedrick, Alejandro Mauro
AMIA Annual Symposium Proceedings, Jan 2009


In this paper, we describe a system that provides drug side-effect data for use as a component in service-oriented architectures. Our system uses ``Web 2.0''techniques to collect data from a variety of public sources, and can provide its output in a variety of human languages (e.g. Spanish and Arabic). To demonstrate our tool's versatility and the ease with which it may be integrated into larger systems, we present several front-ends that use our system, including SMS (``text message''), ``instant messenger'', and iPhone interfaces. We enlisted a panel of Argentinean clinicians to review and rate the quality of our system's Spanish-language output in order to investigate whether freely-available general-purpose machine translation technology (Google's translation API) is adequate for consumer medical applications. Our raters found that Google's translation quality varied greatly among drugs, and we conclude that it is better used as a starting point than as a complete translation solution.

Back to List