The results of our proposed research will extend the usability of MT in healthcare and serve as a foundation for further research into improving the availability of health materials for individuals with Limited English Proficiency. Our description of public health translation work from Aim 1 will provide new understanding of existing barriers to translation. The error analysis from Aim 2 will identify specific focus areas for improving MT.
Aim 3 will provide fundamentally new MT technology designed to adapt generic systems to the health domain, as well as a prototype implementation of a domain-adapted post-processing module. The evaluation studies in Aim 4 will provide a model for evaluation of machine translation technologies and provide benchmarks from which to evaluate advances in the machine translations for health materials in the future. Ultimately, this work will advance us towards the long term goal of eliminating health disparities caused by language barriers and improve access to pertinent multilingual health information for those with limited English proficiency. Review CriteriaSignificanceInvestigator(s)InnovationApproachEnvironmentReviewer 121321Reviewer 212121Reviewer 333453
The ability to access health information in the U.S. depends greatly on the ability to speak English. Yet a growing number of people in this country speak a language other than English. We propose to develop novel domain-specific natural language processing and machine translation technology and evaluate its impact on the process of producing multilingual health materials. Ultimately, this work will advance us towards the long term goal of eliminating health disparities caused by language barriers and improve access to pertinent multilingual health information for individuals with limited English proficiency.
