TransPHorm: Machine translation in public health

In the TransPHorm project, I worked with the Northwest Center for Public Health Practice and bilingual health professionals at the Washington State Department of Health, University of Washington, Public Health – Seattle & King County, and local health departments to advance the understanding of current translation processes, and to examine how novel machine translation technologies can be incorporated into public health practice.


In the US, people with limited ability to read, speak and write English face considerable difficulty accessing health information and preventative care, as well as higher care costs and poorer outcomes. Local health departments are a key part of the care infrastructure for those with limited English proficiency. I helped design and evaluate a novel web-based collaborative translation system called PHAST that links bilingual public health workers to produce multilingual health materials for less time and expense than existing translation methods. Here’s a demo.

For other facets of the study, I have run comparative experiments on post-editing (the process of correcting machine translation), as well as tested existing mobile translation technologies Google Translate and QuickSpeak with emergency responders serving populations with limited English proficiency.

Investigators: Assoc. Prof. Anne M. Turner & Research Assoc. Prof. Katrin Kirchhoff