PHAST: A Collaborative Machine Translation and Post-Editing Tool for Public Health.

Area of Science:

• Public Health Informatics
• Computational Linguistics
• Health Communication

Background:

• Producing multilingual public health materials is crucial but faces significant time and cost barriers.
• Existing machine translation (MT) systems often lack the collaborative features needed for specialized domains like public health.
• Limited bilingual translation resources hinder the creation of accurate, accessible health information for diverse populations.

Purpose of the Study:

• To describe the development and evaluation of PHAST (Public Health Automatic System for Translation), a novel collaborative MT plus post-editing system.
• To introduce a new approach in public health informatics using groupware for sharing limited bilingual translation resources.
• To assess the feasibility and potential impact of PHAST on increasing multilingual public health material production.

Main Methods:

• Design and development of the PHAST system, incorporating collaborative features and groupware functionalities.
• Evaluation of PHAST with potential public health users to gather feedback on usability and effectiveness.
• Analysis of resource sharing mechanisms within the groupware system for optimizing translation workflows.

Main Results:

• PHAST demonstrates a feasible approach to increasing the production of multilingual public health materials.
• The collaborative system effectively reduces barriers related to time and cost in translation.
• User evaluation indicates potential for PHAST to improve the accuracy and accessibility of health information.

Conclusions:

• Collaborative machine translation systems like PHAST can significantly enhance the creation of multilingual public health educational materials.
• Sharing limited bilingual translation resources through groupware is a viable strategy in public health informatics.
• The PHAST design offers a transferable model for other communities aiming to improve MT accuracy via shared language expertise.