Leveraging Machine Translation to Support Distributed Teamwork Between Language-Based Subgroups: The Effects of Automated Keyword Tagging

Yongle Zhang, Dennis Asamoah Owusu, Emily Gong, Shaan Chopra, Marine Carpuat, Ge Gao

Abstract: Modern teamwork often happens between subgroups located in different countries. Members of the same subgroup prefer to communicate in their native language for efficiency, which increases the coordination cost between subgroups. The current study extends previous HCI literature that explores the effects of machine translation (MT) on crosslingual teamwork. We investigated whether automated keyword tagging would assist people's comprehension of imperfect MT outputs and, therefore, enhance the quality of communication between subgroups. We conducted an online experiment where twenty teams performed a collaborative task. Each team consisted of two native English speakers and two native Mandarin speakers. We provided MT support that enabled participants to read all subgroups’ discussions in English before team meetings, but in two forms: with vs. without automated keyword tagging. We found MT with automated keyword tagging affected people's interaction with the translated materials, but it did not enhance translation comprehensibility in the context of teamwork.