Community-Driven Information Accessibility
Online Sign Language Content Creation within d/Deaf Communities
Online Sign Language Content Creation within d/Deaf Communities
The lack of sign language information has made information access one of the most significant challenges facing d/Deaf people who take sign language as their native langauge. HCI researchers have worked hard to develop sign language generation and interpretation systems through computer vision and natural language processing techniques. However, no existing system is reliable enough for real-world adoption due to lack of training datasets, limitations in computational models, etc. Given the challenges in machine-driven solutions, we seek to understand whether and how d/Deaf communities can support the creation of online sign language content in practice. We turn to d/Deaf communities because d/Deaf communities have traditionally played significant roles in supporting d/Deaf people to access information.
Figure 1. A typical example of d/Deaf videos, i.e., sign language videos created by d/Deaf people for d/Deaf people.
Based on interviews with 12 d/Deaf signers in China, we found that many of them enjoy watching Chinese Sign Language (CSL) videos created by and for d/Deaf people (termed as d/Deaf videos in this study) for informational (e.g., COVID-19 information) and/or educational purposes (e.g., basic legal knowledge). Seeing the potential and importance of d/Deaf videos to d/Deaf people's information access, we combined content analysis of 360 d/Deaf videos, diving deep into the creation and sharing of these videos. The results how d/Deaf people in China collaborated online to power community-driven information accessibility.
This work is a two-phased study consisting of:
(1) interviews with d/Deaf signers.
(2) content analysis of d/Deaf videos.
Our study began by interviewing 12 d/Deaf signers
to understand their online information access. Based on the findings from the interviews, we
added content analysis of 360 d/Deaf videos collected from 12 video channels recommended by
our participants and the Deaf professionalist we involved in our research team.
The findings present d/Deaf people's collaborative content generation and sharing on Chinese video-sharing platforms. We found that d/Deaf videos serve as valuable information sources and educational materials to many d/Deaf people, which compensate for the lack of official CSL information in China – the sign language promoted by the Chinese government is best described as Signed Chinese rather than CSL, because it is created based on Mandarin by borrowing words from CSL; as a result, most d/Deaf people in China cannot understand sign language interpretation in official news.
On the other hand, the findings also uncovered the challenges in creating and sharing d/Deaf
videos, including:
(1) difficulties in sign language interpretation caused by language
diversity and unforeseeable words in CSL.
(2) challenges in ensuring content quality.
(3) d/Deaf content creators’ potential negative experience caused by potential online trolls and (micro)aggressions.
We reveal design opportunities to support collaborative content generation within d/Deaf
communities. Such a community-based approach is particularly important in contexts where sign
language has not been standardized yet (e.g., sign language in China, Cambodia, and Indonesia;
American Sign Language in emerging fields such as science).
More importantly, the approach can empower d/Deaf people in accessing information (i.e.,
d/Deaf people can decide what information they can access rather than relying on interpreters).
We also highlight the challenges that d/Deaf people meet in sign language content creation
(e.g., language diversity, unforeseeable words in CSL, inconsistent content quality and possible
microaggressions) that can be further addressed in future research.
Tang, X., Chang, X., Chen, N., Ni, Y., LC, R., & Tong, X. (2023).
CHI '23: Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. April 2023. No.: 50, 1–24.
