Exploring text classification of social support in online health communities for people who are D/deaf and hard of hearing

Biyang Yu,Lynette Gerido,Zhe He
DOI: https://doi.org/10.1002/pra2.2017.14505401179
2017-01-01
Proceedings of the Association for Information Science and Technology
Abstract:ABSTRACT Due to the social oppressions from the mainstream society, social support becomes fundamental for people who are D/deaf and hard of hearing (D/hh) to exchange information within their own community. We aim to examine the support content of messages and how support processes become established by exploring the interactions within a D/hh discussion forum, AllDeaf.com , which is one of the leading online communities for D/hh. In this pilot study, we first employed content analysis to code discussion posts using the Social Support Behavior Code. Two authors collaboratively coded a random sample of 25 threads and 632 posts. Informational support was observed as the most frequently exchanged type of social support for people who are D/hh. Then we identified social support features used for future text classification task. Our preliminary qualitative findings indicated that the linguistic and lexical features have the potential to support automated feature identification and social support classification typology with text mining techniques
What problem does this paper attempt to address?