Dictionary-based diabetes distress detection mechanism using facebook reactions / Marian Cynthia Martin
Over the last decade, the internet has paved the way for Facebook to become the digital hub for social networking transforming the ways of sharing information carrying rich and valuable information of users’ perspectives. Facebook support groups compromised of online diabetes communities facilitates...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/14482/ http://studentsrepo.um.edu.my/14482/2/Marian_Cynthia.pdf http://studentsrepo.um.edu.my/14482/1/Marian_Cynthia.pdf |
| Summary: | Over the last decade, the internet has paved the way for Facebook to become the digital hub for social networking transforming the ways of sharing information carrying rich and valuable information of users’ perspectives. Facebook support groups compromised of online diabetes communities facilitates users to connect to different people with similar conditions and share information wrapped in their own sentiments and emotions. Diabetes being a major life threatening health issue, results in diabetes distress among the online diabetes community. However, the detection of diabetes distress had been carried out
manually using surveys and questionnaires, accentuating the lack of studies in automated diabetes distress detection. Hence, this research aims to leverage on information from public Facebook diabetes support pages to extract extended features such as reactions along with posts to build a diabetes distress detection mechanism. An evaluation illustrates that the developed detection mechanism results with 62% accuracy, indicating that the proposed mechanism provides a feasible solution to detect diabetes distress in the online diabetes community. Finally, a comparison was done with the baseline study and the results depict a significant improvement in the overall accuracy of the proposed mechanism.
|
|---|