Sarsam, SM and Al-Samarraie, H orcid.org/0000-0002-9861-8989 (2022) Early-stage detection of eye diseases on microblogs: glaucoma recognition. International Journal of Information Technology, 14 (1). pp. 255-264. ISSN 2511-2104
Abstract
Glaucoma is the most popular optic neuropathy that causes blindness in people without warning signs. The early detection of glaucoma is crucial for an early treatment that could be useful to delay vision loss. However, since vision loss caused by glaucoma cannot be recovered, this study proposes an early detection mechanism for glaucoma using social media posts. Glaucoma-related tweets were collected using the Twitter streaming application programming interface (API). A hierarchical clustering algorithm was applied to group the tweets that share similar features together. In each cluster, the co-occurrence analysis was applied using the VOSViewer technique to map specific disease-related terminologies. Users’ emotions (e.g., anger, fear, sadness, and joy) and their polarity (positive, neutral, and negative) were extracted using NRC (Affect Intensity Lexicon) and SentiStrength techniques. The detection of glaucoma was achieved by using multinomial logistic regression (Logistic). The classification results showed that the Logistic classifier was able to predict glaucoma tweets with 98.73% accuracy. Our findings revealed that negative, fear, and sadness sentiments can be useful in detecting glaucoma. This study provides an effective mechanism to detect glaucoma disease from Twitter messages.
Metadata
Item Type: | Article |
---|---|
Authors/Creators: |
|
Keywords: | Glaucoma; Blindness; Lexicon-based approach; Classification; Twitter |
Dates: |
|
Institution: | The University of Leeds |
Academic Units: | The University of Leeds > Faculty of Arts, Humanities and Cultures (Leeds) > School of Design (Leeds) |
Depositing User: | Symplectic Publications |
Date Deposited: | 23 May 2022 12:52 |
Last Modified: | 23 May 2022 12:52 |
Status: | Published |
Publisher: | Springer |
Identification Number: | 10.1007/s41870-021-00726-7 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:187120 |