It takes three to tango: understanding the world of online gossips through netnography, big data topic modelling and sentiment analysis
Gossips can be dangerous for business, but some of the content helps businesses identify customers' perceptions, thus permitting opportunities to grow. To date, there has been an increase participation in online gossip, in which people utilise platforms like Facebook, blogs and online forums to...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/99002/ http://dx.doi.org/10.1109/BCD54882.2022.9900719 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Gossips can be dangerous for business, but some of the content helps businesses identify customers' perceptions, thus permitting opportunities to grow. To date, there has been an increase participation in online gossip, in which people utilise platforms like Facebook, blogs and online forums to snoop on and discuss their favourite celebrities. In this study, we propose three approaches for exploring the context and comprehending the content and sentiment users discussed in online gossip threads: Netnography, topic modelling, and sentiment analysis. We retrieved 6055 conversation threads, including three social media celebrity threads, from an online forum called Cari Forum. Using netnography, we were able to identify seven topics in the thread. The thread was then further analysed using the Latent Dirichlet Allocation (LDA) topic modelling approach. We discovered that seven is the optimal number of topics detected in the thread with no overlap. We used the Logistic Regression Model to evaluate each conversation thread's sentiment. We identify each topic's positive and negative sentiment, obtained through our sentiment analysis approach. The three methods helped us to understand consumers' opinions and stories about the lives and businesses of social media celebrities. The implications, research limitations, recommendations for future research and conclusion have been elaborated on in this paper. |
|---|