KDD2020_Poster_Promotion

Paper: https://arxiv.org/abs/2006.07812
Promotion Video: https://www.youtube.com/watch?v=4SOmBmi8g3Y

Transcript

1. Deep Exogenous and Endogenous Influence Combination for Social Chatter Intensity Prediction Subhabrata Dutta 1, Sarah Masud 2, Soumen Chakrabarti 3, Tanmoy Chakraborty 2 1 Jadavpur University, India; 2 IIIT-Delhi, India; 3 IIT Bombay, India ACM SIGKDD 2020 August 23-27 Virtual Conference
2. • Chatter intensity Research Motivation
3. • Chatter intensity • Existing prediction frameworks Research Motivation https://www.theguardian.com/news/blog/2011/jul/04/fox-news-hacked-twitter-obama-dead https://www.vectorstock.com/royalty-free-vector/fake-news-sign-or-stamp-vector-22753049
4. • Chatter intensity • Existing prediction frameworks • External influence Research Motivation
5. Research Motivation • Chatter intensity • Existing prediction frameworks • External influence Data Available Publicly • Discussion threads • Subreddit mapping • Who follows whom
6. GRU GRU ChatterNet: Unified Chatter Model
7. ChatterNet: Unified Chatter Model
8. ChatterNet: Results + and ++ refers to zero-shot and 1hr early observations, respectively
9. Details of our framework at: github.com/LCS2-IIITD/ChatterNet For more interesting research follow us at: @lcs2iiitd E-meet us at KDD 2020 ACM SIGKDD 2020 August 23-27 Virtual Conference