| G. Drakopoulos, A. Karras, C. Karras, C. Makris, Ph. Mylonas, S. Sioutas |
| Decomposing Twitter Graphs Based On Hashtag Trajectories: Mining And Clustering Paths Over MongoDB |
| 12th Conference on Artificial Intelligence (SETN 2022), September 7-9, 2022, Corfu, Greece |
|
ABSTRACT
|
| Social media are widely considered as reflecting to a great extent human behavior including thoughts, emotions, as well as reactions to events. Consequently social media analysis relies heavily on examining the interaction between accounts. This work departs from this established viewpoint by treating the online activity as a result of the diffusion in a social graph of memes, namely elementary pieces of information, with hashtags being the most known ones. The groundwork for a general theory of decomposing a social graph based on hashtag trajectories is lain here. This line of reasoning stems from a functional viewpoint of the underlying social graph and is in direct analogy with the biology tenet where living organisms act as gene carriers with the latter controlling up to a part the behavior of the former. To this end hashtag diffusion properties are studied including the retweet probability, higher order distributions, and the mutation dynamics with patterns drawn from a MongoDB collection. These are evaluated on two benchmark Twitter graphs. The results are encouraging and strongly hint at the possibility of formulating a meme-based graph decomposition.
|
| 07 September, 2022 |
| G. Drakopoulos, A. Karras, C. Karras, C. Makris, Ph. Mylonas, S. Sioutas, "Decomposing Twitter Graphs Based On Hashtag Trajectories: Mining And Clustering Paths Over MongoDB", 12th Conference on Artificial Intelligence (SETN 2022), September 7-9, 2022, Corfu, Greece |
[  PDF] [
BibTex] [
Print] [
 Back] |
