Experimental Data for Rumor Detection Model Based on Dynamic Propagation and Neural Hawkes Process

The Twitter15 and Twitter16 datasets contain 1,490 and 818 tweet events respectively. Each tweet event consists of a tweet propagation graph, where the nodes represent tweet posts and the edges represent reply or retweet relationships between tweet posts.

Each source tweet is labeled with the corresponding category of the tweet event. Both datasets include four labels: True Rumor (TR), False Rumor (FR), Unverified Rumor (UR), and Non-Rumor (NR). In addition, the core code of the DPNHRD model is provided in this data list.