Annotation data released for the CHIME grant under Project Epic. Additional Project Epic annotation data is available at https://github.com/Project-EPIC/epic-annotation
Dataset | # of Tweets | Schema | Link |
Identifying and Categorizing Disaster-Related Tweets | 7,490 | Multi-Layered | CHIME 2016 |
Improving Classification of Twitter Behavior During Hurricane Events | 25,474 | Binary Relevance | CHIME 2018-1 |
Developing and evaluating annotation procedures for Twitter data during hazard events | 1,964 | Refined Multi-Layered | CHIME 2018-2 |
The data provided includes only annotations, and not original tweets, as these are not available through the university. We recommend using the Twitter API to access these tweets: tweets that have been since been deleted or made private are then no longer available. We provide a tool here: Epic Tweet Documentation.
For any questions, please contact
Kevin Stowe
[email protected]