The article presents OmniEvent, a toolkit for event understanding, which refers to comprehending the content and relationship of events within texts. This process involves multiple information extraction tasks, including event detection, event argument extraction, and event relation extraction. OmniEvent supports mainstream modeling paradigms for these tasks and processes 15 widely-used English and Chinese datasets. It also provides off-the-shelf models that can be deployed as web services. The toolkit is designed to be easily used by users with varying needs and ensures fair comparisons between different models.
Publication date: 26 Sep 2023
Project Page: https://github.com/THU-KEG/OmniEvent
Paper: https://arxiv.org/pdf/2309.14258
