Abstract by Tiana Gale
Data Annotation and Automatic Event Extraction
Event extraction is the process of finding events and their components in narrative texts such as news articles and ordering these events in a timeline to reconstruct narratives. The goal of the event extraction project in IDeA Labs is to automate this process. In order to do this, it is necessary to have a collection of events extracted by human annotators that can be used as training data for machine learning. This presentation will discuss our event extraction model, demonstrate our method of manually annotating events in news articles, and explain why our collection of annotated events is important for training computers to do event extraction accurately.