Timelines: Entity-centric Event Extraction from Online News
Automatically extracting structured information on events from online sources for the purpose of intelligence gathering has been acknowledged to be of paramount importance by various organisations worldwide.
This paper reports on an ongoing endeavour on developing a tool that for a given target entity of interest extracts from a stream of online
news articles structured information on events this entity participated in or in whose context it was mentioned. Furthermore, other entity-
related relations that hold are extracted and whenever applicable the events are anchored on a time-scale and classified, which all together
constitutes a target-entity event timeline. The paper first briefly introduces the timeline extraction task and then gives an overview of the core extraction engine and event browsing functionalities. The results of a rudimentary evaluation of the quality of the extracted information
is also provided. The paper is accompanied with a demo of the tool.
PISKORSKI Jakub;
ZAVARELLA Vanni;
ATKINSON Martin;
VERILE Marco;
2021-12-06
-
JRC120446
1613-0073 (online),
Additional supporting files
| File name | Description | File type | |
|---|
