Knowledge Graph Semantic Annotation and Population with Real-Time Events Data from GDELT

Abstract

The availability of large event data brings new opportunities for social analysis and informed policy-making processes. GDELT (the Global Database of Events, Language, and Tone) is a knowledge base that contains event data from news media from all around the world in different languages, and captures the main events about physical activities, entities and concepts. Semantically annotating and representing event data from GDELT for knowledge graph construction in real time is a challenging task. Current related work does not benefit from other freely available background information and Linked Open Data (LOD) sources. This paper presents ongoing work on automatic semantic annotation and representation of the GDELT 2.0 event stream in real time and discusses the benefits of representing events using semantic technologies according to an Event Description Ontology.