The Story of Event Graphs

Goran and Jan’s article covering the large part of Goran’s PhD research has been accepted for publication in the renowned Natural Language Engineering Journal. The article focuses automated extraction of information about events, their participants, and mutual temporal relations from new stories.

News stories are about events: who did what to whom, where and when. Structuring information about events from news stories, written in natural language, is not an easy task, though.

Within his PhD research (under Jan Šnajder’s supervision), TakeLab’s Goran Glavaš developed an approach for constructing the so-called event graphs — graphs in which nodes represent event mentions and edges denote temporal relations between events — from news stories.

Together, Goran and Jan described the process for construction and the methodology for evaluation of event graphs in the article “Construction and Evaluation of Event Graphs” which has been accepted for publication in the Natural Language Engineering Journal, a renowned journal and one of the very few journals dedicated specifically to NLP.