Lost in Recursion: Mining Rich Event Semantics in Knowledge Graphs

Our world is shaped by events of various complexity. This includes bothsmall-scale local events like local farmer markets and large complex eventslike political and military conflicts. The latter are typically not observeddirectly but through the lenses of intermediaries like newspapers or socialmedia. In other words, we do not witness the unfolding of such events directlybut are confronted with narratives surrounding them. Such narratives capturedifferent aspects of a complex event and may also differ with respect to thenarrator. Thus, they provide a rich semantics concerning real-world events. Inthis paper, we show how narratives concerning complex events can be constructedand utilized. We provide a formal representation of narratives based onrecursive nodes to represent multiple levels of detail and discuss hownarratives can be bound to event-centric knowledge graphs. Additionally, weprovide an algorithm based on incremental prompting techniques that mines suchnarratives from texts to account for different perspectives on complex events.Finally, we show the effectiveness and future research directions in a proof ofconcept.