Abstract
AbstractAs the main channel for people to obtain information and express their opinions, online media generate a huge amount of unstructured news data every day, which bring great difficulties for people to perceive social events and grasp the development of events. Event-centric knowledge graph has been used to facilitate the reconstruction of news to form structured event information. Most existing studies generate timeline based on event-centric knowledge graphs without considering the complex relations between events. This paper collects news data from Sina platform, constructs event ontology, and builds event-centric knowledge graph with temporal attribute. Afterwords, we propose a novel storyline generation framework with constraints of coherence and coverage. Experiment results show that our method significantly outperforms two baseline approaches.KeywordsStorylineKnowledge graphEvent evolutionCommunity detection
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call