Automatic Mining of Salient Events from Multiple Documents

Anonymous

Abstract: This paper studies a new event knowledge extraction task, Event Chain Mining. Given multiple documents on a super event, it aims to mine a series of salient events in a temporal order. For example, the event chain of super event Mexico Earthquake in 2017 is {earthquake hit Mexico, destroy houses, kill people, block roads}. This task can help readers capture the gist of texts quickly, thereby improving reading efficiency and deepening text comprehension. To address this task, we regard an event as a cluster of different mentions of similar meanings. In this way, we can identify the different expressions of events, enrich their semantic knowledge and enhance order information among them. Taking events as the basic unit, we propose a novel and flexible unsupervised framework, EMiner. Specifically, we extract event mentions from texts and merge those of similar meanings into a cluster as an event. Then, essential events are selected and arranged into a chain in the order of their occurrences. We then develop a testbed for the proposed task, including a human-annotated benchmark and comprehensive evaluation metrics. Extensive experiments are conducted to verify the effectiveness of EMiner in terms of both automatic and human evaluations.