An Efficient Approach to Summarizing Events from Microblogs

Abstract

In this paper, we present an efficient approach to summarizing events from microblogs. As many events are first reported on microblogging platforms, it is valuable to extract events from microblogs. However, current solutions on event extraction can only provide a course-grained description of events. We propose to generate summarizations for microblog-based events. Particularly, we use a set of short sentences to describe an event and propose a graph model to generate the set of short sentences for describing an event. We conduct experiments on a data set consisting of 14 events crawled from a well-known microblogging platform. The experiment results suggest the effectiveness of the proposed approach.