Joint extraction of entities and relations is an important task ininformation extraction. To tackle this problem, we firstly propose a noveltagging scheme that can convert the joint extraction task to a tagging problem.Then, based on our tagging scheme, we study different end-to-end models toextract entities and their relations directly, without identifying entities andrelations separately. We conduct experiments on a public dataset produced bydistant supervision method and the experimental results show that the taggingbased methods are better than most of the existing pipelined and joint learningmethods. What's more, the end-to-end model proposed in this paper, achieves thebest results on the public dataset.
展开▼