基于指代消解的中文事件融合方法

Approach of Chinese event information fusion based on co-reference resolution

事件抽取是信息抽取领域的一个重要的研究方向。针对事件抽取获得的信息不完整和语义不明确的特点,在此基础上引入指代消解和信息融合理论,通过事件元素归一化和互指事件的合并,在数据和特征两个层次上对事件信息进行融合,进一步明确和完善事件信息,并在不损失信息的情况下精简整个信息系统的信息量。在裁员事件的融合实验中正确率达到86.9%。

Event extraction is an important research direction in the area of information extraction. But the information extracted from single sentence is always incomplete and implicit. The authors provided a novel algorithm based on co-reference resolution and information fusion basic theory. That algorithm including three steps： event element standardization, co- reference event clustering and element fusion. Through those steps we can fuse all co-reference events and get a more explicit and complete event description at last. The experimental results show that the precision reaches as high as 86, 9%.