关键词: 信息融合, 事件抽取, 指代消解, 信息抽取, event extraction

Abstract: 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%.

Key words: information fusion, co-reference resolution, CRE, information extraction, VIPS