Exploiting Syntactic and Semantic Information for Relation Extraction from Wikipedia
Exploiting Syntactic and Semantic Information for Relation Extraction from Wikipedia
复制标题
DOI：
--
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
Y. Matsuo;M. Ishizuka
中科院分区：
文献类型：
--
作者：
Y. Matsuo;M. Ishizuka
最近，维基百科的指数级增长吸引了大量研究人员和从业者的注意。维基百科目前面临的挑战之一是让机器能够处理百科全书。本文讨论了从维基百科的英文文章中提取实体之间的关系的问题，这些关系可以直接转换为语义Web元数据。我们提出了一种利用句法和语义信息进行关系抽取的方法。此外，我们的方法可以利用维基百科的性质来自动获取训练数据。实验的初步结果有力地支持了我们的超级命题，即在维基百科上使用更高描述级别的信息可以更好地进行关系抽取，并表明我们的方法在文本理解中是很有前途的。
The exponential growth of Wikipedia recently attracts the attention of a large number of researchers and practitioners. One of the current challenge on Wikipedia is to make the encyclopedia processable for machines. In this paper, we deal with the problem of extracting relations between entities from Wikipedia's En- glish articles, which can straightforwardly be transformed into Semantic Web meta data. We propose a method to exploit syntactic and semantic information for relation extraction. In addition, our method can utilize the nature of Wikipedia to automatically obtain training data. The preliminary results of our experiments strongly support our hyperthesis that using information in higher level of description is better for relation extraction on Wikipedia and show that our method is promising for text understanding.