中文摘要：

本体构建的重点在于概念的抽取,针对甲骨文卜辞特有的特征和已有的领域概念抽取算法的缺陷,提出了一种基于上下文语义的甲骨文领域概念抽取算法.该算法针对传统的基于DR＋DC的概念抽取算法的缺点进行改进,提出了基于上下文的概念间的相似度的计算方法,并给出了基于语义的领域概念筛选算法.实验数据表明,该方法在准确率和召回率以及困惑度衰减比率都有较大的提高.

英文摘要：

In the process of building ontology , It is focused on concept extraction. A new algorithm based oncontext semantic is proposed for Oracle-bone domain concepts extraction, according to characteristics of the oraclebone inscriptions and overcome the defects of domain concept extraction algorithms. Aiming at the shortcoming of al-gorithm of concept extraction based on DR ＋ DC, this paper makes some improvement on it. Is proposed based onthe calculation method of the degree of similarity between the concept of context, and proposed a sieving algorithmof domain concept based on context. Experimental results show that the algorithm has higher precision and recalland perplexity attenuation ration than the method based on DC ＋ DR.