Abstract: A sentence is composed by semantics chunks.So，ellipsis in sentence means ellipsis of semantics chunks.This dissertation tries to resolve person and organization entity names’ ellipsis based on sentence category analysis and relationship of share between sentences.Ellipsis should be divided into two categories.The first is ellipsis formed by full semantics chunks share.The second is ellipsis formed by integrant semantics chunks share.An algorithm is represented for ellipsis.Result shows that the algorithm is efficient for both two kinds of ellipsis.

Key words: ellipsis, semantics chunks share, relationship between sentences, Hierarchical Network of Concepts（HNC）

摘要： 语义块是句子的语义构成单位，句子内发生的省略现象可以归结为语义块的省略。在句类分析的基础上，从小句间语义块共享关系的角度分析语义块中人名和机构名称的省略。将省略现象分为语义块整块共享形成的省略和语义块部分共享形成的省略，分析了两种情况的特点，并给出了相应的处理算法。测试表明，该算法对于两种省略均有很好的处理效果。

关键词: 省略, 语义块共享, 句间关系, 概念层次网络（HNC）理论