A Sentence-Level Semantic Annotated Corpus Based on HNC Theory

The Sentence-level Semantic annotated Corpus (SSC) is directed by HNC (Hierarchical Network of Concepts) Theory in which the sentence is regarded as the basic unit and semantic and frame information are annotated. This paper introduces the building form of the semantic annotated corpus with XML, describes the content of semantic annotation on sentence level that the annotated contents can be classified three items: sentence category, which is the sentence semantic type; semantic chunks, which is the lower level semantic components in the sentence; sub-sentence (sentence ecdysis and chunk-extension-into-sentence) which is included in semantic chunks. SSC has been an important and valuable knowledge resource for language study and information processing.