基于HNC理论的句类分析系统的设计与实现
文献类型：学位论文
| 作者 | 晋耀红 |
| 学位类别 | 博士 |
| 答辩日期 | 1998 |
| 授予单位 | 中国科学院中科院声学所 |
| 授予地点 | 中科院声学所 |
| 关键词 | 句类分析 语义块感知 概念层次网络 智能调度 |
| 中文摘要 | 概念层次网络理论（HNC）是基于语义的面向整个自然语言的理论框架。它以概念描述体系为基础，以语义块为基本单元，给出了语句深层结构的完备描述——句类。以句类知识为指导的语句分析处理就是句类分析。句类分析既不是基于规则的推理，也不是基于语料库的统计，而是用语句的物理表示式激活语句的全局联想脉络，HNC认为，这正是人脑感知语言过程的模式。本论文实现了黄曾阳先生提出的“中间切入，先上后下”这一句类分析的处理策略，即以语义块感知和句类辨识为切入点，先进行句类假设检验，然后进入语义块构成处理。依据这一处理策略，本论文设计了相应的分析处理算法。语义块感知算法生成语句的结构表示式，句类分析将结构表示式映射到语句的数学和物理表示式，并检验表示式的合理性。句类分析以智能调度为核心，智能调度的本质是数据动，数据是由数据黑板组织的。分析处理的知识来源于三个层面的知识库：概念知识库、语言知识库和常识知识库。句类分析特别适合汉语的分析处理，它实现了语句理解的突破，解决了长期困扰中文信息处理的一系列难题。本论文还简要介绍了基于句类分析而设计开发的“汉语拼音HNC智能输入系统”。 |
| 英文摘要 | The theory of Hierarchical Network of Concepts (HNC) is a semantics-based theoretical framework for the Natural Language Processing (NLP). The analysis/processing of the sentences based on the Sentence Categories is called the Sentence Category Analysis (SCA). SCA is the processing which triggers the associate skeleton of sentence by means of the physical expression of the sentence, rather than of rule reasoning or corpus. In this article, SCA is implemented, following the principle of "To cut-in the problem from the perception of semantic chunk, first Up then Down". That is to say, the chunk perception is conducted to identify Sentence Category. Then the SC hypothesis test is carried out, and the chunk constitution is processed. According to this processing strategy, the necessary algorithms have been presented. The chunk perception algorithm will build the form expression of sentence. Then SCA will map it into the mathematical/physical expression of the sentence, and test its rationality. The kernel part of SCA is an intelligent scheduler which can coordinate the whole processing. The intelligent scheduler is driven by the system data organized by the "Black-Board". The knowledge used in the processing comes from the knowledge bases at three levels: the concept knowledge bases, the language knowledge bases and the world knowledge bases. SCA is especially suitable for the Chinese NLP. It has solved many difficult problems that have puzzled the academic circle of Chinese Information Processing for long time. In this paper is also given a brief introduction of the HNC Intelligent Chinese Pin Yin-to-Text System which is designed and developed based on SCA.(图版 4个; 表格 0个; 参考文献 41个) |
| 语种 | 中文 |
| 公开日期 | 2011-05-07 |
| 页码 | 68 |
| 源URL | [http://159.226.59.140/handle/311008/1412]  |
| 专题 | 声学研究所_声学所博硕士学位论文_1981-2009博硕士学位论文 |
| 推荐引用方式 GB/T 7714 | 晋耀红. 基于HNC理论的句类分析系统的设计与实现[D]. 中科院声学所. 中国科学院中科院声学所. 1998. |
入库方式： OAI收割
来源：声学研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。