基于语义的自动文摘研究综述  

Review of the Research on Semantic-Based Automatic Summarization

作　　者：龚书[1] 瞿有利[1] 田盛丰[1] 

机构地区：[1]北京交通大学计算机与信息技术学院,北京100044

出　　处：《北京交通大学学报》2009年第5期126-131,共6页JOURNAL OF BEIJING JIAOTONG UNIVERSITY

基　　金：教育部科学技术研究重点项目(108126);国家自然科学基金资助项目(10871019/a0107)

摘　　要：基于语义的自动文摘方法能解决自动文摘处理中语言复杂性问题,在克服领域局限性的同时从本质上提高文摘的质量.常用语义模型包括统计主题模型和语义知识模型.基于语义的自动文摘是对传统文摘技术基于语义模型的扩充,在自动文摘的预处理、文档转换、文档候选片段提取和文摘生成4个基本过程中使用各种语义分析方法.Automatic summarization is an important researching branch of natural language processing. Semantic-based automatic summarization method solved the language complexity problem, increased the generality of summarization system, while improved the quality of summary. Statistic topic model and semantic knowledge model are the two main semantic models. Semantic-based automatic summarization enlarges the traditional, summarization technology by semantic models, and using semantic analysis methods in the four basic processes of automatic summarization： preprocessing, document transferring, summarization candidate fragments extracting and summarization generating. Semanticbased automatic summarization still has a wide space to develop.

关 键 词：自然语言处理 自动文摘 语义 语义模型 本体 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]