节点文献
一种面向查询的多文档自动文摘系统实现方法
A System of Approach to Achieve for Query-Focused Multi-Document Automatic Summarization
【Author】 Zhuomin Gui~(12) Tingting He~(12) Jinguang Chen~(12) Fang Li~(12) 1 Department of Computer Science,HuaZhong Normal University,430079,Wuhan 2 Monitor and Research Center of National Language Resource Network Multimedia Sub-branch Center,430079,Wuhan
【机构】 华中师范大学计算机科学与技术系; 国家语言资源监测与研究中心网络媒体分中心;
【摘要】 针对面向查询的多文档自动文摘,本文提出了一种系统实现方法。首先通过对句子结构的分析发现,句子中某些成分并不能反映该句子的重要信息,提出在一定句子的修剪基础上,基于倒几率比的词权计算方法与改进的HAL语言模型方法,并应用于文本的自动摘要。实验证明该方法对自动文摘的质量有一定提高。
【Abstract】 In this paper,we propose an approach to achieve a system for query-focused multi-document summarization.At first,based on the analysis of sentence structure,some of certain components of the sentence do not reflect the important information.This paper put forward the inverse odds ratio of the calculation weight of the words and the improved relevance-based language modeling(HAL) to create the auto-summary based on some pruning of the sentence. Experiments show that our method achieves a certain improvement of the quality of the automatic summarization.
【Key words】 automatic summarization; inverse odds ratio; HAL; query-focused;
- 【会议录名称】 中国计算机语言学研究前沿进展(2007-2009)
- 【会议名称】第十届全国计算语言学学术会议
- 【会议时间】2009-07-24
- 【会议地点】中国山东烟台
- 【分类号】TP391.1
- 【主办单位】中国中文信息学会