摘要
提出了一个面向多文档自动文摘任务的多文本框架(Multiple Document Framework,MDF),该框架通过系统地描述不同层面的文本单元之间的相互关系以及文档集合蕴含的事件在时间上的发生及演变,将多篇文档在不损失文档集合原有信息的前提下实现信息融合.MDF简化了传统交叉文本结构理论的文本集合表示模型,又补充了信息融合理论中缺乏的事件主题的演变性和分布性信息.文中给出了建立MDF、基于MDF的信息融合、文摘生成等一整套算法.通过对32组不同主题的网络文档试验结果表明,MDF策略很好地实现了多知识源的并行融合,并获得了较好的结果.
A Multiple Documents Framework （MDF） is proposed for multi-document automatic summarization task. By representing interrelationship between text units at different levels of granularity and the happen and change of various events at time dimension, this framework can achieve information fusion of multi-document while reserve original information of set of related documents. MDF simplifies traditional multi-document representation in cross structure theory and simultaneously, supplements change and distribution informations of events topics which cannot be obtained in information fusion theory. Concretely, a series of algorithms including building MDF, multi-document information fusion based MDF and summarization generation are proposed. The capability of concurrently fusing multiple knowledge sources of MDF strategies is testified by experiments in 32 different sets of net documents and shows good results.
出处
《计算机学报》
EI
CSCD
北大核心
2007年第11期2048-2054,共7页
Chinese Journal of Computers
基金
国家自然科学基金重点项目(60435020)
国家自然科学基金(60773070)资助.~~
关键词
多文本框架
多文档自动文摘
信息融合
时间
multiple document framework
multi-document automatic summarization
information fusion
time
