基于主题的网络论坛知识转换研究  

Research on Knolwedge Transfer of Internet Forum Based on Topic

作　　者：王力[1] 耿爱静[2] 

机构地区：[1]长春税务学院信息管理系,吉林长春130021 [2]吉林大学管理学院,吉林长春130022

出　　处：《情报科学》2005年第10期1505-1508,共4页Information Science

摘　　要：本文采用现有的中文自动标引与文档自动摘要的技术,将主题讨论区中的内容,通过网络技术自动汇总成常见问答集(Frequently Asked Questions)的知识形式,辅助版主能有效率地将主题讨论区中的知识分享给所有的成员使用。本文通过自动摘要文献的探讨,提出一个FAQ知识转换的概念模式。以混合式自动标引法作为中文关键词抽取的工具并结合相似度计算,将文章整理成FAQ摘要的形式。This research utilizes Chinese segmentation and automatic document summarization techniques to automatically summarize the forum content into an informative abstract, which, in terms, helps forum managers to preserved the tacit knowledge in subject forums efficiently. We propose a knowledge transfer model for FAQ （Frequently Asked Question） format through relevant literature. A hybrid segmentation approach was used to extract Chinese terms among sentences. The similarities between these article sentences were computed using Jaceard coefficient for organizing representative sentences in the FAQ format.

关 键 词：多文档摘要 信息检索 知识管理 

分 类 号：G354[文化科学—情报学]