计算机应用 ›› 2012, Vol. 32 ›› Issue (07): 2030-2032.DOI: 10.3724/SP.J.1087.2012.02030

• 典型应用 • 上一篇    下一篇

绿色网络智能文摘算法研究

龙珑1,邓伟2   

  1. 1. 广西师范学院 计算机与信息工程学院,南宁530023
    2. 广西肿瘤防治研究所 五所,南宁530021
  • 收稿日期:2011-12-23 修回日期:2012-02-09 发布日期:2012-07-05 出版日期:2012-07-01
  • 通讯作者: 龙珑
  • 作者简介:龙珑(1980-),男,广西钦州人,高级工程师,硕士,主要研究方向:计算机安全、人工智能;邓伟(1980-),女,广西钦州人,主管医师,博士,主要研究方向:流行病学、网瘾预防。
  • 基金资助:

    国家创新基金资助项目(10C26224504901)

Green network automatic abstraction algorithm

LONG Long1,DENG Wei2   

  1. 1. College of Computer and Information Engineering, Guangxi Teachers Education University, Nanning Guangxi 530023, China
    2. No. 5 Department, Guangxi Cancer Institute, Nanning Guangxi 530021, China
  • Received:2011-12-23 Revised:2012-02-09 Online:2012-07-05 Published:2012-07-01
  • Contact: LONG Long

摘要: 随着Internet的迅猛发展,我国网民的数量激增,传统的自动文摘算法无法适应绿色网络提供优质内容并过滤不良内容的社会需求。不同于传统采用统计学习自动文摘算法,提出一种新的基于明确语义的算法,利用基于行为分析的绿色网络系统的云数据采集库和维基百科资源作为知识库建立概念空间,对该空间的词语进行语义解释。最后,实验结果证明,所提方法比传统算法用更少的句子即可以获得更高的信息覆盖率,从而大大缩短系统分析过滤不良内容的时间,提升软件的服务质量。

关键词: 绿色网络, 自动文摘算法, 概念空间

Abstract: With the rapid development of Internet, the traditional automatic abstraction algorithm cannot meet the needs of the green network. Different from the traditional automatic abstraction algorithm based on statistical learning, this paper proposed a new algorithm based on clear semantics. It made use of cloud data acquisition library of green network system based on behavior analysis and Wikipedia resources as the knowledge base to establish the concept of space, it performed semantic interpretation on these words. Finally, the experimental results prove that the proposed method can cover more information with fewer sentences compared with the traditional algorithm, thus it greatly shortens the time for the system to analyze and filter inappropriate content and enhances the quality of software services.

Key words: green network, automatic abstraction algorithm, concept space

中图分类号: