摘要
论文介绍了一种具有自学习能力的中文交互式网络搜索引擎INSE(aninteractivenetsearchengineforChi-nesetext),向量空间模型、基于自动机思想的中文分词技术和神经网络BP算法的应用是INSE的主要特点,重点讨论了INSE的自学习能力。基于自动机思想分词是INSE提出的新概念,应用于中文分词可以满足最大匹配且速度较快。INSE自学习能力的实现依靠神经网络的BP算法。该算法应用于交互式网络搜索引擎可以提供更加精确的查询结果。
The INSE(an interactive net search engine for Chinese text)presents in this paper is a kind of net search software,Vector Space Model,Chinese word segmentation based on automata and the Back-propagation algorithm of Ar-tificial Neural Networks are the main characteristics of INSE,This paper focuses on the self-learning ability of INSE.Chinese word segmentation based on automata is INSE's new concept ,which can efficiently segment Chinese word with maximum matching.INSE's self-learning ability comes from the adoption of Back propagate algorithm,which when used in an interactive net search engine,gives more exact query results.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第10期148-150,212,共4页
Computer Engineering and Applications
基金
国家863高科技研究发展计划资助项目(编号:863-104-02-01)