节点文献
基于人工免疫系统的信息过滤技术研究
Study on Information Filtering Based on Artificial Immune System
【作者】 胡可;
【导师】 张家树;
【作者基本信息】 西南交通大学 , 通信与信息系统, 2006, 硕士
【摘要】 伴随着互联网技术的飞速发展,信息获取已经逐渐成为人们生活工作必不可少的内容。但是由于信息量的日渐庞大,信息种类的复杂繁多,一些不必要的,甚至是有害的信息给人们正常的信息获取带来了巨大影响,因此信息过滤技术已经成为国际上研究的热门课题。人工免疫系统作为人工智能领域的重要分支,同神经网络及遗传算法一样也是智能信息处理的重要手段,已经受到越来越多的关注。它通过类似于生物免疫系统的机能,构造具有动态性和自适应性的信息防御体系,以此来抵制外部无用、有害信息的侵入,从而保证接受信息的有效性与无害性。 本文围绕人工免疫系统的优良特性及其在信息处理技术中的应用展开研究,主要内容包括:(1)人工免疫系统的原理与特性分析及其在信息处理技术中的应用;(2)基于人工免疫算法的信息过滤方法与系统构建;(3)用于具体的电子邮件及web网页过滤的信息处理方法与平台设计。论文主要工作如下: 1.对人工免疫系统自身的原理特性与计算方法、中文信息处理技术以及人工免疫系统在信息过滤技术的应用现状进行了总结和分析,并指出了当前信息处理技术中存在的问题。 2.在掌握现有信息处理技术中所存在问题的基础上,分析现有信息处理技术的优缺点,特别是针对汉语的自然语言理解问题,指出现有信息过滤技术不能及时有效地处理信息的多样复杂性,同时对未知信息的识别能力差等缺点,充分利用人工免疫系统中阴性选择算法与克隆选择算法的特性并结合智能信息处理领域的相关技术,提出了一种基于人工免疫系统的信息过滤方案,设计了算法并通过仿真分析了系统的动态性与自适应性性能,从理论和实验两个方面验证了这种方案的可行性与有效性。 3.针对具体的电子邮件应用环境,构建了过滤垃圾邮件的Email免疫系统模型,并且开发了完整的电子邮件收发程序与基于网络的分词算法演示平台,可供实际环境的研究与测试。
【Abstract】 The Information Retrieval becomes a necessary way for people normal living along with the wide use of network. At the same time, to filter nocuous or futile information emerged as side effects of the popularity of information processing over network. So the fast and veracious Retrieval for kinds of information has become an important issue. Artificial Immune System which is one of the important fields of intellective information processing means to filter non-self information is receiving increasing attention.Research works focus on the excellent properties of Artificial Immune System and its use in information filtering, which include: (1) Analysis of the properties of Artificial Immune System; (2) the information filtering methods based on Artificial Immune Algorithm; (3) An idiographic information processing method to Email and Web. The main research fruits are as follows:1. Summarize and analyze the status of the use of Artificial Immune System in information filtering at present and give some examples of the use of information filtering, then point out some problems of exiting information retrieval.2. Analyze the methods of information processing in existence, especially Natural Language Understanding for Chinese and automatic segmentation of Chinese words. Summarize the problem and elicitation of information processing based intelligent technology.3. The function of the immune system is to protect the living body against invaders through the use of defensive mechanisms. This thesis looks at application of the artificial immune system model to protect email users effectively from spam and presents an immune algorithm called EIS that is capable of continuously classifying email as spam and non-spam. Comparisons are drawn with other classifier methods and it is shown that the proposed system performs as well as the other classifier system, especially on recall, and has a great potential for information classifying.
【Key words】 Information filtering; Artificial Immune System; automatic segmentation of Chinese words; Email filtering;
- 【网络出版投稿人】 西南交通大学 【网络出版年期】2006年 04期
- 【分类号】TP393.09
- 【被引频次】4
- 【下载频次】295