【作者】 高静；

【导师】 应吉康；

【作者基本信息】 华东师范大学 ， 计算机应用， 2007， 硕士

【摘要】 信息技术和网络技术的飞速发展，在给人类交流和信息传播带来了革命性变化，为人们的生活、工作和科研带来了巨大方便的同时，也导致了“信息过载”和“信息迷向”。如何帮助人们有效的选择和利用所感兴趣的信息，尽量剔除不相干的信息，使之真正做到“各取所需”，成为信息技术领域的热点问题。推荐系统在此前提下应运而生。经过不到二十年的发展，推荐系统产生了许多新的方法，而最为成熟的技术是协作过滤技术，当然此种技术还是存在着许多不足之处。人工免疫是近几年在智能技术学科上新兴的研究领域。生物免疫是一个高度复杂的自适应系统，具有学习、记忆和模式识别的能力。通过模拟和应用免疫系统的信息处理能力，可以解决许多科学和工程问题。本文利用免疫的有关概念和理论，将之与推荐系统结合，提出了基于人工免疫思想的推荐算法。首先对推荐系统中常用技术进行了详细地阐述，指出这些技术的适用范围和优缺点。然后说明了人工免疫系统的生物学基础以及生物免疫一些基本概念，将免疫思想用于协同过滤技术中，提出了一个基于人工免疫系统的协同过滤算法，有效地解决传统协同过滤算法中出现的推荐准确度不够的问题，在计算推荐结果的过程中，提出了新预测评分方法，进一步提高了推荐结果的推荐质量，经改进后的算法在用户兴趣度预测的结果上更客观、更准确，从而提高了系统推荐的质量。最后实验通过从公共数据库中的收集数据来验证了算法的有效性及可行性。更多还原

【Abstract】 The rapid development of information technology and network technology brings about a revolutionary change for people to live, work and research. This change brought tremendous convenience, also led to the "information overload" and "mislead". It has become a hot issue in the field of information technology that how to help people interested in the effective use of information. Recommendation System came into being in this premise. After less than 20 years of development, recommendation system has produced many methods. Collaborative filtering technology is the most mature technology, of course, there are still many such technical deficiencies.Artificial Immune is an increasingly important area in the field of computational technology. Biological Immune is a strong information-processing system with diversity-recognizing capability, enhanced learning mechanics and distributed associate memory. Simulating biological immune system, the new computational techniques can solve not only the science but also the engineering problems.Based on the immune principles, this paper presents propose a modified recommendation algorithm based on artificial immune system. At the beginning, the recommendation system common technologies are introduced. Then the biology of the immune system and the artificial immune some basic concepts are described. Bring in the ideas to the recommendation system and propose a modified recommendation algorithm based on artificial immune system. This algorithm solves the problem of recommendation accuracy by the traditional collaborative filtering algorithm. In the process of the result calculating recommendation, this essay has brought new methods which having improved recommendation result further. At the end, the experiment to collect data from public databases verified the effectiveness and feasibility of the algorithm.更多还原