互补网络中多级混合过滤系统的研究  

Research of Information Filtering Technologies Based on Multi-Agent in Complementary Architecture Network

作　　者：邢超[1] 王珂[1] XING Chao, WANG Ke （College of Information Science and Technology, Henan University of Technology, Zhengzhou 450001, China）

机构地区：[1]河南工业大学信息科学与工程学院,河南郑州450001

出　　处：《电脑知识与技术》2008年第9期1423-1425,1430,共4页Computer Knowledge and Technology

摘　　要：针对适合互补结构网络信息特点的过滤机制进行研究，提出一种改进的用户兴趣模型，在研究内容过滤和协作过滤的基础上．建立适合互补结构网络的多级混合过滤系统。通过对多级混合过滤模型的实验，表明多级混合过滤模型克服了单独使用内容过滤或协作过滤方法的缺点，提高了过滤的准确度。The paper focuses on mechanism of Information Filtering of the complementary architecture network.An improved user-profile model is proposed. This paper sets up a system of Information filtering based on multi-level and combination strategy in Complementary Architecture Network, on the basis of the content Filtering and Collaborative Filtering. This paper carries out experiments and simulation of the mechanism for the mixed model in Complementary Architecture Network. Laboratorial results show that the model based on multilevel Combination Filtering can overcome shortcomings and enhance the accuracy of the information.

关 键 词：互补结构网络 多智能体 UCL标引 混合过滤 

分 类 号：TP393[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]