基于K最近邻算法的网络不良信息过滤系统研究

汤烈 穆合义 候爱莲 鄢喜爱

摘   要：網络不良信息过滤的难点是如何对信息进行准确分类，而K最近邻分类器作为模式分类领域一种简单而有效的分类方法，在很多领域得到应用。提出了一种基于K最近邻算法的网络不良信息过滤方法，并对K最近邻算法计算量大的问题进行了一些改进。实验表明：采用K最近邻算法的不良信息过滤系统具有良好的应用效果。

关键词：K最近邻;不良信息;过滤;样本优化

中图分类号：TP39                                                    文献标识码：A

Research on Network Bad Information Filtering System Based

on K Nearest Neighbor Algorithms

TANG Lie1？覮，MU He-yi1，HOU Ai-lian2，YAN Xi-ai2

（1.Public Security Department of Hunan Province，Changsha，Hunan 410001，China;

2. Changsha Branch，The People′s Bank of  China，Changsha，Hunan 410005，China;

3. Hunan Police Academy，Changsha，Hunan 410138，China）

Absrtact：The difficulty of bad information filtering is how to classify information accurately. K-nearest neighbor classifier，as a simple and effective classification method in the field of pattern classification，has been applied in many fields. A method of filtering bad information in network based on K-nearest neighbor algorithm is proposed，and some improvements are made on the problem of large computation of K-nearest neighbor algorithm. Experiments show that the bad information filtering system using K-nearest neighbor algorithm has good application effect.

Key words：K nearest neighbor;bad information;filtering;sample optimization