面向android手机平台的网络恶意数据流挖掘研究

由于android手机平台网络中数据流众多,以往研究出的面向android手机平台网络恶意数据流挖掘方法,均无法对网络恶意数据流进行高效、准确挖掘。故提出一种挖掘效率和挖掘准确性均较高的android手机平台网络恶意数据流挖掘方法。网络恶意数据流通常均有自动收发行为,所提方法利用概率分类法和邻近值法对android手机平台网络数据流进行预分类,使具有自动收发行为的网络数据流优先进行网络恶意数据流挖掘,提高挖掘效率和挖掘准确性。该挖掘方法将网络数据流划分成多段行为向量,对具有自动收发行为的网络数据流和不具有自动收发行为的网络数据流采取不同精度的挖掘操作,输出网络恶意数据流,存储网络非恶意数据流,供下次挖掘使用。经实验验证可知,所研究的方法挖掘效率高、挖掘准确性高。

For android mobile phone platform of network malicious data stream mining research

Because of the android platform in the network data flow is numerous, ever developed for android mobile phone platform network malicious data stream mining method, the network is not malicious data flow accurately and efficiently mining. Therefore proposes a mining and mining efficiency high accuracy of the android mobile platform network malicious data stream mining method. Network malicious data flow often have automatic transceiver behavior, the proposed method using the method of probability classification and adjacent values for android mobile phone platform network data flow classification, make the network data flow, with automatic transceiver behavior priority for network malicious data stream mining, improve the efficiency of mining and mining accuracy. The mining method will be divided into many segments behavior vector network data flow, with automatic transceiver behavior of the network data flow and do not have automatic transceiver behavior of the network data flow to different precision of the mining operation, the output network malicious data streams, the network of malicious data flow for storage, for the use of mining next time. After the experimental verification shows that the research method of high efficiency mining, mining accuracy is high.