网络大数据平台中的特征数据分类系统设计与实现

Design and implementation of feature data classification system in network big data platform

网络大数据平台中特征数据的有效分类,是提高网民查询体验、开发新型应用的有效途径。为此,设计稳定性好、资源占用率低的网络大数据平台特征数据分类系统。系统的显示端是网民的直接应用端,其主要进行网络大数据的获取、大数据获取结果的显示和特征分类结果的显示。服务端利用SOA体系结构为网络大数据平台提供特征数据的分类服务,其将特征数据的分类标准纳入到网络大数据中,并传递给逻辑层处理端。逻辑层处理端根据特征数据分类标准,利用云计算和策略设计对网络大数据集合进行特征提取,其特征提取算法于软件中给出。特征数据分类端根据逻辑层处理端所提取出的大数据特征,利用特征向量机进行特征数据的自动分类工作。实验结果表明,所设计的系统稳定性好、资源占用率低。

Effective classification of the characteristics data of the network big data platform is to improve the Internet que- ry experience of netizens, and an effective way to develop new applications. Therefore, a characteristic data classification system with good stability and low resource utilization was designed for the network large data platform. The system＇s display terminal is the direct application client of netizens, which is used to acquire the network big data, and display the data acquisition result and feature classification result. The server utilizes SOA architecture to provide the classification service of the characteristics da- ta for network big data platform. The classification standard of characteristic data is brought into network large data, and passed to the logic layer processing side, which extracts the characteristics of network big data set according to the characteristic data classification standard, cloud computing and strategy design. The feature extraction algorithm is given in the third paragraph of this paper. Feature data classification end classifies the characteristics data automatically by using the feature vector machine （SVM） according to the big data characteristics extracted by logic layer processing side. The experimental results show that the designed system has high stability, low resource utilization.