Abstract:Big data era of the rapid development of information technology, based on all kinds of protective equipment of network hardware architecture of heterogeneous data volume by exponential magnitude increases every day, based on the traditional network security protection technology cannot effectively apply to has the characteristics of huge amounts of data analysis and prediction of network security and etc, so huge amounts of data storage, use and information mining and data analysis, prediction and analysis has become the social from all walks of life value and the current research trend. This paper takes massive heterogeneous data as the research object, identifies typical characteristics of network security big data, and combined with main methods of intelligence prediction, innovatively puts forward network security prediction and analysis technology under the characteristics of big data, so as to improve network security risk identification, prediction and Static ability, and effectively improve network protection effect.