The emerging security analysis challenge in big data environment is getting more attention of security industry and research institute. The current problems are mainly about the volume of security data and the understanding of the data. To solve the problems, we should employ the big data technology, and meantime, consider the semantic processing methods. In this paper, we combine the big data processing with semantic methods and propose a framework for security analysis in big data era with semantics. In this framework, we can merge multisources security data, and process and analyze data using semantic connection knowledge, which is supposed to understand security data semantically and process data with semantic association and inference methods. The proposed framework can be worked with current threat intelligence sharing and exchanging mechanism.