中科大校园网络流量元数据

中科大校园网络流量元数据使用对网络协议的深度报文分析技术，抽取能精确刻画网络通讯特征的元数据。用元数据刻画的特征包括：对每一TCP连接的描述，三次握手完成的情况、传输中的异常、结束时是否正常关闭；对每一 HTTP 交易的描述（URL、Method、UserAgend等）。应用大数据分析对元数据进行任意层次、任意维度、任意时间度的分析，主要用于分析网络流量异常和突发的安全事件。

The USTC campus network traffic metadata dataset is developed through deep packet analysis of network protocols, which extracts metadata that accurately characterizes the core features of network communications. The metadata covers the following characteristic items: descriptions of each TCP connection, including the completion status of the three-way handshake, in-transmission anomalies, and whether the connection was closed normally upon termination; descriptions of each HTTP transaction (such as URL, request method, UserAgent, etc.). Big data analytics can be utilized to perform analyses at arbitrary levels, dimensions and time granularities on this dataset, which is primarily applied to detect abnormal network traffic and emergent security incidents.