The Network Behavior Analysis Based On Stream Data Mining And Its Application Research

With the pervading and developing of the Internet in the whole world, computer network has already connected closely with people's studies and lives, so it becomes essential part of many people's lives. Various new services have been increasing so continuously that the network anomaly behavior is facing serious challenges. The establishment of network behavior detection model resolves theses problems.The dissertation studies and analyzes the network anomaly behavior and stream data mining firstly which builds up foundation for the detection methods of network anomaly behavior.The dissertation introduces the overall design and system architecture of the network behavior detection model. The design plan and the functional architecture follow for the data preparation and network behavior rules mining and managing modules. A new stream data clustering algorithm is presented and applied for mining. The designs and implementations of network behavior rules mining and managing module are highlighted.Last, by using the concrete application of gateway flow analysis, we verify the applicability and validity of the model in the practical use.The detection model outlined in this paper would be able to help the network managers to find the anomaly behavior, which has high practical value. Meanwhile, the research of the stream data clustering algorithm would be useful references to the similar researches.