Recent years,with the policy inclination and the increase of social investment,the overall status of cyber security has been greatly improved,but relevant research on the local area network security within the campus network is relatively rare.In order to understand the internal security status of these small local area networks,we use an open-source software called Zeek to realize data collection of local area network traffic,and then uses Python and related machine learning algorithms to realize network traffic data processing,model training and abnormal behavior detection.The entire anomaly detecting process is convenient,fast,and at a lower economic cost,which is of great help to improve the efficiency of network management.