Research And Application For Visual Data Mining In Network Alarm

Data mining is a new technology which can integrate the human mind's exploration abilities with the enormous processing power of computers to form a powerful knowledge discovery environment that capitalizes on the best of both worlds. Visualization is an intersected science which is based on Computer Graphics , Picture Process, CAD, Computer Vision, User Interface Method , System Design, etc. Data Mining and exploration of large datasets is a new branch which is developed in recent years and has a promising future.In this article, firstly the author introduces some basic conceptions, impacts, applying and problems about the visualization and data mining technologies. Based on these researching, then gives a special emphasis on the combination of them, and its great meaning, applying and tendency. And then fix eyes on the framework of visualizational data mining system.The main research works are following:1. Introduced the concept ,feasibility, development and prospects of network alarm. And the network alarm system Alarm Way, Jiejing Method and the security alarm system were described in detail.2. Designed and implemented a visualization of data mining in the network alarm system in the application of the model. Analysis in detailed the Alarm messages information.3. Improved Bayesian classifier,to strike a conditional probability to join in the weight of amendments.4. Implementation of the Bayesian classifier results and process visualization, And a very good human-computer interface.