Design And Implementation Of Log Analysis System Based On Data Mining Algorithm

According to the concept and guidance from software engineer, I design and implement a system which is based on data mining algorithm and used to analyze logs, from the perspective which concerns about requirement analysis, general and elaborate design of system, coding, software testing and the management of developing procedure in project named with TAT (Trace Analysis Tools).The following problems were resolved in this thesis. In order to implement the updatability and extensively on system, advisable design and architecture must be adopted. Design feasible data structure to transform and process logs. Analyze, optimize and implement related rule algorithm. Adopt the feasible design patterns to promote the running efficiency as well as performance of system, and save space in memory. Display the result which has already been classified and analyzed, in the form of table or picture. At the same time, consolidate the interactive communication function between human and computer.TAT software has been applied in software development of IBM CSDL (China Software Development Laborotry) DB2 for z/OS OSC team. Via the usage of TAT, the administrator of database can optimize the parameter in configuration, so that the performance of operation and access in database can be elevated. Mean while, developers and testers can easily discern the module which characterized with slow execution and frequent exception, and modify logic process in program by utilizing TAT.