Forecasting Web Application Software Aging Damage Based on User Behavior

Software aging threatens to the reliability of the software and computer system seriously, and has become the main bottleneck restricting the development of software reliability. As most of user behavior has certain regularity, so it can be forecast. This paper proposed a prediction method based on the Frequent-User-Access-Patterns-Tree-With-Time-And-Aging (FUAP-Tree). And this method compressed the user behavior into a FUAP-Tree to store the Web log, and then predicted the user behavior using pattern matching by traversing the sub-trees of the FUAP-Tree. The load of the Web application software could be obtained when user behavior is known. So we could predict the Web application software aging. The conclusion we got by the experiment of comparing with another method is that this method got a more effective and accurate result.