A Method of Predicting Software Behavior Risk based on Off-line Runtime Verification
Lei Hu, Guohua Jiang · Advances in engineering research/Advances in Engineering Research · 2016
The current methods of software behavior risk prediction is mainly through the study of the operating rules from the data of the other software of the same type, and that leads to differences between the prediction results and the actual software behavior.Aiming at this problem, this paper presents a software behavior prediction method, which combines prediction of software behavior with runtime verification, using Markov Chain and Hidden Markov Model(HMM), to analys the data from offline runtime verification and predict software behavior.Experiments show that this method can significantly improve the accuracy of the prediction.