Predicting User Dissatisfaction With Internet Application Performance at End-Hosts
The authors design predictors of user dissatisfaction with the performance of applications that use networking. Their approach combines user-level feedback with low level machine and networking metrics. The main challenges of predicting user dissatisfaction, that arises when networking conditions adversely affect applications, comes from the scarcity of user feedback and the fact that poor performance episodes are rare. They develop a methodology to handle these challenges. Their method processes low level data via quantization and feature selection steps. They combine this with user labels and employ supervised learning techniques to build predictors.