Understand System's Relative Effectiveness Using Adapted Confusion Matrix

摘要

The effectiveness of a system refers to the accuracy and completeness with which users achieve specified goals. These two aspects are interpreted as errors and completion in the context of usability testing. However, a holistic view of effectiveness is not straight forward to establish in a comparative test because the two measures focus on different aspects of user outputs. In this paper, we propose a predictive method to measure a system's relative effectiveness based on its own performance prediction. We achieve it by using an adapted confusion matrix to establish a correlation model between the two measures. A real-world use case is provided to demonstrate the usefulness of our method in a comparative study of the two websites.