In this section, we evaluate our system to measure how well it performs in detecting malicious applets. For our experiments, we used two different datasets: A manually- compiled dataset of applets (the manual set), and a set of samples collected and provided to us by the authors of the Wepawet system (the wepawet set).