Improving Kaon identification with CLAS

The JLab program in Cascade physics is based on the process $\gamma p\to K^+K^+\Xi^-$. As such, it is critically dependent on efficient,reliable kaon identification. The largest background for this reaction is uncorrelated events in which one of the $K^+$ particles (or both) is actually a $\pi^+$ which has been misidentified due to its high momentum. This talk will outline the various methods currently used to separate pions from kaons, and will discuss the techniques currently being developed to extract the best possible kaon identification from our existing data set, as well as possibilities for the future.