Realistic, large-scale simulations generate massive amounts of data from which information is extracted to provide analysis. The ability to provide useful, pertinent information requires that data sets be analyzed using data mining, feature detection, and feature extraction techniques. Domain specific methods employ well-formed mathematical descriptions of features often demanding an a priori knowledge of potential areas of interest. This is a cumbersome and tedious task. Domain independent tools, on the other hand, provide an automatic way to provide general characteristics. Because this process is not interactive and can be easily parallelized, these tools scale well with the size of the data. It is necessary to employ the use of domain independent tools to provide general characteristics and the most statistically probable regions of interest followed by the use of domain dependent tools to specifically locate features.