The filtering of a large amount of data is an important process in data mining tasks, particularly for the categorization of unstructured high dimensional data. Therefore, a feature selection process is desired to reduce the space of high dimensional data into small relevant subset dimensions...