Using Rough Set Theory to Find Minimal Log with Rule Generation

The current research aims to enhance real-time analysis and speed of security algorithms by focusing on creating a minimal security log with the same efficiency as the original one through keeping core knowledge. Hence, manipulating this created minimal log with any security algorithm will surely enhance execution time and accordingly push towards real-time analysis. Rough set concepts are being used to reduce log size by using algorithms of the RoughSets package in R language, these algorithms guarantee to find a minimum log at least, and to find a minimal log at most depending on hardware resources available. Embedded knowledge is being extracted from minimum and minimal logs in the form of IF...THEN statements, and finally conclusions were discussed. Keywords: hate speech, text classification, classifier, classifier ensembles, stacking ensemble, text mining, genetic programming, pattern classification.