Abstract

Knowledge discovery in text (KDT) often known as text data mining or simply text mining is the process of applying knowledge discovery methods to unstructured text. A suggested support vector classifier that uses cross validation for the original support vector classifier with a polynomial kernel is described in this study. If we want to optimize classification accuracy and make it better. Using data sets such as intrusion detection in computer networks the feasibility and advantages of the suggested technique are proven. On average the suggested support vector machine achieved a level of accuracy that was over 28.2% higher than that of the original polynomial-kernel support vector machine. Lots of ideas and solutions may be applied to various classifier paradigms since this technique isnt reliant on specific data sets.