Bugging and De-Bugging the Dataset using Improved Hidden Morkov Model (IHMM) for Cybersecurity Applications

Abstract

The priority has been gravely given to cyber security events’ finding, calculation or treatment in the wake of research field for the sake of the centers that are available for response of security. In such a surrounding or situation, the automation is a vital requirement for many processes in quest of finding a solid solution to the damage caused by the threats and assailings to the companies as well to the citizens. The process of bug handling is almost as a manual process which is but as at large a costly administration or maintenance of software systems.  While a vast sum of time as well as effort is poured upon dealing the bug reports, this way automating thus the parts of this process of bug handling can save the wasted vast sum of time as well as effort while dealing the bug reports. Bug cause comes handy for bug perception and bug localization, which to enkindle the developers must have scrutinized the source code during the process of bugging (bug fixing) and debugging.

This paper aims at exploiting the corresponding relationship between bug that fixes to automatically classify bugs from dataset using Improved Hidden Morkov Model (IHMM) which utilize Baum-welch genetic algorithm optimization process for better convergence. The aim of this work is to save the time as well as effort from being wasted in fixing the bugs while modifying (towards the improvement) the overall software’s quality. The model proposed achieves 98.3% accuracy, 81.2% precision, 79.9% recall and 90.2% AUC under Mozilla setup and 97.1% accuracy, 88.4% precision, 91.4% recall and 83.7% comparatively with its performance as when compared with other models under android firefox setup.