Vulnerability Discovery Modeling and Weighted Criteria Based Ranking

Attacks on code based systems have been recent area of concern for the software developers. Of late, this side of the coin has received much attention as the loss happening due to this exploitation has been understood to a good extent. In today’s neck to neck competitive marketplace, firms have to come with their software products as quickly as possible. In order to do so, they are releasing their offerings at a much higher pace as it used to be earlier and so, many bugs sustain in the software at the time of release. Such a code is prone to be easily attacked by any community working in the field. With the goal of predicting or investigating these potential number of loop holes (vulnerabilities); many vulnerability discovery models (VDMs) have been proposed in the literature. In this paper, we develop a model which follows a hump-shaped curve while discovering the security vulnerabilities. Furthermore, we have compared different set of VDMs with the proposed model using the five comparison criteria and each criterion has been assigned different weight in order to capture the ranking of proposed model. For checking the veracity and predictive capabilities of proposed model, validation is done on two different data sets and results shows that the weighted criteria methodology shows very promising results for model comparison.