Bayesian belief network software

A QA process model consists of elements representing software development staffs, software QA staffs, development activities, QA activities and documents generated. Our major concern, quality, is represented as defects density. Each element is represented as a node and is coimected based on its casual relation with other elements. Each node is further designated with 2 5 states representing its status, for example, the undesired event is represented as defect density at high status. In reality, the relation between QA process elements is not static and fixed, i.e., the influence of one node on the other node often exhibits probabilistic and interactive behavior. In order to represent the probabilistic behavior of QA process, we apply Bayesian Belief Network (BBN)(Jensen, 1996) technique. A typical QA process represented in BBN is shown in Fig 3. 

Use of Bayesian belief networks to predict software quality has been proposed previously. [Hall, May, et. al, 1992] working on the FASGEP project used BBNs to measure confidence in the level of integrity of software design process, based on prediction of fault numbers. Fenton in [Fenton, Neil, 1999] present a critique of existing defect prediction models such as Multivariate approach and size and conplexity metrics and also conclude that BBNs offer some attractive benefits conpared to the existing software metrics techniques. 

The BBN structure resulted from the SERENE project [Fenton, Neil, 2004] was designed to support software assurance. More recently, in [Fenton, Neil, 2005a] and [Fenton, Neil, et. al, 2005b]. Bayesian belief network structures were developed to predict the quantity of unknown defects in a software development. Although these BBNs estimate quantities relating to, for exanple, the coders performance and the number of faults in the code, they do not predict the SIL that can be claimed. 

Figure 6- Bayesian belief network model of three phases of the safety software life-cycle.

Techniques do exist for estimating software failure rates based on historical use data (Adelard, 2001) and work has been performed using Bayesian Belief Networks (Strigini, 1996) enabling evidence from diverse assessment activities to be combined. However, historic data cannot, obviously, be used for newly written software. 

Gran, B.A. The use of Bayesian Belief Networks for combining disparate sources of information in the safety assessment of software based systems. Thesis 2002 35, NTNU, Trondheim, Norway (2002)... 

Gran, B.A. Use of Bayesian Belief Networks when Combining Disparate Sources of Information in the Safety Assessment of Software Based Systems. International Journal of Systems Science 33(6), 529-542 (2002)... 

---