1 Research Periodical of University Press, Volume 2, Number 1, June 2007 A Qualitative Approach for Software Project Risk Identification by KeyGraph Gufran Ahmad, Daxin Liu College of Computer Science Harbin Engineering University gufran.ahmad@hrbeu.edu.cn Abstract— Software project risk identification is a crucial step in the process of software risk management. Software risk identification is significant not only for research intent but for industrial intent also. The current work is a successful endeavour to solve the software project risks identification problem with KeyGraph, a graphical tool of chance discovery. The procedure simplifies the software risks identification. We apply KeyGraph method in a realistic software project scenario to discover newly generated hubs which are of prime importance in the identification of software project risks that are associated with the software project. After a careful analysis of the software project scenario, we identify these significant hubs as software project risks within the concerned software project. Keywords— Software project risk, KeyGraph, Software Project, Project Management I. INTRODUCTION Software project risk identification is an initial and integral process in every software risk assessment which is a part of software risk management. Without proper software risk identification, it is tough to control and manage the risk that is associated with the software project [1], [2]. So far, a number of literatures and articles [1-5] are contributed to this effort and a vast majority of techniques have been applied for the concerned problem to solve the issue but still most of them are basically, dependent on rigorous textual procedures to identify risks [4]. The first thing to note that risk management entails additional cost. However, risk management can be considered cost-effective only if the cost of risk management is considerably less than the loss incurred if the risk materializes. Actually, the cost of risk management should be less than the expected value of the loss. Therefore, it is again obligatory to identify software project risk efficiently [6]. In addition, the significance of software project risks has become so dominating in all software development areas that some recent trends have centred its entire efforts on software risks. One of the recent trends is based on project’s particular risks with respect to the use of agile or plan-driven methods in software development. It is known as risk-based method or risk-driven method software development. It is considered as hybrid agile method and is applied to software projects to balance discipline and agility and it can craft big extraordinary change based on project risk patterns. Henceforth, it simplifies all the software development and software project processes and adds a value to software project management as well [3]. II. RESEARCH OBJECTIVE The main objective of this research is to identify the software project risks which are always associated with every software project by means of KeyGraph method. We have classified these software risks as per the technical report of CMU/SEI by Marvin J. Carr, et al [7]. III. RESEARCH METHOD The software risk identification process begins with the creation of software project risk scenarios. These scenarios are at first, made by collecting essential set of data or nodes as input to the formation of KeyGraph [8]. Further, co-occurrence network is created by links between these frequent co-occurring item-pairs or node-pairs of the data-set. The fundamental islands are obtained by deleting links that separates the graphs and deleting the links of highest contribution. This situation is regarded as generation of hubs, candidates of rare event or chance, indeed, risks which are significant items, touching many node-cluster bridges of frequent co-occurrence [9], [10]. Finally, a deliberate analysis is done before the nomination of these risks. These risks are nominated with the aid of software risk categorization. IV. RESEARCH DATA-SET COLLECTION Research data-set is collected from one of the software project scenarios cited in Anne Fuller, et al [5] in their case studies. The cited case is mentioned as follows. “A company has developed a very comprehensive software product for a particular target industry. However, most of the users in the industry cannot afford the costs of deploying the product. The software is too complex for simple users to handle, and potential clients must dedicate a team of people with knowledge of the industry and advanced computer skills if they are to successfully adopt the package. Thus the developers need to scale down the original product in order to increase market share. They are considering two options: to block out some modules in the original system and thus disable some functionality, or to write another small system from scratch, reusing some code from the old system.” On the basis of this case, we create software risk scenario with KeyGraph to identify the software project risks that may occur during such changes by applying KeyGraph method.