Automatic Risk Identification in Software Projects: an Approach based on Inductive Learning

The IT industry has boomed in the past few years with an ever increasing number of risk management applications being developed. There are inherent risks in software development projects and failure to deliver software projects within deadline or failure to develop software according to specifications can be costly. The software risks may occur during the project process. The management process of software risks consists the risk refinement, risk identification, risk monitoring, risk maintenance, risk estimation and risk mitigation. Neural Network has ability to stimulate hidden pattern recognition skill. The primary study of this paper is to focus on various risk management models and how risk tools may help in mitigating software risks during the project development. With the application of Neural Network, We propose short term risk management model which can predict the risk involvement with the upcoming project risks, analyzing from the previous projects causing serious loss in the IT project in terms of values on certain risk factors. Neural Network model can also ability to evaluate the assessment of risks in software development and acts as an effective instrument in analysis and minimizing risks that enable continuous improvement in software processes and products.

Download Full-text

Defining Indicators for Risk Assessment in Software Development Projects

CLEI electronic journal ◽

10.19153/cleiej.16.1.10 ◽

2013 ◽

Vol 16 (1) ◽

Cited By ~ 7

Author(s):

Júlio Menezes Jr. ◽

Cristine Gusmão ◽

Hermano Moura

Keyword(s):

Risk Assessment ◽

Risk Management ◽

Software Development ◽

Financial Market ◽

Management Practices ◽

Relevant Information ◽

Risk Identification ◽

Systematic Mapping Study ◽

Software Projects ◽

Practical Result

The usage of indicators acts in both strategic and tactical levels, it is effective for optimization of processes and also supports managerial decisions. Despite the relevance of risk management in software projects, it is in fact still usually overlooked by the organizations that develop software. One reason for this fact is that the concept of risk is abstract and subjective, and its management does not bring apparent immediate practical result. Differently, for example, in the financial market, where the risk management is consolidated and widely applied, and it is currently a need for the companies that work in this field. Briefly, risk management practices just became effectively consolidated in the financial market field when the uncertainties became measurable. In this context, this paper aims to define and propose indicators that are specific for environments of software projects in order to support risk assessment activities – risk identification and risk analysis. To achieve this objective, we first developed a systematic mapping study in order to collect evidences about metrics, indicators and relevant information for risk assessment. These findings were combined with the identification of measurable risk factors, providing, thus, a set of categorized indicators for software development environments.

Download Full-text