An Analysis of Hybrid Tool Estimator: An Integration of Risk with Software Estimation

Abstract

Problem statement: One important problem with software development projects is to get an early and nevertheless accurate estimation of the effort needed to complete the project within the schedule. In the literature various methods have been developed for this purpose. The most popular examples are Boehm’s COCOMO, Albrecht’s Function Point Method or Sneed’s object-point method. The two last named methods are based on early results of an analysis phase; whereas COCOMO is based on an a priori estimation of the software size in “Lines of Code”. Despite of the increasing needs and available tools and methods, a satisfactory solution is still to be found. During the last two years, has gained some interest in this community an approach based on hybrid technique of software estimation. Approach: In this study, we discuss that traditional Function Point Method does not cover the quality factors and the estimation is fully based on development of systems. Hence, the quality assurance factors were discussed in this study. The comparative analysis of the existing software estimations were also developed and compared with the developed model so that the efficiency of the model can be analyzed with the existing methods. The classification of software system for which the effort estimation is to be calculated is based on the COCOMO model classes. So, our aim is to develop a hybrid method which combines all the important parameters from the various existing method for effort estimation. Once the effort estimation has been found, the same have been extended to risk assessment techniques by considering various risk parameters. So, the developed hybrid model is an integrated model of estimation with risk assessment. Results: A software has been designed (Front End-Java, Back End-MS-ACCESS) which shows the comparison between the traditional Function point method and the proposed method. Conclusion: Detailed comparative analyses have been made based on the result for all the estimation techniques.