Formal Verification Of Extension Of Istar To Support Big Data Projects

Identifying all of the correct requirements of any system is fundamental for its success. These requirements need to be engineered with precision in the early phases. Principally, late correction costs are estimated to be more than 200 times greater than the cost of corrections during requirements engineering (RE), especially in the big data area due to its importance and characteristics. A deep analysis of the big data literature suggests that current RE methods do not support the elicitation of big data project requirements. In this research, we present BiStar (an extension of iStar) to undertake big data characteristics such as volume, variety, etc. As a first step, some missing concepts are identified that are not supported by the current methods of RE. Next, BiStar is presented to take big data-specific characteristics into account while dealing with the requirements. To ensure the integrity property of BiStar, formal proofs are made by performing a Bigraph-based description on iStar and BiStar. Finally, iStar and BiStar are applied on the same exemplary scenario. BiStar shows promising results, so it is more efficient for eliciting big data project requirements