Towards an Enhanced Approach for Big Data Requirement Generation Based on the KAOS Model: A Case Study of Electronic Health Records in Benghazi-Libya

Abstract

Problem statement: Big data systems face a challenge in the requirements engineering phase, given that the nature of their data differs from that of traditional systems. They include structured, semi-structured, and unstructured data, come from multiple sources, and are characterized by the 5Vs (volume, variety, velocity, veracity, and value). Despite increasing research efforts, there is still a lack of a comprehensive framework for managing the requirements of big data systems across their various engineering stages, making their development a complex process. Methodology: The study proposed an approach for generating big data requirements based on the KAOS model, with the aim of organizing and analysing requirements in a systematic manner. The proposed approach was applied through a case study on the electronic health records (EHR) system in Benghazi and evaluated through a questionnaire involving 100 participants who are specialists in the field to assess the effectiveness and applicability of the proposed methodology.  The questionnaire was not just a series of questions; the proposed approach was presented to participants, who were then given the opportunity to express their opinions about it. Findings: The survey results showed that participants had positive impressions of the quality of the proposed approach and its suitability for big data applications. Participants indicated that the approach could be adopted in the future as an effective tool for improving the quality of the requirements engineering process in big data environments. Contribution: This study contributes to filling a knowledge gap in the field of requirements engineering for big data systems by: presenting an applied approach based on the KAOS model for requirements generation; integrating theoretical and practical aspects through a real-world application in the health sector; and providing a quantitative evaluation mechanism to measure the acceptability and quality of the proposed approach. Limitation: The approach was applied to only one area (electronic health records), leaving room to try the proposed approach in other areas so that its results can be more widely disseminated. Furthermore, the approach focused only on the requirements generation process as a supporting approach to the requirements engineering phase, which includes other activities such as requirements documentation, verification, and management.