A Predictive Machine Learning-Driven Approach to Requirements Engineering for Satisfying Stakeholder Needs

Abstract

Requirement engineering (RE) plays a crucial role in the success of software development projects as it ensures that stakeholder needs are met. So that ambiguity, conflict, and incomplete assessments of stakeholder satisfaction are common challenges faced by professional real estate processes. In order to overcome these limitations this study recommended a predictive machine learning based approach to enhance decision making and improve stakeholder satisfaction in the Requirement Engineering (RE) process. Structured surveys and requirement documentation were utilized to gather data from over 1000 responds and the representation was diverse stakeholder groups. Decision trees, Random Forest, Support Vector Machine and Neural Network were used to predict the stakeholder happiness levels are determined by both requirement attributes and project parameters. The suggested technique combines data pre-processing feature selection model training validation and performance to ensure accuracy. Based on the experimental data the machine learning based requirements engineering model has a significant impact on prediction accuracy and gives valuable insights into factors that influence stakeholder satisfaction. Intelligence automation can be utilized to optimize requirement periodization as highlighted by the study enhancing communication between stakeholders and development teams is crucial to reducing projects risks. The research contributes to the growing field of data driven software engineering by introducing a scalable framework for optimizing requirements engineering (RE) by utilizing predictive analysis a design that is focused on stakeholder needs