Enhancing Software Development Effort Estimation: A Cloud-based Data Framework Utilizing Use Case Points, Fuzzy Logic, and Machine Learning

Effort estimation is a crucial task in the software development process, as it determines the type of resources needed, the number of required resources, and the project duration. Nevertheless, client requirement could alter during the development process, reducing the accuracy and dependability of early estimates. By offering a methodology, this study seeks to improve effort estimation. A framework that uses the Use Case Point method to estimate the project’s effort has been developed and made available as software as a service on the cloud. The framework determines how long the project will take, based on the input values for the project’s domain, technical factors, environmental factors, and information. The framework maintains a database called the use case repository, which developers and project managers may use to compare and estimate the length of future projects. To increase accuracy and dependability, fuzzy rules and a triangular membership function are employed in fuzzy logic to predict the time. The estimation’s accuracy is demonstrated by cross-checking the findings against popular datasets. Regression analysis is used to anticipate the project’s duration using the Gaussian Process Regression technique in combination with a machine learning system. The training and testing data, collected from friends, classmates, and other software businesses, are validated against the widely available De-shranais dataset. Ultimately, a business intelligence tool is used to produce reports and do comparative analysis based on various features in the repository data. Developers and administrators may now use a range of criteria to analyze the repository data according to specific project and domain-specific requirements.