Paper Type

Complete

Paper Number

PACIS2026-2107

Description

This paper proposes a model to support resource planning—specifically time and effort estimation—in software development projects using Extreme Programming (XP). While XP ensures software meets user expectations, it burdens project managers with defining scope and estimating resources. The proposed model addresses this by leveraging generative AI to convert user stories into code snippets, enabling rapid prototyping and iterative refinement. This process clarifies software scope and complexity, aiding in mapping resource requirements. The refined prototype also serves as a detailed design artifact to improve communication between managers and developers. For resource estimation, the model introduces an “Applied COCOMO” formula—an adaptation of the basic COCOMO model. Experimental results show that this formula produces estimates closely aligned with those from the “Story Point and Velocity” model, a standard baseline for XP projects, while offering improved precision in estimating programming resources needed to complete the final version of the software.

Comments

17-General

Share

COinS
 
Jul 5th, 12:00 AM

Bridging Resource Estimation Gaps in Extreme Programming Using Generative AI and Applied COCOMO

This paper proposes a model to support resource planning—specifically time and effort estimation—in software development projects using Extreme Programming (XP). While XP ensures software meets user expectations, it burdens project managers with defining scope and estimating resources. The proposed model addresses this by leveraging generative AI to convert user stories into code snippets, enabling rapid prototyping and iterative refinement. This process clarifies software scope and complexity, aiding in mapping resource requirements. The refined prototype also serves as a detailed design artifact to improve communication between managers and developers. For resource estimation, the model introduces an “Applied COCOMO” formula—an adaptation of the basic COCOMO model. Experimental results show that this formula produces estimates closely aligned with those from the “Story Point and Velocity” model, a standard baseline for XP projects, while offering improved precision in estimating programming resources needed to complete the final version of the software.