Scholarlink

Efficient software project management relies on sizing and measurement, facilitating control and performance comparison across different projects. COSMIC ISO 19761, presents a method to quantifying software functional size, catering to specific information requirements regardless of the underlying technology. Concurrently, Arduino’s accessibility, characterized by its user-friendly interface, open-source nature, and affordability, holds promise for a myriad of applications, including IoT devices and smart cities. The identification of common patterns in Arduino code within the COSMIC framework has spurred the development of CosmiCode. This tool, designed to measure software functional size in Cosmic Function Points (CFPs) from source code, yields an accuracy rate of 98.28% in initial tests. Moreover, an extension utilizing machine learning has been integrated into CosmiCode, showcasing promising capabilities in CFP measurement, with an average accuracy of 88.943%.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

CosmiCode: Automated COSMIC Measurement Tool for Arduino Using RegEx and NLP

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Mireille Bishay,
  • Salma Salem,
  • Milad Ghantous,
  • Hassan Soubra

摘要

Efficient software project management relies on sizing and measurement, facilitating control and performance comparison across different projects. COSMIC ISO 19761, presents a method to quantifying software functional size, catering to specific information requirements regardless of the underlying technology. Concurrently, Arduino’s accessibility, characterized by its user-friendly interface, open-source nature, and affordability, holds promise for a myriad of applications, including IoT devices and smart cities. The identification of common patterns in Arduino code within the COSMIC framework has spurred the development of CosmiCode. This tool, designed to measure software functional size in Cosmic Function Points (CFPs) from source code, yields an accuracy rate of 98.28% in initial tests. Moreover, an extension utilizing machine learning has been integrated into CosmiCode, showcasing promising capabilities in CFP measurement, with an average accuracy of 88.943%.