COSMIC Function Points Evaluation for Software Maintenance

被引:0
作者
Hira, Anandi [1 ]
Boehm, Barry [1 ]
机构
[1] Univ Southern Calif, Los Angeles, CA 90089 USA
来源
ISEC'18: PROCEEDINGS OF THE 11TH INNOVATIONS IN SOFTWARE ENGINEERING CONFERENCE | 2018年
关键词
Cost Estimation; Effort Estimation; Project Management; IFPUG; Function Point Analysis; Software Non-Functional Assessment Process; COSMIC Function Points; CFPs; Software Maintenance; Local Calibration; COCOMO; VALIDATION;
D O I
10.1145/3172871.3172874
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
The Common Software Measurement International Consortium (COSMIC) group reviewed the existing functional size methods, such as the International Function Points User Group (IFPUG)'s Function Points (FPs), to develop a functional size metric based on "the basic principles" that applies to a wide range of application domains. Though several empirical studies on the COSMIC method verify that COSMIC Function Points (CFPs) successfully accomplished the goal of being applicable to a wide range of application domains and that its size correlate well with effort over a very wide range of sizes, one study of telecom switching software noticed that the correlation between CFPs and cost is very low for small projects (5 CFPs or less). The COSMIC method does not explicitly size data manipulations (such as, mathematical algorithms), which causes it to be less effective for mathematically-intensive software. IFPUG's FPs method has the same drawback of not explicitly measuring mathematical operations, but IFPUG developed the Software Non Functional Assessment Process (SNAP) to complement a project's functional size. This empirical analysis will determine whether CFPs can be an effective size metric for small, maintenance tasks (between 2 and 12 CFPs) using a dataset consisting of Unified Code Count (UCC)(1)'s maintenance tasks. Additionally, this analysis will consider whether using IFPUG's SNAP with COSMIC's FPs can lead to better effort estimates, as the former provides a method to measure data manipulation. The authors found that tasks adding new features require a different effort estimate model from those that modify existing features.
引用
收藏
页数:11
相关论文
共 39 条
  • [21] AN EMPIRICAL VALIDATION OF SOFTWARE COST ESTIMATION MODELS
    KEMERER, CF
    [J]. COMMUNICATIONS OF THE ACM, 1987, 30 (05) : 416 - 429
  • [22] An exploratory study of how developers seek, relate, and collect relevant information during software maintenance tasks
    Ko, Andrew J.
    Myers, Brad A.
    Coblenz, Michael J.
    Aung, Htet Htet
    [J]. IEEE TRANSACTIONS ON SOFTWARE ENGINEERING, 2006, 32 (12) : 971 - 987
  • [23] Software effort models should be assessed via leave-one-out validation
    Kocaguneli, Ekrem
    Menzies, Tim
    [J]. JOURNAL OF SYSTEMS AND SOFTWARE, 2013, 86 (07) : 1879 - 1890
  • [24] LaToza T. D., 2006, 28th International Conference on Software Engineering Proceedings, P492, DOI 10.1145/1134285.1134355
  • [25] Lavazza Luigi, 2013, 8 INT C SOFTW ENG AD
  • [26] Levesque Ghislain, 2001, P 11 INT WORKSH SOFT
  • [27] Lin CJ, 2016, 2016 INTERNATIONAL COMPUTER SYMPOSIUM (ICS), P555, DOI [10.1109/ICS.2016.0115, 10.1109/ICS.2016.114]
  • [28] Mathew George, 2016, ARXIV161203240
  • [29] McCabe T. J., 1976, IEEE Transactions on Software Engineering, VSE-2, P308, DOI 10.1109/TSE.1976.233837
  • [30] Menzies T, 2005, PROC INT CONF SOFTW, P587