Classification of Effort Estimation Methods

Abstract

In this chapter, we provide a brief overview of the existing software effort estimation methods. First, we identify common characteristics of existing methods and propose a schema for their classification. In the following sections, we provide a brief characterization of each class of methods and provide examples of typical methods belonging to a particular group.

Further Reading

• L.C. Briand and I. Wieczorek (2002), “Resource Modeling in Software Engineering,” in J.J. Marciniak (ed.) Encyclopedia of Software Engineering, 2nd Edition. Wiley & Sons.

  In their chapter, the authors provide a brief classification of effort estimation methods followed by a brief overview and comparative evaluation of selected methods.

• M. Jørgensen, B. Boehm, and S. Rifkin (2009), “Software Development Effort Estimation: Formal Models or Expert Judgment?” IEEE Software, vol. 26, no. 2, pp. 14–19.

  This publication presents a polemic on expert-based and data-driven effort estimation methods. The most prominent strengths and weaknesses of both approaches are presented by “gurus” of both approaches, Magne Jørgensen and Barry Boehm.

• R.W. Jensen, L.H. Putnam, and W. Roetzheim (2006), “Software Estimating Models: Three Viewpoints,” Cross-Talk: The Journal of Defense Software Engineering, vol. 19, no. 2, pp. 23–29.

  Authors discuss three different perspectives on software development effort and its estimation represented by three leading effort estimation tools: SEEM-SER, SLIM, and Cost-Xpert (where the latter one implements the COCOMO model). The article provides a brief overview of the origins, basic principles, and core equations implemented by the three tools.

• C. Schofield (1998), Non-Algorithmic Effort Estimation Techniques. Technical Report TR98-01, Department of Computing, Bournemouth University, UK.

  Author provides a brief overview of software effort approaches based on the techniques from the artificial intelligence domain, such as rule induction, fuzzy systems, regression trees, neural networks, and case-based reasoning. For each approach, the author summarizes its basic principles.

• J. Wen, S. Li, Z. Lin, Y. Hu, and C. Huang (2012), “Systematic literature review of machine learning based software development effort estimation models,” Information and Software Technology, vol. 54, no. 1, pp. 41–59.

  This article presents a comprehensive review of effort estimation methods based on machine-learning techniques such as Case-Based Reasoning (CBR), Artificial Neural Networks (ANN), Decision Trees (DT), Bayesian Networks (BN), Support Vector Regression (SVR), Genetic Algorithms (GA), Genetic Programming (GP), and Association Rules (AR). Authors look at the estimation accuracy of the considered methods (as reported in the related literature) and compare the performance of machine-learning methods against “conventional” non-machine-learning estimation models. Finally, the authors investigate particular project situations in which effort estimation based on machine-learning techniques should be favorable over other data-driven methods.

• M. Jørgensen (2004), “A review of studies on expert estimation of software development effort,” Journal of Systems and Software, vol. 70, no. 1–2, pp. 37–60.

  This article provides an extensive review of studies on expert estimation of software development effort. Based on the review, the author formulates 12 “best practice” guidelines for expert-based software effort estimation.

• J. S. Armstrong (2001), Principles of Forecasting: A Handbook for Researchers and Practitioners, 2nd Edition. Kluwer Academic Publishers, Dordrecht, The Netherlands.

  In Chap. 4, the authors discuss the most relevant aspects of estimation based on human judgment. The section “Improving Judgmental Forecasts” discusses procedures for improving experts’ forecasts. The section “Improving Reliability of Judgmental Forecasts” explains how the accuracy of expert forecasts is reduced when people use unreliable procedures to collect and analyze information. The section “Decomposition for Judgmental Forecasting and Estimation” describes how to decompose problems so that experts can make better estimates and forecasts. Finally, the section “Expert Opinions in Forecasting: Role of the Delphi Technique” provides an overview of forecasting with expert opinions (the authors use the Delphi procedure as a framework to integrate principles for improving expert forecasts).

• I. Ben-Gal (2007), “Bayesian Networks”, in F. Ruggeri, F. Faltin, and R. Kenett (eds.), Encyclopedia of Statistics in Quality & Reliability, Wiley & Sons.

  This short chapter of the Encyclopedia of Statistics and Reliability provides a well-written compact introduction to Bayes’ Theorem and Bayesian Belief Networks. It presents all relevant concepts in a concise and clear way and supports them with intuitive examples.

• Trendowicz (2013), Software Cost Estimation, Benchmarking, and Risk Assessment. Software Decision Makers’ Guide for Predictable Software Development. Springer Verlag.

  This book presents Cost Estimation, Benchmarking, and Risk Assessment (CoBRA), which combines human judgment and measurement data to systematically create custom-specific effort estimation models. Author provides a comprehensive specification of processes for developing a CoBRA effort model and for applying the model in a number of different project management scenarios. Moreover, the book provides a number of practical guidelines on applying these processes, based on industrial experiences regarding project effort estimation in general, and using the CoBRA method, in particular. Several real-world cases of applying the CoBRA method illustrate the practical use of the method.