Giancarlo Succi

A comparative analysis of the efficiency of change metrics and static code attributes for defect prediction

In this paper we present a comparative analysis of the predictive power of two different sets of metrics for defect prediction. We choose one set of product related and one set of process related software metrics and use them for classifying Java files of the Eclipse project as defective respective defect-free. Classification models are built using three common machine learners: logistic regression, naive Bayes, and decision trees. To allow different costs for prediction errors we perform cost-sensitive classification, which proves to be very successful: >75% percentage of correctly classified files, a recall of >80%, and a false positive rate <30%. Results indicate that for the Eclipse data, process metrics are more efficient defect predictors than code metrics.

An empirical study of open-source and closed-source software products

We describe an empirical study of open-source and closed-source software projects. The motivation for this research is to quantitatively investigate common perceptions about open-source projects, and to validate these perceptions through an empirical study. We investigate the hypothesis that open-source software grows more quickly, but does not find evidence to support this. The project growth is similar for all the projects in the analysis, indicating that other factors may limit growth. The hypothesis that creativity is more prevalent in open-source software is also examined, and evidence to support this hypothesis is found using the metric of functions added over time. The concept of open-source projects succeeding because of their simplicity is not supported by the analysis, nor is the hypothesis of open-source projects being more modular. However, the belief that defects are found and fixed more rapidly in open-source projects is supported by an analysis of the functions modified. We find support for two of the five common beliefs and conclude that, when implementing or switching to the open-source development model, practitioners should ensure that an appropriate metrics collection strategy is in place to verify the perceived benefits.

Project management in plan-based and agile companies

Agile methods are a recent set of development techniques that apply a human-centered approach to software production. The agile approach aims to deliver high-quality products faster, producing satisfied customers. We conducted an empirical study to investigate whether agile methods change and improve project management practices in software companies. Survey results show that adopting agile methods appears to improve management of the development process and customer relationships. This article has given a first analysis of the advantages and disadvantages of adopting agile methods from a project management perspective.

Identification of defect-prone classes in telecommunication software systems using design metrics

The goal of this paper is to investigate the relation between object-oriented design choices and defects in software systems, with focus on a real-time telecommunication domain. The design choices are measured using the widely accepted metrics suite proposed by Chidamber and Kemerer for object oriented languages [S.R. Chidamber, C.F. Kemerer, A metrics suite for object oriented design, IEEE Transactions on Software Engineering 20 (6) (1994) 476-493]. This paper reports the results of an extensive case study, which strongly reinforces earlier, mainly anecdotal, evidence that design aspects related to communication between classes can be used as indicators of the most defect-prone classes. Statistical models applicable for the non-normally distributed count data are used, such as Poisson regression, negative binomial regression, and zero-inflated negative binomial regression. The performances of the models are assessed using correlations, dispersion coefficients and Alberg diagrams. The zero-inflated negative binomial regression model based on response for a class shows the best overall ability to describe the variability of the number of defects in classes.

Does refactoring improve reusability

The improvement of the software development process through the development and utilization of high quality and reusable software components has been advocated for a long time. Agile Methods promote some interesting practices, in particular the practice of refactoring, which are supposed to improve understandability and maintainability of source code. In this research we analyze if refactoring promotes ad-hoc reuse of object-oriented classes by improving internal quality metrics. We conduct a case study in a close-to industrial, agile environment in order to analyze the impact of refactoring on internal quality metrics of source code. Our findings sustain the hypothesis that refactoring enhances quality and reusability of – otherwise hard to reuse – classes in an agile development environment. Given such promising results, additional experimentation is required to validate and generalize the results of this work.

Managing uncertainty in requirements: a survey in documentation-driven and agile companies

This paper investigates commonalities and differences between agile and documentation-driven approaches in managing uncertainty in requirement gathering. The research method is a survey collected interviewing sixteen project managers of Italian software companies, 8 using agile methods, and 8 using documentation-driven methods. The results show that agile and document-driven companies consider in a different way the problem of changing requirements and the related uncertainty; thus, they manage differently requirements gathering and the relationship with the customer

An Empirical Exploration of the Distributions of the Chidamber and Kemerer Object-Oriented Metrics Suite

The object-oriented metrics suite proposed by Chidamber and Kemerer (CK) is a measurement approach towards improved object-oriented design and development practices. However, existing studies evidence traces of collinearity between some of the metrics and low ranges of other metrics, two facts which may endanger the validity of models based on the CK suite. As high correlation may be an indicator of collinearity, in this paper, we empirically determine to what extent high correlations and low ranges might be expected among CK metrics.To draw as much general conclusions as possible, we extract the CK metrics from a large data set (200 public domain projects) and we apply statistical meta-analysis techniques to strengthen the validity of our results. Homogenously through the projects, we found a moderate (∼0.50) to high correlation (>0.80) between some of the metrics and low ranges of other metrics.Results of this empirical analysis supply researchers and practitioners with three main advises: a) to avoid the use in prediction systems of CK metrics that have correlation more than 0.80 b) to test for collinearity those metrics that present moderate correlations (between 0.50 and 0.60) c) to avoid the use as response in continuous parametric regression analysis of the metrics presenting low variance. This might therefore suggest that a prediction system may not be based on the whole CK metrics suite, but only on a subset consisting of those metrics that do not present either high correlation or low ranges.

Effort Prediction in Iterative Software Development Processes -- Incremental Versus Global Prediction Models

Estimation of development effort without imposing overhead on the project and the development team is of paramount importance for any software company. This study proposes a new effort estimation methodology aimed at agile and iterative development environments not suitable for description by traditional prediction methods. We propose a detailed development methodology, discuss a number of architectures of such models (including a wealth of augmented regression models and neural networks) and include a thorough case study of Extreme Programming (XP) in two semi-industrial projects. The results of this research evidence that in the XP environment under study the proposed incremental model outperforms traditional estimation techniques most notably in early phases of development. Moreover, when dealing with new projects, the incremental model can be developed from scratch without resorting itself to historic data.

Data description

The study is concerned with a granular data description in which we propose a characterization of numeric data by a collection of information granules so that the key structure of the data, their topology and essential relationships are described in the form of a family of fuzzy sets - information granules. A comprehensive design process is introduced in which we show a two-phase development strategy: first, numeric prototypes are built with the use of Fuzzy C-Means (FCM) that is followed by their augmentation resulting in a collection of information granules. In the design of information granules we engage the fundamental ideas of Granular Computing, especially the principle of justifiable granularity. A series of experiments is presented to visualize the key steps of the construction of information granules.

Failure prediction based on log files using Random Indexing and Support Vector Machines

Research problem: The impact of failures on software systems can be substantial since the recovery process can require unexpected amounts of time and resources. Accurate failure predictions can help in mitigating the impact of failures. Resources, applications, and services can be scheduled to limit the impact of failures. However, providing accurate predictions sufficiently ahead is challenging. Log files contain messages that represent a change of system state. A sequence or a pattern of messages may be used to predict failures. Contribution: We describe an approach to predict failures based on log files using Random Indexing (RI) and Support Vector Machines (SVMs). Method: RI is applied to represent sequences: each operation is characterized in terms of its context. SVMs associate sequences to a class of failures or non-failures. Weighted SVMs are applied to deal with imbalanced datasets and to improve the true positive rate. We apply our approach to log files collected during approximately three months of work in a large European manufacturing company. Results: According to our results, weighted SVMs sacrifice some specificity to improve sensitivity. Specificity remains higher than 0.80 in four out of six analyzed applications. Conclusions: Overall, our approach is very reliable in predicting both failures and non-failures.