Entity

Time filter

Source Type

Salvador, Brazil

Novais R.L.,Federal Institute of Bahia | Nunes C.,Pontifical Catholic University of Rio de Janeiro | Garcia A.,Pontifical Catholic University of Rio de Janeiro | Mendonca M.,Federal University of Bahia
IEEE International Conference on Software Maintenance, ICSM | Year: 2013

Program comprehension is an essential activity to perform software maintenance and evolution. Comprehensibility often encompasses the analysis of individual logical units, called features, which are often scattered through many program modules. Understanding how the feature code is implemented along the software evolution history is essential, for instance, to perform refactoring activities. However, existing tools do not provide means to comprehend the feature code evolution. To overcome this shortcoming, this paper presents a tool called Source Miner Evolution (SME) that provides multiple interactive and coordinated views to comprehend feature code evolution. SME implements a feature-sensitive comparison of multiple program versions. Our usability assessment with experienced developers indicated that SME allows them to efficiently perform recurring comprehension tasks on evolving feature code. The developers' performance was influenced by the combination of visual SME mechanisms, such as colors, tool tips and menu-popup interactions over the features' code elements. © 2013 IEEE. Source


Novais R.L.,Federal University of Bahia | Novais R.L.,Federal Institute of Bahia | De F. Carneiro G.,Federal University of Bahia | Simoes Jr. P.R.M.,Federal University of Bahia | Mendonca M.G.,Federal University of Bahia
Lecture Notes in Business Information Processing | Year: 2012

Software evolution is one of the most important topics in modern software engineering research. This activity requires the analysis of large amounts of data describing the current software system structure as well as its previous history. Software visualization can be helpful in this scenario, as it can summarize this complex data into easy to interpret visual scenarios. This paper presents an interactive differential approach for visualizing software evolution. The approach builds multi-view structural descriptions of a software system directly from its source code, and uses colors to differentiate it from any other previous version. This differential approach is highly interactive allowing the user to quickly brush over many pairs of versions of the system. As a proof of concept, we used the approach to analyze eight versions of an open source system and found out it was useful to quickly identify hot spot and code smell candidates in them. © 2012 Springer-Verlag. Source


Simas Filho E.F.,Federal University of Rio de Janeiro | Simas Filho E.F.,Federal Institute of Bahia | de Seixas J.M.,Federal University of Rio de Janeiro | CalOba L.P.,Federal University of Rio de Janeiro
Neurocomputing | Year: 2010

The nonlinear independent component analysis (NLICA) is an extension of the standard ICA model that does not restrict the mixing system to be linear. Different algorithms have been proposed to solve the NLICA problem, but, as the dimension of the problem increases, most of them present limitations such as poor accuracy and high computational cost. In this work, a novel structural model is proposed for the overdetermined NLICA problem (when there exist more sensors than sources), by adding a signal compaction block to the standard post-nonlinear (PNL) de-mixing model. The proposed methodology proves to be efficient in the feature extraction phase of a challenging high-dimensional online neural discrimination problem. © 2010 Elsevier B.V. Source


Novais R.,Federal University of Bahia | Novais R.,Federal Institute of Bahia | Nunes C.,Opus Research Group | Lima C.,Federal University of Bahia | And 4 more authors.
Proceedings - International Conference on Software Engineering | Year: 2012

Program comprehension is a key activity through maintenance and evolution of large-scale software systems. The understanding of a program often requires the evolution analysis of individual functionalities, so-called features. The comprehension of evolving features is not trivial as their implementations are often tangled and scattered through many modules. Even worse, existing techniques are limited in providing developers with direct means for visualizing the evolution of features' code. This work presents a proactive and interactive visualization strategy to enable feature evolution analysis. It proactively identifies code elements of evolving features and provides multiple views to present their structure under different perspectives. The novel visualization strategy was compared to a lightweight visualization strategy based on a tree-structure. We ran a controlled experiment with industry developers, who performed feature evolution comprehension tasks on an industrial-strength software. The results showed that the use of the proposed strategy presented significant gains in terms of correctness and execution time for feature evolution comprehension tasks. © 2012 IEEE. Source


Isidio De Lima J.J.,Federal University of Bahia | Caetano Da Silva J.,Federal Institute of Bahia | Rodriguez-Esquerre V.F.,Federal University of Bahia
Proceedings of SPIE - The International Society for Optical Engineering | Year: 2015

The absorption and reflection characteristics of multilayered nanoplasmonic gratings with sub wavelength sizes are analyzed in details by using an efficient finite element method. The multilayered structures are composed by several layers of nanoparticles of metals such as Silver, Gold and Aluminum embedded in dielectric such as amorphous silicon over a metallic substrate. The propagations characteristics for several geometrical configurations are obtained and a broadband reflection or absorption covering the near infrared wavelengths has been observed. The proposed nanoplasmonic structures have a great potential for applications in photovoltaic cells or polarizers by improving their reflection or absorption efficiency. Peaks of reflection or absorption larger than 80% were obtained and their performance over the near infrared can be improved by adequately tuning their geometrical parameters, the refractive index and thickness of the layers as well as the nanoparticles shape and size. © 2015 SPIE. Source

Discover hidden collaborations