Salesian Journal on Information Systems

2

Authors: LINDEN, R.

This is a book review about "Ciência: da filosofia à publicação", from Gilson Volpato

6

Authors: TELLES, G.P., ALMEIDA, N.F., ALVAREZ,P. et al.

In this work we propose a simple method to obtain groups of homologous genes across multiple (k) organisms, called kGC. Our method takes as input all-against-all Blastp comparisons and produces groups of homologous sequences. First, homologies among groups of paralogs of all the k compared genomes are found, followed by homologies of groups among k - 1 genomes and so on, until groups belonging exclusively to only one genome, that is, groups of one genome not presenting strong similarities with any group of any other genome, are identified. We have used our method to determine homologous groups across six Actinobacterial complete genomes. To validate kGC, we first investigate the Pfam classification of the homologous groups, and after compare our results with those produced by OrthoMCL. Although kGC is much simpler than OrthoMCL it presented similar results with respect to Pfam classification.

8

Authors: FALCAO, H. S.; LOVATO, A. V.; SANTOS A. F. DOS; OLIVEIRA, L. S. et al.

This project intends to develop a prototype for the identification of free spots in open air parking area where there is a good aerial view without obstacles, allowing for the identification of occupied and free spots. We used image processing techniques and pattern recognition using Artificial Neural Networks (ANN). In order to help implement the prototype, we used Matlab. In order to simulate the parking area, we created a model so that we could acquire the images using a webcam, process them, train the neural network, classify the spots and finally, show the results. The results show that it is viable to apply pattern recognition through image capture to classify parking spots.

13

Authors: AFFONSO, F. J., CARNEIRO, M. C. V. S., RODRIGUES, E. L. L. et al.

This paper presents a reference model as an automated process to assist the adaptive software development at runtime, also known as Self-adaptive Systems (SaS) at runtime. This type of software has specific characteristics in comparison to traditional one, since it allows that changes (structural or behavioral) to be incorporated at runtime. Automated processes have been used as a feasible solution to conduct software adaptation at runtime by minimizing human involvement (developers) and speeding up the execution of tasks. In parallel, reference models have been used to aggregate knowledge and architectural artifacts, since they capture the systems essence in specific domains. However, presently no there is reference model based on reflection for the automation of software adaptation at runtime. In this scenario, this paper presents a reference model based on reflection, as an automated process, for the development of software systems that require adaptation at runtime. To show the applicability of the model, a case study was conducted and a good perspective to efficiently contribute to the area of SaS has been obtained.

6

Authors: MELO JR., A. ADÁN-COELLO, J. M

The Proportional Differentiation Model (PDM) is currently one of the main service proposals for the Next Generation Internet. This paper presents a new packet scheduling algorithm for implementing the PDM model using measurement windows and a mechanism of dynamic adjustment of packet delay estimation. Window Based Waiting-Time Priority Plus (WBWTP+), the proposed algorithm, is an evolution of the WBWTP algorithm that adjusts dynamically the relative weights of transmitted and waiting for transmission packets according to the current state of the system. The WBWTP+ delay estimator makes possible to accelerate or to delay the transmission of backlogged packets. Simulations performed to asses the performance of the WBWTP+ show that it presents significant improvement in the attendance of the PDM objective in relation to WBWTP in most scenarios, excepted when the link utilization rate is 100%. Even in that case the performance of WBWTP+ was better than that of WTP and PAD algorithms.

9

Authors: BOLINA, A. C., PEREIRA, D. A., ESMIN, A. A. A., PEREIRA, M. R. et al.

The main goal of data mining is to discover relevant information on digital content. The Apriori algorithm is widely used to this objective, but its sequential version has a low performance when execu- ted over large volumes of data. Among the solutions for this problem is the parallel implementation of the algorithm, and among the parallel implementations presented in the literature that based on Apriori, it highlights the DPA (Distributed Parallel Apriori) [10]. This paper presents the DMTA (Distributed Multithread Apriori) algorithm, which is based on DPA and exploits the parallelism level of threads in order to increase the performance. Besides, DMTA can be executed over heterogeneous hardware platform, using different number of cores. The results showed that DMTA outperforms DPA, presents load balance among processes and threads, and it is effective in current multicore architectures.

10

Authors: ANDRADE, A., JOST, T., RIGHI, R. et al.

Electronic Funds Transfer represents an upward trend, which fosters the proximity among consumers and suppliers. Each transaction is sent to a Computer Center, in charge of decoding, processing and returning the results as fast as possible. Particularly, the present article covers the GetNet Company day-by-day, focusing on one of their subsystems. In the article, we model the incoming transaction volume and the corresponding processing to answer the following questions: (i) how is the idleness of the company transaction system settings and what are the rates involed on that? (ii) Given an annual growth of 20% in the transaction volume, which modifications should be made in the current Computer Center to fulfill the need in terms of transactions until 2020? The tests were based on transactions execution logs during one day, which corresponds to the greater volume of 2011. As expected, the results show that the 10 machines composing GetNet system are overestimated for the current situation, which could support the operational load with only 4 machines. In addition, the current configuration could be sustained, regarding the growth predicted before, until the middle of 2017 without loss of transactions.