Context: The speed with which the Fake News spread today has encouraged work in various areas to minimize the damage and the public insecurity caused by their proliferation. Objective: To characterize and analyze Fake News threat detection. Method: Systematic Mapping, since the area youthfulness still prevents a complete meta-analysis. Results: The most used algorithms were LSTM (17.14%), Naive-Bayes and Similarity Algorithm (11.43%). Conclusions: There is still the absence of more controlled experiments in the Big Data context. Fake News is a national security problem, requiring effective solutions to combat it. Situations like the Covid-19 virus (coronavirus) reinforce this fact.

Background: Software repositories provide large amount of data encompassing software changes throughout its evolution. Those repositories can be effectively used to extract and analyze pertinent information and derive conclusions related to the software history or its current snapshot. Objective: This work aims to investigate recent studies on Mining Software Repositories (MSR) approaches collecting evidences about software analysis goals (purpose, focus, and object of analysis), data sources, evaluation methods, tools, and how the area is evolving. Method: A systematic mapping study was performed to identify and analyze research on mining software repositories by analyzing five editions of Working Conference on Mining Software Repositories -- the main conference on this area. Results: MSR approaches have been used for many different goals, mainly for comprehension of defects, analysis of the contribution and behavior of developers, and software evolution comprehension. Besides, some gaps were identified with respect to their goals, focus, and data source type (e.g. lack of usage of comments to identify smells, refactoring, and issues of software quality). Regarding the evaluation method, our analysis pointed out to an extensive usage of some types of empirical evaluation. Conclusion: Studies of the MSR have focused on different goals, however there are still many research opportunities to be explored and issues associated with MSR that should be considered

The majority of fatal car crashes are caused by reckless driving. With the sophistication of vehicle instrumentation, reckless maneuvers, such as abrupt turns, acceleration, and deceleration, can now be accurately detected by analyzing data related to the driver-vehicle interactions. Such analysis usually requires very specific in-vehicle hardware and infrastructure sensors (e.g. loop detectors and radars), which can be costly. Hence, in this paper, we investigated if off-the-shelf smartphones can be used to online detect and classify the driver’s behavior in near real-time. To do so, we first modeled and performed an intrinsic evaluation to assess the performance of three outlier detection algorithms formulated as a data stream processing network which receives as input and processes data streams of smartphone and vehicle sensors. Next, we implemented a novel scoring mechanism based on online outlier detection to quantitatively evaluate drivers’ maneuvers as either cautious or reckless. Thus, we adapted a data mining mechanism which takes into account a sensor’s data rates and power to determine driver behavior in the scoring process. Finally, as the intrinsic evaluation does not necessarily reveal how well an algorithm will perform in a real-world scenario, we evaluated the algorithm that achieved the best result in a real-world case study to assess drivers’ driving behavior. Our results indicate that the algorithm performs quickly and accurately; the algorithm classifies driver behavior with 95.45% accuracy. Moreover, such results are obtained within 100 milliseconds of processing time on average.

Through the increase in deep learning study and use, in the last years there was a development of specific libraries for Deep Neural Network (DNN). Each one of these libraries has different performance results and applies different techniques to optimize the implementation of algorithms. Therefore, even though implementing the same algorithm and using different libraries, the performance of their executions may have a considerable variation. For this reason, developers and scientists that work with deep learning need scientific experimental studies that examine the performance of those libraries. Therefore, this paper has the aim of evaluating and comparing these two libraries: TensorFlow and PyTorch. We have used three parameters: Hardware utilization, hardware temperature and execution time in a context of heterogeneous platforms with CPU and GPU. We used the MNIST database for training and testing the LeNet Convolutional Neural Network (CNN). We performed a scientific experiment following the Goal Question Metrics (GQM) methodology, data were validated through statistical tests. After data analysis, we show that PyTorch library presented a better performance, even though the TensorFlow library presented a greater GPU utilization rate.

Stemming algorithms are commonly used during textual preprocessing phase in order to reduce data dimensionality. However, this reduction presents different efficacy levels depending on the domain that it is applied to. Thus, for instance, there are reports in the literature that show the effect of stemming when applied to dictionaries or textual bases of news. On the other hand, we have not found any studies analyzing the impact of radicalization on Brazilian judicial jurisprudence, composed of decisions handed down by the judiciary, a fundamental instrument for law professionals to play their role. Thus, this work presents two complete experiments, showing the results obtained through the analysis and evaluation of the stemmers applied on real jurisprudential documents, originating from the Court of Justice of the State of Sergipe. In the first experiment, the results showed that, among the analyzed algorithms, the RSLP (Removedor de Sufixos da Lingua Portuguesa) possessed the greatest capacity of dimensionality reduction of the data. In the second one, through the evaluation of the stemming algorithms on the legal documents retrieval, the RSLP-S (Removedor de Sufixos da Lingua Portuguesa Singular) and UniNE (University of Neuchâtel), less aggressive stemmers, presented the best cost-benefit ratio, since they reduced the dimensionality of the data and increased the effectiveness of the information retrieval evaluation metrics in one of analyzed collections.