“"In the age of Big Data, previously defining rules for finding all the uncountable patterns of events of interest, in complex and critical applications, is infeasible, if not impossible. A big challenge in this kind of scenario is to find previously unknown data that do not conform to the expected behavior. Anomaly Detection methods are techniques well-suited for tackling this category of problem due to its ability of identifying data that significantly deviate from an expected pattern, many times without any training example. Isolation Forest is a state-of-the-art technique in the unsupervised Anomaly Detection area, this, in addition of having a low computational cost. However, unsupervised Anomaly Detection many times suffer from high false-positive rate and high false-negative rate. Semi-supervised techniques can significantly improve the unsupervised algorithms with low human effort, and can also aggregate semantic knowledge into the models. Hybrid Isolation Forest is an Isolation Forest semi-supervised variation, which aggregates known anomalies, but only in a single class. This research proposed a Hybrid Isolation Forest-based model capable of aggregating known anomalies in distinct classes. The research work included a comprehensive literature aiming at identifying possibilities of aggregating expert feedback into Anomaly Detection techniques. The proposed model was accompanied by a very large set of experiments presented. Our proposed multiple anomaly classes semi-supervised model showed better performance in some distinct datasets and scenarios and showed an ability to significantly improve the underlying unsupervised algorithm.”
“Compreender o fenômeno da desinformação e sua disseminação pela internet tem sido uma tarefa cada vez mais desafiadora, mas é necessária, uma vez que os efeitos desse tipo de conteúdo têm seus impactos nas mais diversas áreas e geram cada vez mais impactos na sociedade. Uma forma de lidar com esse problema é o desenvolvimento de sistemas automatizados de verificação de fatos utilizando técnicas de inteligência computacional. O primeiro desafio destes sistemas, no contexto Brasil, é relacionado à disponibilidade de conjuntos de dados contendo notícias classificadas entre verdadeiras e falsas em língua portuguesa, para poder compreender a desinformação e dos seus subgrupos. O segundo desafio está relacionado à uma etapa essencial na geração de modelos de classificação que é o pré-processamento de dados e a identificação de dados que porventura possam ser ruidosos e estejam atrapalhando o processo de classificação. Este trabalho propõe: um novo Corpus contendo 19.446 notícias; a busca pelas melhores técnicas no processo de transformação, normalização e seleção de recursos; a exploração dos perfis da desinformação através de técnicas de agrupamento hierárquico e; a identificação de elementos ruidosos através da técnica t-SNE. Como resultado final, um modelo classificador com uma acurácia de 97,33% utilizando a técnica Random Forest foi proposto e implementado no Confere.ai, um projeto para automação de checagem de fatos.”
“Agile Software Development (ASD) is defined by the principles and values present in the Agile manifesto. One of the agile principles refers to the creation of value as: "our highest priority is to satisfy the customer through early and continuous delivery of valuable software". Despite this, initial ASD research followed other trends, addressing various agile methods, such as eXtreme Programming, Scrum, and Lean Software Development (LSD). However, in recent years, the value creation in ASD has shown to be a strong research trend and being widely discussed in the software industry. However, identifying practices that foster value creation becomes difficult since this concept has many aspects. Identifying these practices can change the mindset of agile teams, as research indicates that value creation is poorly understood from the agile team's point of view. Thus, this study aims to develop a guide composed of practices for creatingvalue in the context of ASD. This work applied some research methods: a Systematic Literature Review (SLR) and Grey Literature Review (GLR), seeking to identify practices for fostering value creation in ASD. As a result, some practices for promoting valuecreation have been identified in the grey literature; lastly, the interview technique was conducted with software development professionals to obtain evidence for constructing a guide composed of the practices to fostering value creation in ASD.”