Abstract
Este trabalho apresenta um Mapeamento Sistemático da Literatura sobre evasão escolar, em que se buscou identificar tecnologias de mineração de dados e fatores indutores para evasão escolar, que vem sendo exploradas para desvendar as possíveis causas da evasão escolar. As buscas foram realizadas em quatro bases de dados científicas, com o objetivo de responder a seguinte questão de pesquisa: “Quais ferramentas, técnicas e fatores indutores vem sendo utilizados para desvendar possíveis causas da evasão escolar?”. Observou-se que a ferramenta Weka é a mais utilizada para auxiliar a desvendar as causas da evasão escolar. Entre as técnicas, destaca-se a utilização da classificação. Por fim, o mapeamento identificou que os principais trabalhos da área se concentram em estudar fatores relacionados às características individuais do aluno.

Abstract
MixAR, a full-stack system capable of providing visualization of virtual reconstructions seamlessly integrated in the real scene (e.g. upon ruins), with the possibility of being freely explored by visitors, in situ, is presented in this article. In addition to its ability to operate with several tracking approaches to be able to deal with a wide variety of environmental conditions, MixAR system also implements an extended environment feature that provides visitors with an insight on surrounding points-of-interest for visitation during mixed reality experiences (positional rough tracking). A procedural modelling tool mainstreams augmentation models production. Tests carried out with participants to ascertain comfort, satisfaction and presence/immersion based on an in-field MR experience and respective results are also presented. Ease to adapt to the experience, desire to see the system in museums and a raised curiosity and motivation contributed as positive points for evaluation. In what regards to sickness and comfort, the lowest number of complaints seems to be satisfactory. Models' illumination/re-lightning must be addressed in the future to improve the user's engagement with the experiences provided by the MixAR system.

Abstract
Enterprises have been challenged to adopt practices of sustainability to benefit shareholders and society with goods standing much beyond monetary profit or required by law. In combination with environmental and economic concerns, Corporate Social Responsibility (CSR) has become an option to leverage businesses with good reputation and to attract sustainability-aware market segments. In line with such a demand, this paper presents a systematic literature review of conceptual modeling studies referring explicitly to certifications, laws or norms of CSR. The more specific research goal of this work is to discover ontologies for representing CSR best practices, design patterns or policies. In total, 921 peer-reviewed papers were analyzed, from which only 17 were considered relevant for data extraction. The main result of this work is the identification of a research gap in explicit knowledge representation of CSR practices for Information Systems design, which ought to be filled to complement the (dominant) economic perspective on sustainability.

Abstract
Climate change is projected to be a key influence on crop yields across the globe. Regarding viticulture, primary climate vectors with a significant impact include temperature, moisture stress, and radiation. Within this context, it is of foremost importance to monitor soils' moisture levels, as well as to detect pests, diseases, and possible problems with irrigation equipment. Regular monitoring activities will enable timely measures that may trigger field interventions that are used to preserve grapevines' phytosanitary state, saving both time and money, while assuring a more sustainable activity. This study employs unmanned aerial vehicles (UAVs) to acquire aerial imagery, using RGB, multispectral and thermal infrared sensors in a vineyard located in the Portuguese Douro wine region. Data acquired enabled the multi-temporal characterization of the vineyard development throughout a season through the computation of the normalized difference vegetation index, crop surface models, and the crop water stress index. Moreover, vigour maps were computed in three classes (high, medium, and low) with different approaches: (1) considering the whole vineyard, including inter-row vegetation and bare soil; (2) considering only automatically detected grapevine vegetation; and (3) also considering grapevine vegetation by only applying a normalization process before creating the vigour maps. Results showed that vigour maps considering only grapevine vegetation provided an accurate representation of the vineyard variability. Furthermore, significant spatial associations can be gathered through (i) a multi-temporal analysis of vigour maps, and (ii) by comparing vigour maps with both height and water stress estimation. This type of analysis can assist, in a significant way, the decision-making processes in viticulture.

Abstract
To ensure high availability in large scale distributed systems, Conflict-free Replicated Data Types (CRDTs) relax consistency by allowing immediate query and update operations at the local replica, with no need for remote synchronization. State-based CRDTs synchronize replicas by periodically sending their full state to other replicas, which can become extremely costly as the CRDT state grows. Delta-based CRDTs address this problem by producing small incremental states (deltas) to be used in synchronization instead of the full state. However, current synchronization algorithms for delta-based CRDTs induce redundant wasteful delta propagation, performing worse than expected, and surprisingly, no better than state-based. In this paper we: 1) identify two sources of inefficiency in current synchronization algorithms for delta-based CRDTs; 2) bring the concept of join decomposition to state-based CRDTs; 3) exploit join decompositions to obtain optimal deltas and 4) improve the efficiency of synchronization algorithms; and finally, 5) experimentally evaluate the improved algorithms.

Abstract
The hypergraph-of-entity is a joint representation model for terms, entities and their relations, used as an indexing approach in entity-oriented search. In this work, we characterize the structure of the hypergraph, from a microscopic and macroscopic scale, as well as over time with an increasing number of documents. We use a random walk based approach to estimate shortest distances and node sampling to estimate clustering coefficients. We also propose the calculation of a general mixed hypergraph density based on the corresponding bipartite mixed graph. We analyze these statistics for the hypergraph-of-entity, finding that hyperedge-based node degrees are distributed as a power law, while node-based node degrees and hyperedge cardinalities are log-normally distributed. We also find that most statistics tend to converge after an initial period of accentuated growth in the number of documents. © 2020, Springer Nature Switzerland AG.