Abstract
In InfoVis design, visualizations make use of pre-attentive features to highlight visual artifacts and guide users' perception into relevant information during primitive visual tasks. These are supported by visual marks such as dots, lines, and areas. However, research assumes our pre-attentive processing only allows us to detect specific features in charts. We argue that a visualization can be completely perceived pre-attentively and still convey relevant information. In this work, by combining cognitive perception and psychophysics, we executed a user study with six primitive visual tasks to verify if they could be performed pre-attentively. The tasks were to find: horizontal and vertical positions, length and slope of lines, size of areas, and color luminance intensity. Users were presented with very simple visualizations, with one encoded value at a time, allowing us to assess the accuracy and response time. Our results showed that horizontal position identification is the most accurate and fastest task to do, and the color luminance intensity identification task is the worst. We believe our study is the first step into a fresh field called Incidental Visualizations, where visualizations are meant to be seen at-a-glance, and with little effort.

Abstract
Telecommunication networks evolution is driving the development of new applications for mobile devices. Some of these applications are resource-intensive and push computational and energy demands of mobile devices beyond the mobile hardware capabilities. In this context, Mobile Cloud Computing (MCC) architecture emerges as a solution for offloading mobile devices that allows to execute these applications in cloud datacenters thus reducing the processing demand in mobile devices. However, more demanding applications, e.g. interactive and realtime applications, are sensitive to processing and communications delay. For these applications, Mobile Edge Computing (MEC) can be used as an intermediary technology, providing computing and storage resources in the network edge. This paper presents a study carried out to evaluate the performance of MEC and MCC architectures when executing two applications, Fluid and FaceSwap, representative of real time and computing intensive applications. A set of scenarios were designed to quantify the performance of these architectures in different settings.

Abstract
Software is a critical factor in the reliability of computer systems. While the development of hardware is assisted by mature science and engineering disciplines, software science is still in its infancy. This situation is likely to worsen in the future with quantum computer systems. Actually, if quantum computing is quickly coming of age, with potential groundbreaking impacts on many different fields, such benefits come at a price: quantum programming is hard and finding new quantum algorithms is far from straightforward. Thus, the need for suitable formal techniques in quantum software development is even bigger than in classical computation. A lack of reliable approaches to quantum computer programming will put at risk the expected quantum advantage of the new hardware. This position paper argues for the need for a proper quantum software engineering discipline benefiting from precise foundations and calculi, capable of supporting algorithm development and analysis. © 2020 ACM.

Abstract
Escalonamento na arquitetura cloud e no paradigma fog continuam a apresentar alguns desafios aliciantes. Na cloud, segundo o conhecimento dos autores, ela é amplamente estudada e em muitas pesquisas é abordada na perspetiva de provedores de serviço. Na fog, é muito complexo e, existem poucos estudos. Procurando trazer contributos inovadores nas áreas de escalonamento de tarefas, neste artigo, propomos uma solução para o problema de escalonamento de aplicações móveis sensíveis ao contexto para o paradigma fog computing onde diferentes parâmetros de contexto são normalizados através da normalização Min-Max, as prioridades são definidas através da aplicação da técnica da Regressão Linear Múltipla (RLM) e o escalonamento é feito recorrendo a técnica de Otimização de Programação Não Linear Multi-objetivo (MONLP).;Scheduling in cloud architecture and in the fog paradigm continue to present some exciting challenges. In the cloud, according to the authors' knowledge, it is widely studied and in many researches, it is addressed from the perspective of service providers. In fog, it is very complex and there are few studies. Trying to bring innovative contributions in the areas of task scheduling, in this paper we propose a solution to the problem of context-aware scheduling of mobile applications for the fog computing paradigm, where different context parameters are normalized through Min-Max normalization, priorities are defined by applying the Multiple Linear Regression (MLR) technique and scheduling is performed using Multi-Objective Nonlinear Programming Optimization (MONLP) technique.

Abstract
Consuming 360 audiovisual content using a Head-Mounted Display (HMD) has become a standard feature for Immersive Virtual Reality (IVR). However, most applications rely only on visual and auditory feedback whereas other senses are often disregarded. The main goal of this work was to study the effect of tactile and olfactory stimuli on participants' sense of presence and cybersickness while watching a 360 video using an HMD-based IVR setup. An experiment with 48 participants and three experimental conditions (360 video, 360 video with olfactory stimulus, and 360 video with tactile stimulus) was performed. Presence and cybersickness were reported via post-test questionnaires. Statistical analysis showed a significant difference in presence between the control and the olfactory conditions. From the control to the tactile condition, mean values were higher but failed to show statistical significance. Thus, results suggest that adding an olfactory stimulus increases presence significantly while the addition of a tactile stimulus only shows a positive effect. Regarding cybersickness, no significant differences were found across conditions. We conclude that an olfactory stimulus contributes to higher presence and that a tactile stimulus, delivered in the form of cutaneous perception of wind, has no influence in presence. We further conclude that multisensory cues do not affect cybersickness.

Abstract
Background As of today, cancer is still one of the most prevalent and high-mortality diseases, summing more than 9 million deaths in 2018. This has motivated researchers to study the application of machine learning-based solutions for cancer detection to accelerate its diagnosis and help its prevention. Among several approaches, one is to automatically classify tumor samples through their gene expression analysis. Methods In this work, we aim to distinguish five different types of cancer through RNA-Seq datasets: thyroid, skin, stomach, breast, and lung. To do so, we have adopted a previously described methodology, with which we compare the performance of 3 different autoencoders (AEs) used as a deep neural network weight initialization technique. Our experiments consist in assessing two different approaches when training the classification model - fixing the weights after pre-training the AEs, or allowing fine-tuning of the entire network - and two different strategies for embedding the AEs into the classification network, namely by only importing the encoding layers, or by inserting the complete AE. We then study how varying the number of layers in the first strategy, the AEs latent vector dimension, and the imputation technique in the data preprocessing step impacts the network's overall classification performance. Finally, with the goal of assessing how well does this pipeline generalize, we apply the same methodology to two additional datasets that include features extracted from images of malaria thin blood smears, and breast masses cell nuclei. We also discard the possibility of overfitting by using held-out test sets in the images datasets. Results The methodology attained good overall results for both RNA-Seq and image extracted data. We outperformed the established baseline for all the considered datasets, achieving an average F(1)score of 99.03, 89.95, and 98.84 and an MCC of 0.99, 0.84, and 0.98, for the RNA-Seq (when detecting thyroid cancer), the Malaria, and the Wisconsin Breast Cancer data, respectively. Conclusions We observed that the approach of fine-tuning the weights of the top layers imported from the AE reached higher results, for all the presented experiences, and all the considered datasets. We outperformed all the previous reported results when comparing to the established baselines.