Abstract
We present MONARCH, a framework-agnostic storage middleware that transparently employs storage tiering to accelerate Deep Learning (DL) training. It leverages existing storage tiers of modern supercomputers (i.e., compute node's local storage and shared parallel file system (PFS)), while considering the I/O patterns of DL frameworks to improve data placement across tiers. MONARCH aims at accelerating DL training and decreasing the I/O pressure imposed over the PFS. We apply MONARCH to TensorFlow and PyTorch, while validating its performance and applicability under different models and dataset sizes. Results show that, even when the training dataset can only be partially stored at local storage, MONARCH reduces TensorFlow's and PyTorch's training time by up to 28% and 37% for I/O-intensive models, respectively. Furthermore, MONARCH decreases the number of I/O operations submitted to the PFS by up to 56%.

Abstract
Geolocation data is fundamental to businesses relying on vehicles such as logistics and transportation. With the advance of the technology, collecting geolocation data become increasingly accessible and affordable, which raised new opportunities for business intelligence. This paper addresses the application of geolocation data for monitoring logistics processes, namely for detecting vehicle-based operations in real time. A stream of geolocation entries is used for inferring stationary events. Data from an international logistics company is used as a case study, in which operations of loading/unloading of goods are not only identified but also quantified. The results of the case study demonstrate the effectiveness of the solution, showing that logistics operations can be inferred from geolocation data. Further meaningful information may be extracted from these inferred operations using process mining techniques.

Abstract
-This work presents a study on how an immersive virtual environment's level of interaction and fidelity can affect the quality of experience (QOE) in a real estate context. Four versions of the virtual space were created with the level of interaction and the level of fidelity varying between them. The QoE dimensions considered in this work are user satisfaction, lighting quality, interior space quality, and interaction features. The sample comprises 28 participants, of which 21 are men and 7 are women, aged between 18 and 29 years. Results show that, overall, the level of fidelity is more relevant when the level of interaction is low, assuming the movement around the apartment is statistically higher in high-fidelity experiences.

Abstract
The paper addresses one of the techniques mor used by Advanced Persistent Threats attacks, phishing. The paper demonstrates the complexity of the technique, explains how attacks can be carried out, and presents defense techniques, and strategies against phishing attacks. The article also presents a summary description of what an Advanced Persistent Threat attack is. This description characterizes this type of attack.

Abstract
The Internet of Things (IoT) emerged to describe a network of connected things on a large scale to offer services to a large number of applications in different environments and domains. Middleware is software that seeks to facilitate the management and communication of all these things, providing the necessary functionalities to manage things, to discover, to compose services, and perform communication. For this reason, several proposals for middleware solutions for IoT have been developed. In this article, we conducted a systematic review of the literature to bring together middleware solutions for IoT, identifying the requirements and communication protocols used. In addition, we present some gaps and directions for future research in the development of IoT middleware.

Abstract

Internet-based interventions can be effective in the treatment of depression. However, internet-based interventions for older adults with depression are scarce and little is known about their feasibility and effectiveness.