Abstract
In recent years, we have been witnessing a real explosion of information, due in large part to the development in Information and Knowledge Technologies (ICTs). As in-formation is the raw material for the discovery of knowledge, there has been a rapid growth, both in the scientific community and in ICT itself, in the approach and study of the phenomenon called Big Data (BD) [1]. The concept of Smart Grids (SG) has emerged as a way of rethinking how to produce and consume energy imposed by economic, political and ecological issues [2]. To become a reality, SGs must be sup-ported by intelligent and autonomous IT systems, to make the right decisions in real time. Knowledge needed for real-time decision-making can only be achieved if SGs are equipped with systems capable of efficiently managing all the information sur-rounding their ecosystem. Multi-Agent systems have been increasingly used from this purpose. This work proposes a system for the management of information in the context of agent based SG to enable the monitoring, in real time, of the events that occur in the ecosystem and to predict upcoming events.

Abstract
Predicting the inherent traffic behaviour of a network is an essential task, which can be used for various purposes, such as monitoring and managing the network's infrastructure. However, the recent surge of dynamic environments, such as Internet of Things and Cloud Computing have hampered this task. This means that the traffic on these networks is even more complex, displaying a nonlinear behaviour with specific aperiodic characteristics during daily operation. Traditional network traffic predictors are usually based on large historical data bases which are used to train algorithms. This may not be suitable for these highly volatile environments, where the strength of the force exerted in the interaction between past and current values may change quickly with time. In light of this, a taxonomy for network traffic prediction models, including the review of state of the art, is presented here. In addition, an analysis mechanism, focused on providing a standardized approach for evaluating the best candidate predictor models for these environments, is proposed. These contributions favour the analysis of the efficacy and efficiency of network traffic prediction among several prediction models in terms of accuracy, historical dependency, running time and computational overhead. An evaluation of several prediction mechanisms is performed by assessing the Normalized Mean Square Error and Mean Absolute Percent Error of the values predicted by using traces taken from two real case studies in cloud computing.

Abstract
The PERFoRM project, an innovation action promoted within the scope of the EU Horizon 2020 program, advocates the use of an Industrie 4.0 compliant system architecture for the seamless reconfiguration of robots and machinery. The system architecture re-uses the innovative results from previous successful R&D projects on distributed control systems domain, such as SOCRADES, IMC-AESOP, GRACE and IDEAS. This paper, after describing the main pillars of the PERFoRM system architecture, focuses on mapping the system architecture into four industrial use cases aiming to validate the system architecture design before its deployment in the real environments. © Springer International Publishing AG 2017.

Abstract
This paper presents the study of variable-speed Pump Storage Powerplant (PSP) in the Portuguese power system. It evaluates the progressive integration in three major locations and compares the power system performance following a severe fault event with consequent disconnection of non-Fault Ride-through (FRT) compliant Wind Farms (WF). To achieve such objective, a frequency responsive model was developed in PSS/E and was further used to substitute existing fixed-speed PSP. The results allow identifying a clear enhancement on the power system performance by the presence of frequency responsive variable-speed PSP, especially for the scenario presented, with high level of renewables integration.

Abstract
This paper describes a preliminary study for producing and distributing a large-scale database of embeddings from the Portuguese Twitter stream. We start by experimenting with a relatively small sample and focusing on three challenges: volume of training data, vocabulary size and intrinsic evaluation metrics. Using a single GPU, we were able to scale up vocabulary size from 2048 words embedded and 500K training examples to 32768 words over 10M training examples while keeping a stable validation loss and approximately linear trend on training time per epoch. We also observed that using less than 50% of the available training examples for each vocabulary size might result in overfitting. Results on intrinsic evaluation show promising performance for a vocabulary size of 32768 words. Nevertheless, intrinsic evaluation metrics suffer from over-sensitivity to their corresponding cosine similarity thresholds, indicating that a wider range of metrics need to be developed to track progress.

Abstract
Software developers have always found it difficult to adopt Field-Programmable Gate Arrays (FPGAs) as computing platforms. Recent advances in HLS tools aim to ease the mapping of computations to FPGAs by abstracting the hardware design effort via a standard OpenCL interface and execution model. However, OpenCL is a low-level programming language and requires that developers master the target architecture in order to achieve efficient results. Thus, efforts addressing the generation of OpenCL from high-level languages are of paramount importance to increase design productivity and to help software developers. Existing approaches bridge this by translating MATLAB/Octave code into C, or similar languages, in order to improve performance by efficiently compiling for the target hardware. One example is the MATISSE source-to-source compiler, which translates MATLAB code into standard-compliant C and/or OpenCL code. In this paper, we analyse the viability of combining both flows so that sections of MATLAB code can be translated to specialized hardware with a small amount of effort, and test a few code optimizations and their effect on performance. We present preliminary results relative to execution times, and resource and power consumption, for two OpenCL kernels generated by MATISSE, and manual optimizations of each kernel based on different coding techniques. © 2018 The authors and IOS Press.