Abstract
Through IoT, humans and objects can be connected seamlessly, to guaranty improved quality of service (QoS). IoT-driven e-Health systems benefit from such rich network setting, to transmit health information and deliver health services. It is expected to grow massively in scale, but for that to happen, several issues need to be addressed, including security and trust. Edge computing paradigms, such as Fog computing and Cloudlet, are already popular in IoT based e-Health domain. Fog nodes are leveraged to reduce latency between IoT devices and remote cloud computing infrastructure. In this work, we explain how Mobile edge-clouds, which is a less popular edge computing paradigm, can be employed to achieve similar or lower latency, at a lower cost. We also propose a lightweight mechanism for security and fairness in e-Health protocols that are based on mobile edge-clouds and other paradigms. Detailed simulation experiments show that the proposed method is scalable and can efficiently mitigate attacks that are targeted at e-Health information and the network.

Abstract
Given an heterogeneous social network, can we forecast its future? Can we predict who will start using a given hashtag on twitter? Can we leverage side information, such as who retweets or follows whom, to improve our membership forecasts? We present TENSORCAST, a novel method that forecasts time-evolving networks more accurately than current state-of-the-art methods by incorporating multiple data sources in coupled tensors. TENSORCAST is (a) scalable, being linearithmic on the number of connections; (b) effective, achieving over 20% improved precision on top-1000 forecasts of community members; (c) general, being applicable to data sources with different structure. We run our method on multiple real-world networks, including DBLP, epidemiology data, power grid data, and a Twitter temporal network with over 310 million nonzeros, where we predict the evolution of the activity of the use of political hashtags.

Abstract
Counting subgraph occurrences is a hard but very important task in complex network analysis, with applications in concepts such as network motifs or graphlet degree distributions. In this paper we present a novel approach for this task that takes advantage of knowing that a large fraction of subgraph types does not appear at all on real-world networks. We describe a pattern-growth methodology that is able to iteratively build subgraph patterns that do not contain smaller non-occurring subgraphs, significantly pruning the search space. By using the g-trie data structure, we are able to efficiently only count those subgraphs that we are interested in, reducing the total computation time. The obtained experimental results are very promising allowing us to avoid the computation of up to 99.78% of all possible subgraph patterns. This showcases the potential of this approach and paves the way for reaching previously unattainable subgraph sizes. © Springer Nature Switzerland AG 2019.

Abstract

Abstract
Solving programming exercises is the best way to promote practice in computer programming courses and, hence, to learn a programming language. Meanwhile, programming courses continue to have an high rate of failures and dropouts. The main reasons are related with the inherent domain complexity, the teaching methodologies, and the absence of automatic systems with features such as intelligent authoring, profile-based exercise sequencing, content adaptation, and automatic evaluation on the student's resolution. At the same time, gamification is being used as an approach to engage learners' motivations. Despite its success, its implementation is still complex and based on ad-hoc and proprietary solutions. This paper presents PROud as a framework to inject gamification features in computer programming learning environments based on the usage data from programming exercises. This data can be divided into two categories: generic data produced by the learning environmentsuch as, the number of attempts and the duration that the students took to solve a specific exerciseor code-specific data produced by the assessment toolsuch as, code size, use memory, or keyword detection. The data is gathered in cloud storage and can be consumed by the learning environment through the use of a client library that communicates with the server through an established Application Programming Interface (API). With the fetched data, the learning environment can generate new gamification assets (e.g., leaderboards, quests, levels) or enrich content adaptations and recommendations in the inner components such as the sequencing tools. The framework is evaluated on its usefulness in the creation of a gamification asset to present dynamic statistics on specific exercises.

Abstract
Computer programming is a difficult subject that can only be mastered with lots of practice. It is therefore of primary importance to rise and retain students' engagement during a programming course, a task in which gamification has been proven as a competent method. Even though there are numerous reports on applying gamification to programming courses, there are no available open resources or dedicated platforms that could be used by programming teachers to gamily their courses, meeting both the requirements of being easy to adopt and leaving the decisions on the scope of the course and the level of gamification to the teachers themselves. In order to fulfill this gap, a consortium of four European institutions initiated a common project to develop open gamified programming exercises and interactive course materials for popular programming languages. In this paper, we report the results of the first stage of this work, which defined the range of gamification concepts to be covered within the developed framework and its evaluation by students.