Abstract
Crowd-sourcing the resources of mobile devices is a hot topic of research given the game-changing applications it may enable. In this paper we study the feasibility of using edge-clouds of mobile devices to reduce the load in traditional WiFi infrastructures for video dissemination applications. For this purpose, we designed and implemented a mobile application for video dissemination in sport venues that retrieves replays from a central server, through the access points in the WiFi infrastructure, into a smartphone. The fan's smartphones organize themselves into WiFi-Direct groups and exchange video replays whenever possible, bypassing the central server and access points. We performed a real-world experiment using the live TV feed for the Champions League game Benfica-Besiktas with the help of a group of volunteers using the application at the student's union lounge. The analysis of the logs strongly suggests that edge-clouds can significantly reduce the load in the access points at such large venues and improve quality of experience. Indeed, the edge-clouds formed were able to serve up to 80% of connected users and provide 56% of all downloads requested from within.

Abstract
In the last decade, technological advances and improved manufacturing processes have significantly dropped the price tag of mobile devices such as smartphones and tablets whilst augmenting their storage and computational capabilities. Their ubiquity fostered research on mobile edge-clouds, formed by sets of such devices in close proximity, with the goal of mastering their global computational and storage resources. The development of crowdsourcing applications that take advantage of such edge-clouds is, however, hampered by the complexity of network formation and maintenance, the intrinsic instability of wireless links and the heterogeneity of the hardware and operating systems in the devices. In this paper we present a middleware to deal with this complexity, providing a building block upon which crowd-sourcing applications may be built.We motivate the development of the middleware through a discussion of real-world applications, and present the middleware's architecture along with the associated components and current development status. The middleware takes form as a Java API for Android devices that allows for the establishment of links using heterogeneous communication technologies (e.g., Wifi-Direct, Bluetooth), and the combination of these links to form a logical edge-cloud network. On top of this functionality, services for edge computation, storage, and streaming are also being developed. © 2017 Association for Computing Machinery.

Abstract
Author name disambiguation in bibliographic databases has been, and still is, a challenging research task due to the high uncertainty there is when matching a publication author with a concrete researcher. Common approaches normally either resort to clustering to group author's publications, or use a binary classifier to decide whether a given publication is written by a specific author. Both approaches benefit from authors publishing similar works (e.g. subject areas and venues), from the previous publication history of an author (the higher, the better), and validated publicationauthor associations for model creation. However, whenever such an algorithm is confronted with different works from an author, or an author without publication history, often it makes wrong identifications. In this paper, we describe a feature extraction method that aims to avoid the previous problems. Instead of generally characterizing an author, it selectively uses features that associate the author to a certain publication. We build a Random Forest model to assess the quality of our set of features. Its goal is to predict whether a given author is the true author of a certain publication. We use a bibliographic database named Authenticus with more than 250, 000 validated author-publication associations to test model quality. Our model achieved a top result of 95.37% accuracy in predicting matches and 91.92% in a real test scenario. Furthermore, in the last case the model was able to correctly predict 61.86% of the cases where authors had no previous publication history. Copyright 2017 ACM.

Abstract

Abstract
Automated assessment is an essential part of eLearning. Although comparatively easy for multiple choice questions (MCQs), automated assessment is more challenging when exercises involve languages used in computer science. In this particular case, the assessment is more than just grading and must include feedback that leads to the improvement of the students’ performance. This paper presents ongoing work to develop Kora, an automated diagram assessment tool with enhanced feedback, targeted to the multiple diagrammatic languages used in computer science. Kora builds on the experience gained with previous research, namely: a diagram assessment tool to compute di erences between graphs; an IDE inspired web learning environment for computer science languages; and an extensible web diagram editor. Kora has several features to enhance feedback: it distinguishes syntactic and semantic errors, providing specialized feedback in each case; it provides progressive feedback disclosure, controlling the quality and quantity shown to each student after a submission; when possible, it integrates feedback within the diagram editor showing actual nodes and edges on the editor itself. © Hélder Correia, José Paulo Leal, and José Carlos Paiva

Abstract