Abstract

Abstract
The increasing capabilities of smartphones is paving way to novel applications through the crowd-sourcing of these untapped resources, to form hyperlocal meshes commonly known as edge-clouds. While a relevant body-of-work is already available for the underlying networking, computing and storage facilities, security and privacy remain second class citizens. In this paper we present Panoptic, an edge-cloud system that enables the search for missing people, similar to the commonly known Amber alert system, in high density scenarios where wireless infrastructure might be limited (WiFi and LTE), e.g. concerts, while featuring privacy and security by design. Since the limited resources present in the mobile devices, namely battery capacity, Panoptic offers a computing offloading that tries to minimize data leakage while offering acceptable levels of performance. Our results show that it is achievable to run these algorithms in an edge-cloud configuration and that it is beneficial to use this architecture to lower data transfer through the wireless infrastructure while enforcing privacy. Results from our experimental evaluation show that the security layer does not impose a significant overhead, and only accounts for 2% of the total execution time for an edge cloud comprised by, but not limited to, 8 devices.

Abstract
The expansion of social networks has contributed to the propagation of information relevant to general audiences. However, this is small percentage compared to all the data shared in such online platforms, which also includes private/personal information, simple chat messages and the recent called ‘fake news’. In this paper, we make an exploratory analysis on two social networks to extract features that are indicators of relevant information in social network messages. Our goal is to build accurate machine learning models that are capable of detecting what is journalistically relevant. We conducted two experiments on CrowdFlower to build a solid ground truth for the models, by comparing the number of evaluations per post against the number of posts classified. The results show evidence that increasing the number of samples will result in a better performance on the relevancy classification task, even when relaxing in the number of evaluations per post. In addition, results show that there are significant correlations between the relevance of a post and its interest and whether is meaningfully for the majority of people. Finally, we achieve approximately 80% accuracy in the task of relevance detection using a small set of learning algorithms. © 2017 Copyright is held by the owner/author(s).

Abstract
Social networks are becoming a wide repository of information, some of which may be of interest for general audiences. In this study we investigate which features may be extracted from single posts propagated throughout a social network, and that are indicative of its relevance, from a journalistic perspective. We then test these features with a set of supervised learning algorithms in order to evaluate our hypothesis. The main results indicate that if a text fragment is pointed out as being interesting, meaningful for the majority of people, reliable and with a wide scope, then it is more likely to be considered as relevant. This approach also presents promising results when validated with several well-known learning algorithms.

Abstract
An overwhelming quantity of messages is posted in social networks every minute. To make the utilization of these platforms more productive, it is imperative to filter out information that is irrelevant to the general audience, such as private messages, personal opinions or well-known facts. This work is focused on the automatic classification of public social text according to its potential relevance, from a journalistic point of view, hopefully improving the overall experience of using a social network. Our experiments were based on a set of posts with several criteria, including the journalistic relevance, assessed by human judges. To predict the latter, we rely exclusively on linguistic features, extracted by Natural Language Processing tools, regardless the author of the message and its profile information. In our first approach, different classifiers and feature engineering methods were used to predict relevance directly from the selected features. In a second approach, relevance was predicted indirectly, based on an ensemble of classifiers for other key criteria when defining relevance-controversy, interestingness, meaningfulness, novelty, reliability and scope-also in the dataset. The first approach achieved a F (1)-score of 0.76 and an Area under the ROC curve (AUC) of 0.63. But the best results were achieved by the second approach, with the best learned model achieving a F (1)-score of 0.84 with an AUC of 0.78. This confirmed that journalistic relevance can indeed be predicted by the combination of the selected criteria, and that linguistic features can be exploited to classify the latter.

Abstract
The use of Social Media applications in educational settings has gained attention ever since educators became aware of their growing role in student's daily routine. These arise as privileged tools for social interactions, information exchange, collaborative knowledge building, immediate communication and persistent attention retaining, among others. Consequently, these tools impose themselves as complements to the profoundly established use of the traditional LMS, either being propelled by educators or requested by students. In previous research, we have already identified Facebook groups as one of the social media applications with the highest potential to foster the development of social learning communities. We have acknowledged the need to integrate Facebook groups and corresponding learning analytics into formal learning environments, such as the institutional LMS, and we have developed and presented a system which performs that integration. However, as the educational settings diversify in terms of pedagogy, coursework and student's profile and cultural background, we have identified the need to extend this integration to other social media tools, such as the instant messaging app WhatsApp, and to provide valuable learning analytics on its usage. Mobile, instant messaging based learning communities differ a lot from forum-alike communities, where threads, topics, conversations and interactions are easily trackable and, for instance, social network analysis can be conducted to profile members, roles and relationships. Therefore, research presented in this paper adds to previous consolidated work both on the technological and analytical dimensions. We address the challenges posed by the integration of WhatsApp based learning analytics in the LMS Moodle, starting by the fact that, unlike Facebook groups, WhatsApp does not provide an API for developers, nor any stream of structured data that can feed a real-time monitoring system. We then focus research on revealing an actual set of visual learning analytics that characterize a learning community of about thirty foreign master students, who used WhatsApp as a complementary tool during a semester. We discuss which type of learning analytics and corresponding visualizations best suit WhatsApp learning communities; what can educators draw from the analytics of such communities; and how that information can strengthen student assessment and profiling.