Since the harmful consequences of the online publication of fake news have emerged clearly, many research groups worldwide have started to work on the design and creation of systems able to detect fake news and entities that share it consciously. Therefore, manifold automatic, manual, and hybrid solutions have been proposed by industry and academia. In this article, we describe a deep investigation of the features that both from an automatic and a human point of view, are more predictive for the identification of social network profiles accountable for spreading fake news in the online environment. To achieve this goal, the features of the monitored users were extracted from Twitter, such as social and personal information as well as interaction with content and other users. Subsequently, we performed (i) an offline analysis realized through the use of deep learning techniques and (ii) an online analysis that involved real users in the classification of reliable/unreliable user profiles. The experimental results, validated from a statistical point of view, show which information best enables machines and humans to detect malicious users. We hope that our research work will provide useful insights for realizing ever more effective tools to counter misinformation and those who spread it intentionally.

Nowadays, several location-based services (LBSs) allow their users to take advantage of information from the Web about points of interest (POIs) such as cultural events or restaurants. To the best of our knowledge, however, none of these provides information taking into account user preferences, or other elements, in addition to location, that contribute to define the context of use. The provided suggestions do not consider, for example, time, day of week, weather, user activity or means of transport. This article describes a social recommender system able to identify user preferences and information needs, thus suggesting personalized recommendations related to POIs in the surroundings of the user's current location. The proposed approach achieves the following goals: (i) to supply, unlike the current LBSs, a methodology for identifying user preferences and needs to be used in the information filtering process; (ii) to exploit the ever-growing amount of information from social networking, user reviews, and local search Web sites; (iii) to establish procedures for defining the context of use to be employed in the recommendation of POIs with low effort. The flexibility of the architecture is such that our approach can be easily extended to any category of POI. Experimental tests carried out on real users enabled us to quantify the benefits of the proposed approach in terms of performance improvement.

In recent years, Automated Machine Learning (AutoML) has become increasingly important in Computer Science due to the valuable potential it offers. This is testified by the high number of works published in the academic field and the significant efforts made in the industrial sector. However, some problems still need to be resolved. In this paper, we review some Machine Learning (ML) models and methods proposed in the literature to analyze their strengths and weaknesses. Then, we propose their use—alone or in combination with other approaches—to provide possible valid AutoML solutions. We analyze those solutions from a theoretical point of view and evaluate them empirically on three Atari games from the Arcade Learning Environment. Our goal is to identify what, we believe, could be some promising ways to create truly effective AutoML frameworks, therefore able to replace the human expert as much as possible, thereby making easier the process of applying ML approaches to typical problems of specific domains. We hope that the findings of our study will provide useful insights for future research work in AutoML.