Abstract
A critical factor in immersive educational narratives is identification by students with the characters. In this work-in-progress analyzes the process of rendering characters from textual narratives into visual form by non-artists (i.e., instructors). We tried to match archetypes with their visual representation through the platforms: Pixton, Powtoon (both 2D) and The Sims4 (3D). The limitations of characterization can impact students' narrative immersion. As future work we intend to test with the target group and observe the improvements needed to increase identification and sense of immersion in the narrative.

Abstract
Research data on the activities of student teams in online learning environments are relevant for evaluating instructional methods, strategies, tools, and materials. For research data sharing and publication purposes, these personal data must be anonymized or pseudonymized as recommended by data protection and privacy policies. This paper addresses issues related to anonymizing and pseudonymizing student data on the Slack teamwork platform, one often employed in educational and business settings. Issues are discussed from two perspectives: data extraction and data transformation. Difficulties and challenges concerning data extraction and transformation are described. The complexities of these two processes are considered, and a starting point for developing more efficient methods is put forward.

Abstract
Virtual reality (VR) is still a field that is in constant development, and people are trying to use it to have a close representation of reality by creating immersive environments. However, despite the existence of some tools that have been adapted to work with VR, they require some experience to work with, and there is a considerable amount of resources that need to be spent to create and maintain the VR experiences, which prevents the adoption and use of all the benefits that VR can bring. This work proposes an architecture for an authoring tool that allows users to create their own virtual experiences without the need for an extensive understanding of it and use them to create a virtual training exercise. This paper uses a case study built upon a real training context scenario applied to the agroforestry field. To validate this proposal, a prototype was built and subject to usability and satisfaction tests that demonstrated the ease of understanding and learning of the interfaces and all the functionalities implemented.

Abstract
A key aspect of any hash map design is the problem of dynamically resizing it in order to deal with hash collisions. Compression in tree-based hash maps is the ability of reducing the depth of the internal hash levels that support the hash map. In this context, elasticity refers to the ability of automatically resizing the internal data structures that support the hash map operations in order to meet varying workloads, thus optimizing the overall memory consumption of the hash map. This work extends a previous lock-free hash trie map design to support elastic hashing, i.e., expand saturated hash levels and compress unused hash levels, such that, at each point in time, the number of levels in a path is adjusted, as closely as possible, to the set of keys that is stored in the data structure. To materialize our design, we introduce a new compress operation for hash levels, which requires redesigning the existing search, insert, remove and expand operations in order to maintain the lock-freedom property of the data structure. Experimental results show that elasticity effectively improves the search operation and, in doing so, our design becomes very competitive when compared to other state-of-the-art designs implemented in Java.

Abstract
Virtual reality applications aimed at worker training to train professionals are more common with the virtual reality advancements observed in this day and age. More companies search for ways to improve the efficiency and efficacy of their training programs, whilst also reducing training costs. There are several training applications found in the literature, but not many focus on the theme of computer assembly, and only a few have options like an observer's menu or a scoring system. With that in mind, a training application for assembling computer towers was designed. This article will focus on the application's functionalities, the results of questionnaires made to evaluate its quality and usability and potential future work. The study realized had good results and a good, varied sample of volunteers, with a score of 93.4% in the custom-made questionnaire, a cyber-sickness (SSQ) score of 26.53%, a usability score (SUS) of 90% and a satisfaction (ASQ) score of 17.67%, being that a higher score is better in custom made and SUS questionnaires, and a lower score is better in the SSQ and ASQ questionnaires. Although this project is just a proof of concept, it focuses on a theme that will certainly be explored soon, with the rise of demand for training applications, the ever-growing gamer market, and workstations for the design of virtual reality applications, like the one described on this paper.

Abstract
Permission managers in mobile devices allow users to control permissions requests, by granting of denying application's access to data and sensors. However, existing managers are ineffective at both protecting and warning users of the privacy risks of their permissions' decisions. Recent research proposes privacy protection mechanisms through user profiles to automate privacy decisions, taking personal privacy preferences into consideration. While promising, these proposals usually resort to a centralized server towards training the automation model, thus requiring users to trust this central entity. In this paper we propose a methodology to build privacy profiles and train neural networks for prediction of privacy decisions, while guaranteeing user privacy, even against a centralized server. Specifically, we resort to privacy-preserving clustering techniques towards building the privacy profiles, that is, the server computes the centroids (profiles) without access to the underlying data. Then, using federated learning, the model to predict permission decisions is learnt in a distributed fashion while all data remains locally in the users' devices. Experiments following our methodology show the feasibility of building a personalized and automated permission manager guaranteeing user privacy, while also reaching a performance comparable to the centralized state of the art, with an F1-score of 0.9.