Abstract
Classifying web queries into a set of categories is a crucial task to better understand the user's intent behind a query, contextualize their search and provide more relevant results to the user. However, web queries are typically short and ambiguous making the query classification a non-trivial problem. In this article, we present a new automatic approach for identifying and characterizing queries in the health domain. This method makes use of the search engine counts through a semantic similarity measure called Normalized Google Distance (NGD) combined with Support Vector Machines to classify queries into three dimensions: health-related, severity and semantic type. To evaluate our methods, we used two datasets in different languages, Portuguese and English, and built another for evaluating the last dimension. Overall, the results achieved were satisfactory. The most generic classification obtains better results than more specific ones. The NGD proved to be a valuable assent in query classification.

Abstract
Readability is a linguistic feature that indicates how difficult it is to read a text. Traditional readability formulas were made for the English language. This study evaluates their adequacy to the Portuguese language. We applied the traditional formulas in 10 parallel corpora. We verified that the Portuguese language had higher grade scores (less readability) in the formulas that use the number of syllables per words or number of complex words per sentence. Formulas that use letters by words instead of syllables by words output similar grade scores. Considering this, we evaluated the correlation of the complex words in 65 Portuguese school books of 12 schooling years. We found out that the concept of complex word as a word with 4 or more syllables, instead of 3 or more syllables as originally used in traditional formulas applied to English texts, is more correlated with the grade of Portuguese school books. In the end, for each traditional readability formula, we adapted it to the Portuguese language performing a multiple linear regression in the same dataset of school books. © Springer Nature Switzerland AG 2019.

Abstract
Sentiment analysis is a set of techniques that deal with the verification of sentiment and emotions in written texts. This introductory work aims to explore the combination of scores and polarities of sentiments (positive, neutral and negative) provided by different sentiment analysis tools. The goal is to generate a final score and its respective polarity from the normalization and arithmetic average scores given by those tools that provide a minimum of reliability. The texts analyzed to test our hypotheses were obtained from forum posts from participants in a massive open online course (MOOC) offered by Universidade Aberta de Portugal, and were submitted to four online service APIs offering sentiment analysis: Amazon Comprehend, Google Natural Language, IBM Watson Natural Language Understanding, and Microsoft Text Analytics. The initial results are encouraging, suggesting that the average score is a valid way to increase the accuracy of the predictions from different sentiment analyzers. © Springer Nature Switzerland AG 2019.

Abstract
O nosso estudo propôs a implementação de Gamificação numa Plataforma Social Académica de uma Universidade Virtual. Gamificação foi definida como o uso de elementos derivados dos jogos em contextos que não são jogos. Adaptando-a ao nosso contexto, quisemos verificar de que forma a sua implementação influenciava a Interação, Colaboração, Cooperação e Aprendizagem Social e ainda as presenças da Comunidade de Investigação, no espaço da Plataforma. Para isso usámos uma Metodologia de Design Based Research numa configuração de Métodos Mistos. Identificámos como foco de análise as formas sociais para a aprendizagem. Começámos por recolher opiniões dos utilizadores, usando entrevistas semi-estruturadas. Os resultados, informaram na construção de um protótipo “gamificado” construído com a ferramenta Elgg. Seguidamente efetuaram-se testes de usabilidade, recolhendo dados da performance e das opiniões dos utilizadores e foi efetuada uma implementação gamificada. Foram depois enviados questionários aos utilizadores e recolhidas estatísticas do uso. Os dados foram descritos, analisados e discutidos, segundo uma ordem cronológica. É discutido o impacto da implementação nas dimensões propostas e possibilidades de investigação futuras.;The article reports on the implementation of Gamification in a Social Platform of an Open University. Gamification is the use of game elements in non-game context. We adapted the implementation to context, to inspect the impact on the Interaction, Collaboration, Cooperation, Social Learning and on the Community of Inquiry. We used Design Based Research in a Mixed-Methods configuration. We looked at the Social Forms for Learning as a lens for the Elements. We present our study in a chronological order, explaining how each phase leads to the other. We first gathered information from the users of the academic platform via semi-structured interviews. We then designed a prototype with the Elgg engine, containing Gamification Elements. We made usability tests and gathered data from users´ performance and opinions. After we designed a second implementation, this time for students enrolled in different curricular units. This time we deployed a Survey and gathered information about the use of the platform. Lastly, we present our results discussing the impact on the proposed dimensions, and making suggestions for future research.

Abstract
Este artigo apresenta reflexões seminais acerca da utilização de tutores virtuais no sistema de e-learning da Universidade Aberta. O tutor virtual é uma analogia do tutor humano, adota uma interface antropomórfica com características emocionais, comumente denominado de avatar, que integra inteligência artificial e pode constituir uma abordagem interessante, tanto para o professor como para os estudantes, se se perspetivar o futuro dos ambientes típicos de ensino e educação a distância online e em rede e a crescente massificação da individualização do acompanhamento do estudante. Embora atraente de muitas formas, especialmente quando se leva em linha de conta a redução da carga da interação e do apoio que um professor (humano) deve fornecer nesses ambientes, várias questões se colocam visando a plena e efetiva integração de avatares como artefactos de mediação da tutoria virtual. Este artigo fornece um conjunto de perspetivas iniciais sobre este desafio, introduzindo o estado da arte no campo, enquanto apresenta alguns desenvolvimentos recentes e em experimentação na Universidade Aberta pela instanciação de dois avatares – a Maria e o João.;This article presents seminal reflections on the use of virtual tutors in the Aberta University e-learning system. The virtual tutor is an analogy of the human tutor, adopting an anthropomorphic interface with emotional characteristics, commonly called an avatar, which integrates artificial intelligence, and can be an interesting approach for both the teacher and the students, if the future of environments typical of online and networked distance learning and education, and the increasing massification of the individualization of student follow-up. Although attractive in many ways, especially when one takes into account the reduction of the interaction load and the support that a (human) teacher must provide in these environments, several questions are posed for the full and effective integration of avatars as mediation artifacts of virtual tutoring. This article provides a set of initial perspectives on this challenge, introducing the state of the art in the field, while presenting some recent developments and experimentation at the Aberta University by the instantiation of two avatars - Mary and John.

Abstract
This paper explores two editions of a MOOC delivered at Universidade Aberta's "AULAbERTA" (Open Class) platform. This platform offers different kinds of open learning paths and concerning specifically the MOOC learning path it is relevant to understand if the pedagogical model and the strategy achieve the desired results: (i) to improve education and enlarge the access to knowledge; (ii) to deliver high quality open and online education in Portuguese. The methodology used was based on a cross analysis of learner activity statistics against certain characteristics of the participants' profile like profession and motivations. Four categories of participants were identified and some aspects of their behaviours and engagement in the MOOC were examined, providing some insights concerning the level of participation and interaction. These elements can help us to understand some differences in the learning experiences connected with internal and external motivations and also how participants appreciate the online experience and course design.