Departamento de Ciência da Computação

Before the web, it was a challenge to historians to find historical sources for the research work. In recent decades, the problem was reversed. It became impractical to read all available sources and, thus, the data mining became fundamental in supporting the search in the domain. Here, we propose to the history domain, a semantic search engine, which considers the historical context and meaning of the queries. We show, for instance, that it is possible to identify relations between people and a particular historical event, or even statistics on a particular event. In this work, we indicate that the semantic search adapted to the domain is able to transform the research in the field of History.

Software tests have a high impact on the cost of software development. In practice, they are generally created at random and without any methodology, and do not have sufficient documentation. Commonly used approaches also perform the tests outside the application environment (e.g. web servers and containers). Besides, the test cases are usually restricted to target business components behavior, leaving a huge gap by not evaluating the presentation layer. Most of these practices can be explained by the overhead required to maintain manually the whole test artifacts. Applying a Model-based Approach (MBA), the creation and maintenance of test artifacts can be automated. This paper proposes a method that applies the Model-driven Architecture (MDA), a strategy of MBA, to determine the flow of test cases. The proposed method was based on the use of Unified Modeling Language (UML) activity diagrams. These diagrams allow determining the test flows and the objective of each activity, such as testing of business and presentation layers. Moreover, the generated test artifacts allow for performing the tests inside the application environment.

Collaborative Filtering (CF) is a well-known approach for Recommender Systems (RS). This approach extrapolates rating predictions from ratings given by user on items, which are represented by a user-item matrix filled with a rating ri,jri,j given by an user i on an item j. Therefore, CF has been confined to this data structure relying mostly on adaptations of supervised learning methods to deal with rating predictions and matrix decomposition schemes to complete unfilled positions of the rating matrix. Although there have been proposals to apply Machine Learning (ML) to RS, these works had to transform the rating matrix into the typical Supervised Learning (SL) data set, i.e  ., a set of pairwise tuples (x,y)(x,y), where y  is the correspondent class (the rating) of the instance x∈Rkx∈Rk. So far, the proposed transformations were thoroughly crafted using the domain information. However, in many applications this kind of information can be incomplete, uncertain or stated in ways that are not machine-readable. Even when it is available, its usage can be very complex requiring specialists to craft the transformation. In this context, this work proposes a domain-independent transformation from the rating matrix representation to a supervised learning dataset that enables SL methods to be fully explored in RS. In addition, our transformation is said to be straightforward, in the sense that, it is an automatic process that any lay person can perform requiring no domain specialist. Our experiments have proven that our transformation, combined with SL methods, have greatly outperformed classical CF methods.

This work proposes a new methodology for the Group Recommendation problem. In this approach we choose the Most Representative User (MRU) as the group medoid in a user space projection, and then generate the recommendation list based on his preferences. We evaluate our proposal by using the well-known dataset Movie lens. We have taken two different measures so as to evaluate the group recommender strategies. The obtained results seem promising and our strategy has shown an empirical robustness compared with the baselines in the literature.