Abstract
The telecommunications industry is highly competitive, as operators engage in fierce attacks, especially in bundled services, to acquire new customers originating high churn rate. The objective of this paper is to gain a comprehensive understanding of the factors influencing the switching of operators for bundled services among telecom operators. The paper includes a quantitative study with 3,004 customers utilizing bundled services from a Portuguese telecom operator. Employing covariance-based structural equation modeling and logit regression, the research shows that internet service, television service, and the service provided by the contact center exert the greatest impact on loyalty to the operator. In contrast, landline service has an insignificant effect, while loyalty has a negative influence on customer churn. This study offers telecommunications managers insights for identifying the main factors to retain customers and curbing customer defection. Additionally, it provides a framework for assessing customer experience within bundled telecom services, which is useful for researchers, managers and marketing practitioners alike.

Abstract
Context. The region of protoplanetary disks closest to a star (within 1-2 au) is shaped by a number of different processes, from accretion of the disk material onto the central star to ejection in the form of winds and jets. Optical and near-IR emission lines are potentially good tracers of inner disk processes if very high spatial and/or spectral resolution are achieved. Aims. In this paper, we exploit the capabilities of the VLTI-GRAVITY near-IR interferometer to determine the location and kinematics of the hydrogen emission line Br gamma. Methods. We present VLTI-GRAVITY observations of the Br gamma line for a sample of 26 stars of intermediate mass (HAEBE), the largest sample so far analysed with near-IR interferometry. Results. The Br gamma line was detected in 17 objects. The emission is very compact (in most cases only marginally resolved), with a size of 10-30 R*(1-5 mas). About half of the total flux comes from even smaller regions, which are unresolved in our data. For eight objects, it was possible to determine the position angle (PA) of the line-emitting region, which is generally in agreement with that of the inner-dusty disk emitting the K-band continuum. The position-velocity pattern of the Br gamma line-emitting region of the sampled objects is roughly consistent with Keplerian rotation. The exception is HD 45677, which shows more extended emission and more complex kinematics. The most likely scenario for the Br gamma origin is that the emission comes from an MHD wind launched very close to the central star, in a region well within the dust sublimation radius. An origin in the bound gas layer at the disk surface cannot be ruled out, while accreting matter provides only a minor fraction of the total flux. Conclusions. These results show the potential of near-IR spectro-interferometry to study line emission in young stellar objects.

Abstract
The concept of patterns and pattern languages, although very common in software nowadays, was first approached by Christopher Alexander, in the area of architecture, in the book A pattern language: towns, buildings, construction. However, it was only in 1980 that the term was adapted for software development, gaining its popularity in 1994. Despite the fact that the concept of patterns has been used in the area of software development for more than 40 years, there is still no consensus on the best method to validate patterns and patterns languages, and the existing methods are scattered in several different papers and research across the scientific community. As such, in this paper, we conduct a systematic literature review about the existing methods in the scientific community to validate patterns and pattern languages.

Abstract
One of the indicators of possible occurrences of cardiovascular diseases is the amount of coronary artery calcium. Recently, approaches using new technologies such as deep learning have been used to help identify these indicators. This work proposes a segmentation method for calcification of the coronary arteries that has three steps: (1) extraction of the ROI using U-Net with batch normalization after convolution layers, (2) segmentation of the calcifications and (3) removal of false positives using Modified U-Net with EfficientNet. The method uses histogram matching as preprocessing in order to increase the contrast between tissue and calcification and normalize the different types of exams. Multiple architectures were tested and the best achieved 96.9% F1-Score, 97.1% recall and 98.3% in the OrcaScore Dataset.

Abstract
Integration of introductory programming into higher education programs beyond computer science has lead to an increase in the failure and drop out rates of programming courses. In this context, programming instructors have explored new methodologies by introducing dynamic elements in the teaching-learning process, such as automatic code evaluation systems and gamification. Even though these methods have shown to be successful in improving students' engagement, they do not address all the existing problems and new strategies should be explored. In this work, we propose a new approach that combines the strengths of the Kumon method for personalized learning and progressive skill acquisition with the ability of online judge systems to provide automated assessment and immediate feedback. This approach has been used in teaching Programming I to students in several bachelor degrees and led to a 10% increase in exam approval rates compared to the baseline editions in which our Kumon-inspired methodology was not implemented. © Ivone Amorim, Pedro Baltazar Vasconcelos, and João Pedro Pedroso;

Abstract
The world economy depends on information systems. Business value resides in the data stored on Information Technology (IT) systems and on the processes run with those data. Malicious actors target these IT systems to extract value out of them using different cyberattacks. Denial of Service (DoS) attacks are a common and harmful method to render IT systems connected to the Internet and, consequently, business processes running on top of them, unavailable. Cybersecurity researchers in the industry and academia are in search of early warning and detection systems to detect and mitigate these DoS attacks. This article proposes a novel early detection strategy for DoS attacks based on the information provided by Horizontal Visibility Graphs (HVG) and Natural Visibility Graphs (VG) obtained out of the network flows monitored at the receiving end of the attack. From the primary results obtained, this strategy can detect a DoS attack under 70ms and 30 packets. These results point out that the application HVGs and VGs is a relevant and promising research direction to prevent or mitigate the impact of a DoS attack.