Abstract

Abstract
This study investigates the main motivations, difficulties and good practices in the migration process of software development models from Waterfall to Agile experienced by IT Portuguese companies. For that, we adopted a quantitative methodology based on multiple case studies that allowed us to explore five research questions. In a first step, we propose to identify the main reasons that motivate the companies to adopt Agile development processes and we intend to characterize this migration process. We also explore a set of good practices that can be followed in this migration process and we suggest a set of measures that can be used to assess the success of projects in an Agile environment. Finally, we explore the organizational and management impacts with the introduction of Agile processes. Copyright © 2019, IGI Global.

Abstract
One of the main strategies for fostering the entrepreneurial mindset in students has been the development of training activities inside and outside classrooms using serious games. The chapter makes a brief description regarding five serious games initiatives that intend to promote and facilitate entrepreneurship education for different educational stages. Furthermore, the authors discuss the key benefits and challenges by the introduction of serious games in the learning process. The authors concluded by identifying that serious games have the potentiality of improving the learning process by providing an immersive experience for students that increases their motivation and potentiates the acquisition of multidisciplinary competencies. However, there are some challenges that should be properly considered and mitigated, namely in terms of assessing learning goals, integration with didactical systems, and support for different styles of learning.

Abstract
Technological advancements have created a media ecosystem in which traditional journalism sees its existence strongly threatened by the emergence of new players. Social networks have created a competitive environment that, whether due to its dispersion or its capillarity, has relegated the mainstream media to a secondary role in the media ecosystem. Ironically, the technologies that threaten traditional journalism are also those that can save it; provided they are used correctly. Journalism, weakened by the economic crisis and with increasingly smaller newsrooms, has artificial intelligence as an opportunity to recover a certain centrality in the media ecosystem. This paper studies AIDA, a project from the Brazilian television network Globo. This project looked to automation as a way to avoid errors and ambiguities in the news. The study of the AIDA case, complemented by interviews, presents the challenges to achieve the automatization of news regarding electoral polls.

Abstract
This paper describes the research, development and evaluation process of a solution based on computer vision for the detection and prevention of Computer Vision Syndrome, a type of eye fatigue characterized by the appearance of ocular symptoms during or after prolonged periods watching digital screens. The system developed targets users of computers and mobile devices, detecting and warning users to the occurrence of eye fatigue situations and suggesting corrective behaviours in order to prevent more complicated health consequences. The implementation resorts to machine learning techniques, using eye images datasets for training the eye state detection algorithm. OpenCV Lib was used for eye's segmentation and subsequent fatigue analysis. The final goal of the system is to provide users and health professionals with quality data analysis of eye fatigue levels, in order to raise awareness over accumulated stress and promote behaviour change. © 2019 IEEE.

Abstract
In this paper, a method for estimation of human pose is proposed, making use of ToF (Time of Flight) cameras. For this, a YOLO based object detection method was used, to develop a top-down method. In the first stage, a network was developed to detect people in the image. In the second stage, a network was developed to estimate the joints of each person, using the image result from the first stage. We show that a deep learning network trained from scratch with ToF images yields better results than taking a deep neural network pretrained on RGB data and retraining it with ToF data. We also show that a top-down detector, with a person detector and a joint detector works better than detecting the body joints over the entire image.