Abstract
The outermost regions, characterized by their remoteness, insularity, small size, difficult topography and climate, and economic dependence on certain sectors, are increasingly looking to invest in new technologies. The application of e-marketing, which arose as a variant of conventional marketing, has demonstrated over recent years more and more it's potential. With new developments in technology, the companies' promotion possibilities have increased, so as their e-marketing actions. The e-marketing new techniques supported in the Internet allowed to improve customer relations, distribution, promotion of products and services and increase firms income. The wide possibilities of e-marketing, associated with their low cost and ease of implementation, are key factors in this choice. Through this study it was possible.

Abstract

Abstract
The development platform of a digital game is frequently defined before any game design document. Without a proper examination, this could exclude other viable solutions. Although the online virtual world Second Life is sometimes erroneously denominated a game, it can be used as a development platform for many purposes. This paper analyses game creation within the Second Life virtual world and possible approaches for the implementation of game assets and game logic. There is also an analysis of what game types could be more suited for development under this platform. Experiences learned from the developed prototype are shown and hints are given as to evaluate different techniques and expose best practices. © 2011 AISTI.

Abstract
We provide an extension to Merton's famous continuous time model of optimal consumption and investment, in the spirit of previous works by Pliska and Ye, to allow for a wage earner to have a random lifetime and to use a portion of the income to purchase life insurance in order to provide for his estate, while investing his savings in a financial market consisting of one risk-free security and an arbitrary number of risky securities whose diffusive terms are driven by a multi-dimensional Brownian motion. The wage earner's problem is to find the optimal consumption, investment, and insurance purchase decisions in order to maximize expected utility of consumption and of the size of the estate in the event of premature death, and of the size of the estate at the time of retirement. Dynamic programming methods are used to obtain explicit solutions for the case of constant relative risk aversion utility functions, and new results are presented together with the corresponding economic interpretations.

Abstract
Relational reinforcement learning is a promising branch of reinforcement learning research that deals with structured environments. In these environments, states and actions are differentiated by the presence of certain types of objects and the relations between them and the objects that are involved in the actions. This makes it ultimately suited for tasks that require the manipulation of multiple, interacting objects, such as tasks that a future house-holding robot can be expected to perform like cleaning up a dinner table or storing away done dishes. However, the application of relational reinforcement learning to robotics has been hindered by assumptions such as discrete and atomic state observations. Typical robotic observation systems work in a streaming setup, where objects are discovered and recognized and their placement within their surroundings is determined in a quasi continuous manner instead of a state based one. The resulting information stream can be compared to a set of multiple inter-related data streams. In this paper, we propose an adaptive windowing strategy for generating a stream of learning examples and enabling relational learning from this kind of data. Our approach is independent from the learning algorithm and is based on a gradient search over the space of parameter values, i.e., window sizes, guided by the estimation of the testing error. The proposed algorithm performs online and is data driven and flexible. To the best of our knowledge, this is the first work addressing this problem. Our ideas are empirically supported by an extensive experimental evaluation in a controlled setup using artificial data. © 2011 IEEE.

Abstract
The e-learning systems are, currently, extremely useful tools in education, helping to reduce the distance between students and teachers and providing better methods for monitoring the learning process. These systems can play a particularly important role in programming courses, due to the need that students master new programming languages to code algorithms. This is usually achieved by solving a high number of exercises that allow them to learn from their mistakes. The overall learning process can be improved by the use of an automated assessing tool, since it is very important to present the student with the feedback he needs to refine his programming skills. Moreover, teachers will benefit from not having to carry these tasks manually, which allows them to provide immediate feedback and more frequent assessment. In this paper a new Computer-Based Assessment (CBA) system is proposed, based on Moodle and the DOMjudge platform, and developed for supporting the programming components of the courses of the Faculty of Engineering of the University of Porto (FEUP). From an initial prototype that had been implemented with a set of essential features, improvements were introduced in order to provide more complex functionalities that increase the effectiveness of this tool. This paper starts by presenting a study of state-of-the-art CBA systems in order to compare the most advanced features they implement on top of the basic functionality of a CBA system, followed by an identification of the main points which needed to be addressed to assure its effectiveness. Between the advanced features outlined, the ones that distinguish themselves are the plagiarism detection and the static analysis of code quality. The former aims to motivate students to learn by themselves, while the latter to help students improve their programming style and skill. The architecture of the CBA system is also presented, bearing in mind modularity and scalability. The system currently supports grading parameters such as the definition of the start and end time, maximum number of submissions and configuring the level of feedback on a per exercise basis. This section covers details of the CBA system implementation as well as the advanced modules developed, with emphasis on the plagiarism detection, static analysis of code quality, feedback system and assessment creation possibilities. The CBA system was used in an assignment activity proposed by the teachers of a course unit and a report on this case study is also presented in the following section, yielding important information on the feedback of teachers and students, as well as some limitations that were found on system hardware. Since the developed modules were implemented, tested and integrated with extensibility in mind, the section also provides guidelines on how they can be expanded and how to improve a few incomplete advanced features. Finally, it was concluded that the CBA system can be used in a real working environment, providing it is installed in production servers and all users are warned that the platform is still in test. However, in order to further unleash the educational potential of the system, the advanced modules could benefit from some optimizations and extended functionalities.