Abstract
With the use of new computational technologies and novel methodologies for problem solving, recurring to the use of Group Decision Support Systems, normally the problem of incomplete information is marginalized as if we were living in an ideal world. Common sense tells us that in the precise time a decision is make it is impossible to know all the information regarding to it, however decisions must be made. What we propose is, in the ambit of the VirtualECare project, is a possible solution to decision making, through the use of Group Decision Support Systems, aware of incomplete information but, even so, able to make decisions based in the quality of the information and its source.

Abstract

Abstract
Memory is one's mental ability to store, to retain, and recall information, representing past and future, our dreams and/or expectations. However, as the human been ages, the capacity of remembering decreases as well the ability to pile up new memories, therefore affecting our quality-of-living and lowering our self- esteem. This configures a social and human dilemma. With the present work we intend to address some of these problems, in terms of a Personal Memory Assistant (PMA), in order to help the user to remember things and occurrences, making it in a proactive mode. It will also cater for some form of relaxation on the part of the user.

Abstract
In this work it is addressed the problem of knowledge evaluation in a VirtualECare Group Decision Supporting System (GDSS), in terms of an Multi-valued Extended Logic Programming language, which is aimed at sustaining online healthcare services. Indeed, reasoning with incomplete and uncertain knowledge have to be dealt with, due to the particular nature of the healthcare services, where the awful consequences of bad decisions, or lack of timely ones, demand for a responsible answer.

Abstract
Web growth has brought several problems to users. The large amount of information that exists nowadays in some particular Websites turns the task of finding useful information very difficult. Knowing users' visiting pattern is crucial to owners, so that they may transform or customize the Website. This problem originated the concept known as Adaptive Website: a Website that adapts itself for the purpose of improving the user's experience. This paper describes a proposal for a doctoral thesis. The main goal of this work is to follow a multi-agent approach for Web adaptation. The idea is that all knowledge administration about the Website and its users, and the use of that knowledge to adapt the site to fulfil user's needs, are made by an autonomous intelligent agent society in a negotiation environment. The complexity of the problem and the inherently distributed nature of the Web, which is an open, heterogeneous and decentralized network, are reasons that justify the multi-agent approach. It is expected that this approach enables real-time Web adaptation with a good level of benefit to the users.

Abstract
We propose an unsupervised method for propagating automatically extracted fine-grained topic labels among news items to improve their topic description for subsequent text classification procedure. This method compares vector representations of news items and assigns to each news item the label of its closest neighbour with a different topic label. Results obtained show that high precision can be achieved in propagating the top ranked topic label, and that 2-gram and 3-gram feature representations optimize the precision.