4

this project will address. We think that designing and developing such a platform will allow citizens to trust their administrators more and will help the policy making process to be more adherent to citizens' will.

1.1.2 Objectives

The main goal of this project is to build an extensible, secure and transparent web platform (or toolbox) for e-participation and e government tools. The platform will also provide the functionality of collecting, learning and synthesizing knowledge coming from a broad fo1ksonomy, i.e. the practice and method of collaboratively creating and managing tags to annotate and categorize content.

We also aim to define a general framework, which allows both i) the governments to build trusted application (by relying on platform services) and ii) any citizen to develop tools for e-participation, as well as to define new services to be used by other user-defined tools.

We will also give examples of e-participation and e-government tools relying on our platform.

1.1.2.1 Knowledge

Recommendation system. We will provide the user with personalized recommendations based on his profile, comparable to recommender systems such as the ones in Amazon or lastFM. But unlike these examples we do not know anything in advance about the content the users will add to EUGAGER. This is why the recommender system has to be flexible in this regard. We will provide different kind of dimensions the user may choose from to get recommendations. E.g. he may search for content he agrees to or content he is interested in. This is not necessarily the same: a citizen does not have to agree to or like a proposal to find it interesting or important and therefore get it recommended, but on the other hand he should agree to or like an application to get it recommended,

By letting the user choose which dimensions are important for every search, we provide him with the necessary flexibility to filter new content of EUGAGER in different ways. Of course, we will also offer predefined templates for standard filtering problems and the possibility to create and share such search profiles in the community.

Tagging system (Knowledge learning and tracking). As citizens input their opinions on the net, we need to produce a search tool that permits them to efficiently retrieve existing opinions. This is to avoid dispersing their efforts through many different, but equivalent, opinions. And this is also to see different opinions on a topic, with the aim of integrating part of them, to gain a wider endorsement. There are many ways to produce such a search tool. Our aim is to use a bottom-up approach, which results in an organization of the information that mirrors the culture of the people that are using them.

We should thus produce three separate, but intertwined search tools. Each opinion that has been inserted should be linked with related opinions. There should also be a search box, where the user can search directly all the opinions inserted, and finally the opinions should be clustered in topics of knowledge. Those topics of knowledge should be emergent from the opinions inserted, and from the keywords inserted by the users. As such, the way they will appear will mirror and give information on the group of people that generated them.

Knowledge synthesis. Citizens need to be able not only to voice their opinion, but also to endorse each others' positions; to synthesize them into a coherent whole that takes into consideration different, but equally valid, points of view. From an IT perspective the aim of this process is to let people search in the space of possible opinions, looking for the opinion with the widest possible endorsement. To achieve this goal we want to produce a Darwinian system of knowledge. In a Darwinian system the evolution is generated by subsequent evaluation of existing partial solutions, and productions of different mutations based on them. Such mutations are effectively searching the space of possible solutions to a problem.

1.1.2.2 Security

The objective is to realize a layered architecture, based on a safe, reliable and trustworthy toolbox, that will be designed to assure an