eVoto

The introduction of electronic voting (e-voting) in National Election Systems has recently received a significant interest in several countries (e.g. USA, Canada, UK, Portugal). The studies conducted until today clearly demonstrate that this type of system is perceived as having a high level of risk. As researchers in the complex systems, organization and software engineering fields, we are trying to analyze the problems associated to general e-voting systems, relating them with the system architecture and components, and finding new ways to deploy systems with reduced risk levels. This problem is a challenging one, since it will be necessary to identify the multiple dimensions that characterize e-voting systems and associated risks. These risks do not roll up from one specific research topic, but rather have a strong multidisciplinary nature. This project does not aim at tackling one or two well-known issues, e.g. security, but leaving the whole problem unsolved. On the contrary, our perspective is that a whole new integrated architecture must be developed, one that is capable to tackle the intrinsic complex nature of e-voting systems. This new approach must be multidisciplinary, since problems with e-voting lay in different research areas, such as distributed systems, information security, HCI, software design or organizationware. This new approach must also be visionary, since most of the problems with e-voting are large-scale in several dimensions (users, components, requirements, risks). Our proposed solution will consist of an architectural framework with high-level components conceived for handling risk in e-voting systems. This project will integrate teams from the following research areas: - Complex systems architecture specification (ADETTI, STAPE, ITIJ) – With specialists in National Election Systems, organizationware, complex systems analysis and modeling to define the main lines of an e-voting system that fill legal, social and technical requirements - Software systems design (FCUL, ITIJ) – With specialists in the development of complex socio-technical systems, new analysis and design techniques such as participatory design, low-fidelity prototyping and application of use- and misuse-cases - Human-computer interaction (FCUL, STAPE) – With experts in HCI, usability engineering, accessibility and multimodal interactions - Information security (IEETA, ADETTI, INESC) – With experts in designing and building information security systems and strong authentication mechanisms - Distributed systems (INESC, IEETA) – With specialists in large-scale systems, Internet technology and secure voting protocols. The project will be organized according to three major areas of concern: 1) security-related risks; 2) community-related risks; and 3) usefulness-related risks.

A introdução de votação electrónica (e-voto) nos Sistemas Eleitorais Nacionais tem recentemente recebido um interesse significativo em diversos países (e.g. EUA, UK, Portugal). Os estudos realizados até hoje claramente demonstram que este tipo de sistema é percebido como tendo um elevado nível de risco. Como investigadores em sistemas complexos, engenharia organizacional e engenharia de software, nós estamos a tentar analisar os problemas associados aos sistemas de e-voto, relacionando-os com a arquitecturas e componentes do sistema, e procurando novas formas de desenvolver sistemas com risco reduzido. Estre problema é um desafio, dado que será necessário identificar as multiplas dimensões que caracterizam os sistemas de e-voto e riscos associados. Estes riscos não surgem de uma única área de investigação, mas antes têm uma natureza fortemente multidisciplinar. Este projecto não se propôe resolver 1 ou 2 temas bem conhecidos, e.g. segurança, deixando os restantes por resolver. Ao invés, a nossa perspectiva é que uma nova arquitectura tem de ser desenvolvida, uma que seja capaz de resolver a natureza intrinsecamente complexa dos sistemas de e-voto. Esta nova abordagem tem de ser multidisciplinar, dado que os problemas se distribuem por diversas áreas, como de sistemas distribuídos, segurança da informação, IHC, design de software ou organizationalware. Esta nova abordagem tem também de ser visionária, dado que a maioria dos problemas dos sistemas de e-voto são de larga-escala em diversas dimensões (utilizadores, componentes, requisitos, riscos). A nossa proposta consiste numa arquitectura de componentes de alto nível concebida para lidar com o risco em sistemas de e-voto.