Perceiving is Believing. Authentication with Behavioural and Cognitive Factors

• Autor(a) principal: Malhadas, Luis Filipe Machado
• Data de Publicação: 2016
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/19960
• topic: Authentication; Security; Multi-Modal; Multi-Factor; Multi-Mode; Biometrics; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática

Most computer users have experienced login problems such as, forgetting passwords, loosing token cards and authentication dongles, failing that complicated screen pattern once again, as well as, interaction difficulties in usability. Facing the difficulties of non-flexible strong authentication solutions, users tend to react with poor acceptance or to relax the assumed correct use of authentication procedures and devices, rendering the intended security useless. Biometrics can, sort of, solve some of those problems. However, despite the vast research, there is no perfect solution into designing a secure strong authentication procedure, falling into a trade off between intrusiveness, effectiveness, contextual adequacy and security guarantees. Taking advantage of new technology, recent research onmulti-modal, behavioural and cognitive oriented authentication proposals have sought to optimize trade off towards precision and convenience, reducing intrusiveness for the same amount of security. But these solutions also fall short with respect to different scenarios. Users perform currently multiple authentications everyday, through multiple devices, in panoply of different situations, involving different resources and diverse usage contexts, with no "better authentication solution" for all possible purposes. The proposed framework enhances the recent research in user authentication services with a broader view on the problems involving each solution, towards an usable secure authentication methodology combining and exploring the strengths of each method. It will than be used to prototype instances of new dynamic multifactor models (including novel models of behavioural and cognitive biometrics), materializing the PiB (perceiving is believing) authentication. Ultimately we show how the proposed framework can be smoothly integrated in applications and other authentication services and protocols, namely in the context of SSO Authentication Services and OAuth.