Utilize este identificador para referenciar este registo: http://hdl.handle.net/10362/30072
Título: Model-driven Personalisation of Human-Computer Interaction across Ubiquitous Computing Applications
Autor: Madeira, Rui Miguel Neves Gonçalves
Orientador: Correia, Nuno
Postolache, Octavian
Palavras-chave: Personalisation
Adaptation
Model-driven engineering
User modelling
Ubiquitous computing
Mobile computing
Data de Defesa: Dez-2017
Resumo: Personalisation is essential to Ubiquitous Computing (Ubicomp), which focuses on a human-centred paradigm aiming to provide interaction with adaptive content, services, and interfaces towards each one of its users, according to the context of the applications’ scenarios. However, the provision of that appropriated personalised interaction is a true challenge due to different reasons, such as the user interests, heterogeneous environments and devices, dynamic user behaviour and data capture. This dissertation focuses on a model-driven personalisation solution that has the main goal of facili-tating the implementation of a personalised human-computer interaction across different Ubicomp scenarios and applications. The research reported here investigates how a generic and interoperable model for personalisation can be used, shared and processed by different applications, among diverse devices, and across different scenarios, studying how it can enrich human-computer interaction. The research started by the definition of a consistent user model with the integration of context to end in a pervasive model for the definition of personalisations across different applications. Besides the model proposal, the other key contributions within the solution are the modelling frame-work, which encapsulates the model and integrates the user profiling module, and a cloud-based platform to pervasively support developers in the implementation of personalisation across different applications and scenarios. This platform provides tools to put end users in control of their data and to support developers through web services based operations implemented on top of a personalisa-tion API, which can also be used independently of the platform for testing purposes, for instance. Several Ubicomp applications prototypes were designed and used to evaluate, at different phases, both the solution as a whole and each one of its components. Some were specially created with the goal of evaluating specific research questions of this work. Others were being developed with a pur-pose other than for personalisation evaluation, but they ended up as personalised prototypes to better address their initial goals. The process of applying the personalisation model to the design of the latter should also work as a proof of concept on the developer side. On the one hand, developers have been probed with the implementation of personalised applications using the proposed solution, or a part of it, to assess how it works and can help them. The usage of our solution by developers was also important to assess how the model and the platform respond to the developers’ needs. On the other hand, some prototypes that implement our model-driven per-sonalisation solution have been selected for end user evaluation. Usually, user testing was conducted at two different stages of the development, using: (1) a non-personalised version; (2) the final per-sonalised version. This procedure allowed us to assess if personalisation improved the human-com-puter interaction. The first stage was also important to know who were the end users and gather interaction data to come up with personalisation proposals for each prototype. Globally, the results of both developers and end users tests were very positive. Finally, this dissertation proposes further work, which is already ongoing, related to the study of a methodology to the implementation and evaluation of personalised applications, supported by the development of three mobile health applications for rehabilitation.
URI: http://hdl.handle.net/10362/30072
Designação: Doutor em Informática
Aparece nas colecções:FCT: DI - Teses de Doutoramento

Ficheiros deste registo:
Ficheiro Descrição TamanhoFormato 
Madeira_2017.pdf7,27 MBAdobe PDFVer/Abrir


FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpace
Formato BibTex MendeleyEndnote Degois 

Todos os registos no repositório estão protegidos por leis de copyright, com todos os direitos reservados.