Model-driven generative programming for BIS mobile applications

• Autor(a) principal: Silva, Luís Miguel Pires Teixeira da
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10071/8442
• topic: Model-driven development; Source code generation; Unified Modeling Language (UML); Design patterns; Software prototyping; Object oriented software development; Desenvolvimento orientado por modelos; Geração de código fonte; Linguagem de Modelação Unificada (UML); Padrões de desenho; Prototipagem de software; Desenvolvimento de software orientado a objetos

The burst on the availability of smart phones based on the Android platform calls for cost-effective techniques to generate mobile apps for general purpose, distributed business information systems (BIS). To mitigate this problem our research aims at applying model-driven techniques to automatically generate usable prototypes with a sound, maintainable, architecture. Following three base principles: model-based generation, separation of concerns, paradigm seamlessness, we try to answer the main guiding question – how to reduce development time and cost by transforming a given domain model into an Android application? To answer this question we propose to develop an application that follows a generative approach for mobile BIS apps that will mitigate the identified problems. Its input is a platform independent model (PIM), with business rules specified in OCL (Object Constraint Language). We adopted the Design Science Research methodology, that helps gaining problem understanding, identifying systemically appropriate solutions, and in effectively evaluating new and innovative solutions. To better evaluate our solution, besides resorting to third party tools to test specific components integration, we demonstrated its usage and evaluated how well it mitigates a subset of the identified problems in an observational study (we presented our generated apps to an outside audience in a controlled environment to study our model-based centered and, general apps understandability) and communicated its effectiveness to researchers and practitioners.