Uma abordagem para a identificação automática de problemas de usabilidade em interfaces de sistemas web através de reconhecimento de padrões

• Ano de defesa: 2013
• Autor(a) principal: Santana, Gisele Alves
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Cornelio Procopio; Programa de Pós-Graduação em Engenharia Elétrica
• Palavras-chave em Português: Sites da Web - Avaliação e classificação; Sistemas de reconhecimento de padrões; Web sites - Ratings and rankings; Pattern recognition systems
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/783

Recently, some systems have been transferred to the web-based platform. Many services and applications, including those of power systems simulating and planning and automation systems, are developed with Internet-based interface. Usability is mainly a characteristic of the interface and is associated with the functionalities of the systems. It describes how well a product can be used for its intended purpose by its users with efficiency, effectiveness and satisfaction. This paper presents the application of pattern recognition techniques in automatic detection and classification of usability problems in the interface of a web system. The initial focus of this work is focused on identifying potential usability problems in web forms. The potential usability problems of the web form are defined based on the recommendations described in the literature. The tasks performed by the user are obtained through analysis of user interaction stored in log files. The classification of tasks which are performed as expected and what are considered potential usability problems is performed by an Artificial Neural Network.