Abstract

Abstract

Abstract

Abstract
This article presents the results of a qualitative study of a multi-channel bank. It aims at developing new methods of gathering user requirements for web interfaces, joining HCI and Marketing perspectives. The results obtained so far indicate that, as most of financial operations are functionally available in the different service channels, experience requirements become increasingly important. In this context, essential use cases are particularly valuable in improving the process of gathering customer requirements. As they allow the analysis of users' interaction needs in a channel-independent way, their use can improve decisions on what services are best suited to each channel, to effectively address customer needs across different interaction modes, and make an efficient allocation of resources among channels.

Abstract
The commercial use of the Internet for service provision has deeply changed the environment where human-computer interaction takes place. Web interfaces are now integrated in overall service provision, and are designed for a huge and diversified set of potential users, in an uncontrolled context. With the aim of understanding customer interaction needs and improving the methods of requirements elicitation in the web context, a qualitative study of a multi-channel Portuguese bank was made. The results obtained so far indicate that Interaction Design and Services Marketing have strong complementarities. The Marketing perspective is especially useful with regard to the study of customer experience requirements, which are increasingly influential in customer decisions to adopt Internet services. Essential use cases are also very useful in the multiple platform service context, as they allow the elicitation of experience requirements in a technology-independent way, and therefore allow an integrated management of the different interaction channels.

Abstract
Large-scale applications that require executing very large numbers of tasks are only feasible through parallelism. In this work we present a system that automatically handles large numbers of experiments and data in the context of machine learning. Our system controls all experiments, including re-submission of failed jobs and relies on available resource managers to spawn jobs through pools of machines. Our results show that we can manage a very large number of experiments, using a reasonable amount of idle CPU cycles, with very little user intervention.