In-memory cluster computing platforms have gained momentum in the last years, due to their ability to analyse big amounts of data in parallel. These platforms are complex and difficult-to-manage environments. In addition, there is a lack of tools to better understand and optimize such platforms that consequently form backbone of big data infrastructure and technologies. This directly leads to underutilization of available resources and application failures in such environment. One of the key aspects that can address this problem is optimization of the task parallelism of application in such environments. In this paper, we propose a machine learning based method that recommends optimal parameters for task parallelization in big data workloads. By monitoring and gathering metrics at system and application level, we are able to find statistical correlations that allow us to characterize and predict the effect of different parallelism settings on performance. These predictions are used to recommend an optimal configuration to users before launching their workloads in the cluster, avoiding possible failures, performance degradation and wastage of resources. We evaluate our method with a benchmark of 15 Spark applications on the Grid5000 testbed. We observe up to a 51% gain on performance when using the recommended parallelism settings. The model is also interpretable and can give insights to the user into how different metrics and parameters affect the performance.

In this paper we present validation tests that we have carried out on gestures that we have designed for an embodied conversational agent (ECAs), to assess their soundness with a view to applying said gestures in a forthcoming experiment to explore the possibilities ECAs can offer to overcome typical robustness problems in spoken language dialogue systems (SLDSs). The paper is divided into two parts: First we carry our a literature review to acquire a sense of the extent to which ECAs can help overcome user frustration during human-machine interaction. Then we associate tentative, yet specific, ECA gestural behaviour with each of the main dialogue stages, with special emphasis on problem situations. In the second part we describe the tests we have carried out to validate our ECA's gestural repertoire. The results obtained show that users generally understand and naturally accept the gestures, to a reasonable degree. This encourages us to proceed with the next stage of research: evaluating the gestural strategy in real dialogue situations with the aim of learning about how to favour a more efficient and pleasant dialogue flow for the users.

In this paper we explore the possibilities that conversational agent technology offers for the improvement of the quality of human-machine interaction in a concrete area of application: the multimodal biometric authentication system. Our approach looks at the user perception effects related to the system interface rather than to the performance of the biometric technology itself. For this purpose we have created a multibiometric user test environment with two different interfaces or interaction metaphors: one with an embodied conversational agent and the other with on-screen text messages only. We present the results of an exploratory experiment that reveals interesting effects, related to the presence of a conversational agent, on the user's perception of parameters such as privacy, ease of use, invasiveness or system security.