Preface

Hamid Aghajan, Juan Carlos Augusto
2015 Journal of Ambient Intelligence and Smart Environments  
whom we thank for their work. As in previous volumes, in 2015 JAISE will keep the tradition of alternating Thematic Issues with Regular Issues. The list of issues for the coming year is included at the end of this article. This issue The vision of the future of Intelligent Environments includes fixed and mobile display screens with socially-aware virtual embodied agents. The paper "Mirror mirror on the wall" by Mattheij et al. examines the potential of virtual embodied agents to generate a
more » ... l connection with the inhabitants of Intelligent Environments. The hypothesis that users mimic the behavior of an interactive agent is explored in the paper with an experiment in which participants interact verbally with such an agent. Two modalities of vocal pitch and affective facial expressions of the agent are used to measure the similarity of expressions of the participants and the agents. Domotic Effects is an ontology-based modeling approach which defines a high-level abstract layer for defining user goals in a smart environment in a declarative way. The paper "Real-Time Monitoring of High-Level States in Smart Environments" by Corno and Razzak describes an approach for the automatic evaluation of domotic effects combined through Boolean expressions for monitoring of the domotic structure of an environment. The paper addresses the problem of finding the new values of all the domotic effects de-fined for the environment when devices change their state or new sensors are detected in the environment. The health condition of the aging population can be generally described through scenarios, such as those described in the World Health Organization's International Classification of Functioning, Disability and Health (ICF). However, these descriptions have not been implemented in the development of smart homes. The paper "Public health resources for smart-home scenario development: A methodological approach" by Brink et al. proposes a methodology for the generation of health condition scenarios based on using public health resources. Specific user characteristics are taken from the descriptive documents while common daily user activities are obtained by research into common activities practiced by the endusers. As a result, a functional scenario is generated for the target product by combining user characteristics and user activities. Activity recognition classifiers, which label an activity based on sensor data, have typically a decreased level of classification accuracy when trained on a population and then used in the real world with a particular person. The paper "Adapting Activity Recognition to a Person with Multi-Classifier Adaptive Training" by Cvetkovic et al. proposes a multi-classifier adaptive training algorithm, which aims to adapt activity recognition classifier to a particular person by using four classifiers when new user data is acquired. A general classifier is trained on the labelled data available before deployment and used in the new environment. A user-specific classifier is trained on a limited amount of labelled data belonging to the user in the new environment. A domain-independent meta-classifier decides whether to classify a new instance with the general or specific classifier. And a second meta-classifier decides whether to include the new instance into the 1876-1364/15/$35.00
doi:10.3233/ais-150313 fatcat:gpeu5s4t3vfytmp36haz4pdfiy