Search

Twenty-eight Benelux Conference on Artificial Intelligence, Item does not contain fulltext, Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to provide natural interaction with humans. This work proposes to extend conversational agents' benefits for social skills training by analysing the emotion conveyed by the user's speech. For that, we developed a new system that captures emotions from the human voice and, combined with the context of a particular situation, uses this to influence the internal state of the agent and change its behaviour. An example of the system’s use is shown and its limitations and advantages are discussed, together with the internal workflow of the system.

ECAI 2016: European Conference on Artificial Intelligence, The Hague, the Netherlands, from August 29 to September 2, 2016, Item does not contain fulltext, Artificial Intelligence techniques are increasingly being used to develop smart training applications for professionals in various domains. This paper presents an intelligent training system that enables professionals in the public domain to practice their aggression de-escalation skills. The system is one of the main products of the STRESS project, an interdisciplinary research project involving partners from academia, industry and society. The system makes use of a variety of AI-related techniques, including simulation, virtual agents, sensor fusion, model-based analysis and adaptive support. A preliminary evaluation of the system has been conducted with two groups of potential end users, namely tram conductors and police academy students.

BNAIC 2016 (Amsterdam, November 10-11, 2016), Contains fulltext : 161897.pdf (publisher's version ) (Open Access), Research on human interaction has shown that attributing agency to another agent has substantial consequences for the way we perceive and evaluate its actions. Specifically, considering an agent's actions related to either effort or ability can have important consequences for the attribution of responsibility. This study indicates that participants' interpretation of a robot failure in terms of effort - as opposed to ability - significantly increases their attribution of agency and - to some extent - moral responsibility to the robot. However a robot displaying lack of effort does not lead to the level of affective and behavioural reactions of participants normally found in reactions to other human agents.

Contains fulltext : 171122.pdf (Publisher’s version ) (Open Access), In this review, we provide an overview of the clinical aspects, histopathology, molecular genetics, and treatment options for Vulvar Paget's Disease (VPD), a rare skin disease, most commonly found in postmenopausal Caucasian women. The underlying cause of VPD remains not well understood. VPD is rarely associated with an underlying urogenital, gastrointestinal or vulvar carcinoma. In approximately 25% of the cases, VPD is invasive; in these cases, the prognosis is worse than in non-invasive cases. Recurrence rates in invasive VPD are high: 33% in cases with clear margins, and even higher when surgical margins are not clear, regardless of invasion. Historically, surgical excision has been the treatment of choice. Recent studies show that imiquimod cream may be an effective and safe alternative.

8th International Conference on Social Informatics, SocInfo 2016, Bellevue, WA, USA, November 11-14, 2016, Item does not contain fulltext, Social media are an effective means for people to share everyday problems with their peers. Although this often leads to empathic responses which help alleviate the experienced stress, such peer support is not always available. As an alternative solution for such situations, this paper explores the possibilities to develop an 'artificial friend' that offers online social support through text messages. To formulate the requirements for such a system, a pilot study was performed in which 230 participants were asked (via a crowdsourcing platform) to report their experiences regarding online peer support. The results have been converted into a number of working hypotheses about online peer support. Based on these hypotheses, a conceptual framework has been developed that describes the functioning of the proposed support system.

The 28th Benelux Conference on Artificial Intelligence, 11 november 2016, Contains fulltext : 162394.pdf (publisher's version ) (Open Access), We propose a system that assigns topical labels to automatically detected events in the Twitter stream. The automatic detection and labeling of events in social media streams is a 'big data' problem. The early detection of future social events, specifically those associated with civil unrest, has a wide applicability in areas such as security, e-governance, and journalism. We used machine learning algorithms and encoded the social media data using a wide range of features. Experiments show a high-precision (but low-recall) performance in the first step. We designed a second step that exploits classification probabilities, boosting the recall of our category of interest, social action events.

Item does not contain fulltext

Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to provide natural interaction with humans. This work proposes to extend conversational agents’ benefits for social skills training by analysing the emotion conveyed by the user’s speech. For that, we developed a new system that captures emotions from the human voice and, combined with the context of a particular situation, uses this to influence the internal state of the agent and change its behaviour. An example of the system’s use is shown and its limitations and advantages are discussed, together with the internal workflow of the system.

NLDA Research Paper no. 110, Cultura is een virtuele trainingsmethode op het grensvlak van cultuur en ethiek. De effectiviteit van dit onderwijsleermiddel is onderzocht door de Nederlandse Defensie Academie in samenwerking met de Vrije Universiteit Amsterdam. Het Cultura onderzoek beschrijft naast het ontwikkelingsproces van Cultura ook het verloop van het onderzoek en de resultaten. Het nut van didactische herhaling komt sterk naar voren; de stof op verschillende manieren herhalen zet aan tot nadenken over de cultuurverschillen welke men gedurende missies aantreft. Cultura kan worden ingezet om een hoger niveau van training te bereiken dan wordt bereikt met enkel een informerend college. Nog meer culturele diepgang wordt echter bereikt wanneer nog een extra leermiddel wordt ingezet; bijvoorbeeld een rollenspel met acteur. Hiernaast bevat het Cultura onderzoek een advies om leermiddelen als Cultura in te zetten binnen het huidige culturele onderwijs.

12th International Conference on Engineering Psychology and Cognitive Ergonomics, EPCE 2015, Held as Part of HCI International 2015, Los Angeles, CA, USA, August 2-7, 2015, Item does not contain fulltext, Being able to de-escalate aggressive behavior during face-to-face interactions is an important skill for employees in a variety of domains. To do this effectively, employees should learn to recognize the emotional state of their conversation partner. However, this task can be seriously hindered by the stress triggered by the aggressive encounter. To gain more insight in the impact of threat on task performance during emotion recognition, 30 participants were asked to perform an emotion recognition task using pictures of virtual characters. Each participant performed the task two times, once under normal circumstances, and once in a 'stress' condition in which threatening stimuli were presented whenever a wrong answer was given. Additionally, all 30 participants performed a second, mathematical task, also under a normal and a 'stress' condition. Counterbalancing was used to control for order effects. The results indicate that there was a negative impact of the (simulated) threat on performance in the emotion recognition task, but not in the mathematical task. In follow-up research, these results will be used to develop an adaptive serious game for public service workers, with which they can train their aggression de-escalation skills in a personalized manner.

Item does not contain fulltext

2015 International Conference on Autonomous Agents and Multiagent Systems, Istanbul, Turkey - May 04 - 08, 2015, Item does not contain fulltext, Most research into intelligent virtual agents focuses on agents with a positive stance towards the user. Nevertheless, the development of virtual agents that show aggressive behavior may also be interesting for a range of application domains, varying from aggression de-escalation training to anti-bullying education. However, ensuring that such aggressive agents achieve the desired effect is not easy, as they need to be believable in a number of aspects. In particular, they need to bring their human conversation partners into a serious state of anxiety. To investigate to what extent this can be achieved using state-of-the-art virtual agent technology, an experiment was performed in which the impact of an aggressive virtual agent was compared with that of an aggressive human. By randomly distributing a group of 28 participants over two conditions (virtual and human) and measuring their physiological and subjective emotional state before and after an aggressive outburst of their conversation partner, the difference between virtual and human aggression was studied. The results point out that both types of aggression induced a substantial stress response, but that the impact of the human aggression was higher than that of the virtual aggression.

16th Annual European GAMEON Conference on Simulation and AI in Computer Games, December 2-4, 2015, University of Amsterdam (UvA) , Amsterdam, The Netherlands, Item does not contain fulltext, Public transport employees are confronted with aggressive behavior on a regular basis. As such encounters can have serious consequences, employees need to be well prepared, so that they know how to deal with incidents of aggression. The current paper describes an ongoing endeavor that is aimed at the development and evaluation of a simulation-based training environment for public transport employees, by which they can practice their verbal aggression de-escalation skills during face-to-face conversations. A prototype of the training environment is presented, as well as an experiment to evaluate the environment in several steps. The results indicate that the prototype is evaluated positively with respect to user satisfaction, whereas there is room for improvement with respect to learning effectiveness.

15th International Conference on Intelligent Virtual Agents, Delft, The Netherlands, August 26-28, 2015, Proceedings, Item does not contain fulltext, Embodied conversational agents (ECAs) have been put forward as a promising means for the training of social skills. The traditional approach to drive the behaviour of ECAs during human-agent dialogues is to use conversation trees. Although this approach is easy to use and very transparent, an important limitation of conversation trees is that the resulting behaviour of the ECAs is often perceived as predictable. To provide ECAs with more sophisticated behaviour, the current paper proposes an approach to endow them with mental states. The approach is illustrated by a motivational example in the domain of aggression de-escalation training.

18th International Conference on Principles and Practice of Multi-Agent Systems, Bertinoro, Italy, October 26-30, 2015, Proceedings, Item does not contain fulltext, Traditionally, Embodied Conversational Agents communicate with humans using dialogue systems based on conversation trees. To enhance the flexibility and variability of dialogues, this paper proposes an approach to integrate conversation trees with cognitive models. The approach is illustrated by a case study in the domain of aggression de-escalation training, and a preliminary evaluation in the context of a practical application is presented.

Item does not contain fulltext

First International Symposium, ALIA 2014, Bangor, UK, November 5-6, 2014. Revised Selected Papers, Item does not contain fulltext, The ability to de-escalate confrontations with aggressive individuals is a useful skill, in particular within professions in public domains. Nevertheless, offering appropriate training that enables students to develop such skills is a nontrivial matter. As a complementary approach to real-world training, the STRESS project proposes a simulation-based environment for training of aggression de-escalation. The main focus of the current paper is to make this system adaptive to the performance of the trainee. To realize this, first a number of learning goals have been identified. Based on these, several levels of difficulty were established, as well as a mechanism to switch up and down between these levels based on the user’s score. A preliminary evaluation demonstrated that the system successfully adapts its difficulty level to the performance of the user, and that users are generally positive about the adaptation mechanism.

18th International Conference on Principles and Practice of Multi-Agent Systems, Bertinoro, Italy, October 26-30, 2015, Item does not contain fulltext, Human space flight is a prototypical example of a complex, dynamic, and safety-critical domain in which missions are performed by collaborative teams of humans and technical systems. In such domains, intelligent electronic partners (ePartners) can play a useful role in supporting human-robot teams in their problem solving process whenever a non-nominal situation is encountered. To enhance the supportive capabilities of such ePartners, this paper presents an approach to formally represent the functionality of human-robot teams in terms of different levels of abstraction. By establishing formal relations between domain knowledge at different abstraction levels and introducing reasoning rules to navigate through these relations, ePartners are endowed with a number of supportive functions, such as the ability to reason about the mission status, make suggestions in non-nominal situations, and provide explanations. The approach is applied to a use case in the context of a manned space mission to Mars. It has been implemented within a mobile application to assist robot-astronaut teams during space missions, and has been tested in a pilot experiment at the European Space Research and Technology Centre.

Item does not contain fulltext, To avoid the development of negative emotion in their teams, team leaders may benefit from being aware of the emotional dynamics of the team members. To this end, the use of intelligent computer systems that analyze emotional processes within teams is a promising direction. As a first step toward the development of such systems, this paper uses an agent-based approach to formalize and simulate emotion contagion processes within groups, which may involve absorption or amplification of emotions of others. The obtained computational model is analyzed both by explorative simulation and by mathematical analysis. In addition, to illustrate the applicability of the model, it is shown how the model can be integrated within a computational 'ambient agent model' that monitors and predicts group emotion levels over time and proposes group support actions based on that. Based on this description, a discussion is provided of the main contribution of the model, as well as the next steps needed to incorporate it into real-world applications.

BACKGROUND: Adjuvant therapy increases disease-free survival in endometrial cancer (EC), but has no impact on overall survival and negatively influences the quality of life. We investigated the discriminatory power of classical and immunological predictors of recurrence in a cohort of EC patients and confirmed the findings in an independent validation cohort. METHODS: We reanalysed the data from 355 EC patients and tested our findings in an independent validation cohort of 72 patients with EC. Predictors were selected and Harrell's C-index for concordance was used to determine discriminatory power for disease-free survival in the total group and stratified for histological subtype. RESULTS: Predictors for recurrence were FIGO stage, lymphovascular space invasion and numbers of cytotoxic and memory T-cells. For high risk cancer, cytotoxic or memory T-cells predicted recurrence as well as a combination of FIGO stage and lymphovascular space invasion (C-index 0.67 and 0.71 vs 0.70). Recurrence was best predicted when FIGO stage, lymphovascular space invasion and numbers of cytotoxic cells were used in combination (C-index 0.82). Findings were confirmed in the validation cohort. CONCLUSIONS: In high-risk EC, clinicopathological or immunological variables can predict regional or distant recurrence with equal accuracy, but the use of these variables in combination is more powerful.

The first joint European Society for Medical Oncology (ESMO), European SocieTy for Radiotherapy & Oncology (ESTRO) and European Society of Gynaecological Oncology (ESGO) consensus conference on endometrial cancer was held on 11–13 December 2014 in Milan, Italy, and comprised a multidisciplinary panel of 40 leading experts in the management of endometrial cancer. Before the conference, the expert panel prepared three clinically-relevant questions about endometrial cancer relating to the following four areas: Prevention and screening, surgery, adjuvant treatment and advanced and recurrent disease. All relevant scientific literature, as identified by the experts, was reviewed in advance. During the consensus conference, the panel developed recommendations for each specific question and a consensus was reached. Results of this consensus conference, together with a summary of evidence supporting each recommendation, are detailed in this article. All participants have approved this final article.