Juan Martínez-Miranda

Emotions in human and artificial intelligence

Intelligence and emotions differentiate humans from animals. Emotion is part of a persons behaviour and certain feelings can affect his/her performance, emotions can even prevent a person from producing an intelligent outcome. Therefore, when a computer aims to emulate human behaviour, not only should this computer think and reason, but it should also be able to show emotions. This paper presents a review of recent research that shows the importance of the emotions in human intelligence. This paper also presents the research that has been carried out into the incorporation of emotions to intelligent systems, how a computer can show affections and how to create intelligent agents that show emotions to other agents that communicate with them in the same environment.

Modelling Two Emotion Regulation Strategies as Key Features of Therapeutic Empathy

Computational models of affective processes have allowed the construction of synthetic characters able to produce empathic behaviours. The use of empathy, as a strategy to enhance engagement and cooperation with human pairs has proved good results in different application domains. Mental care is a particular area where the use of empathic virtual characters would offer several advantages facilitating the self-treatment management. Empathic responses in counselling and psychotherapy differ from “natural” empathy produced in everyday situations. Therapeutic empathy requires an emotional involvement of the therapist with the patient and an emotional detachment for a more objective appraisal of the situation. This paper introduces a model of emotion regulation as the first steps to get therapeutic empathy responses in a virtual agent constructed to support the treatment of major depression. The modelling of two specific strategies of emotion regulation based on Gross theory (cognitive change and response modulation) is described.

Simulation of work teams using a multi-agent system

When a complex project starts in industry the selection of team(s) is one of the first steps that must be done. The correct selection of people to integrate a team within a complex engineering project is not a trivial task. Team configuration is a type of business decision-making typically done by a manager based on his/her past experience and with the available information about the behaviour and interaction between the potential team members. In this work we propose a Multi-Agent System that provides information about the possible overall behaviour of a work team and present the first results obtained.

TEAKS: simulation of human performance at work to support team configuration

The management of a complex engineering project is a difficult task that initially involves the division of the project into tasks; the selection of the right people; and the correct allocation of those tasks for the selected people. Team configuration process is typically performed by a manager based on his/her past experience and the available (though frequently scarce, uncertain and dynamic) information about the cognitive, personal and social characteristics of the potential team members. To support this decision-making process we propose TEAKS, a knowledge-based tool that given an initial team configuration and a set of tasks, simulates the most possible team performance. Its formal bases are Fuzzy Set Theory and Fuzzy Logic and it is implemented using Multi-Agent Systems technology in Java, JADE, JESS and FuzzyJESS. This tool was validated with an industrial project involving 23 team members and 23 tasks (ranging from task assignment to SAP administration, SQL programming and equipment connectivity).

Look on the Bright Side: A Model of Cognitive Change in Virtual Agents

The modelling of empathic reactions in virtual agents has been recognised as a key aspect to improve agent-user relationship. Empathy is particularly important in a virtual agent used in computer-based psychotherapy applications. From a clinical perspective, it is more useful to produce therapeutic-empathy responses in the agent and not only natural empathic reactions as response to patient’s inputs. Based on Gross’s theory of emotion regulation, this paper presents a model of cognitive change used to reappraise with a more positive perspective those negative situations reported by the user and provide an adequate emotional reaction during the interaction. An initial evaluation of the model is also presented and the further work is described.

Age-based differences in preferences and affective reactions towards a robot's personality during interaction

Abstract The design and use of social robots addressed to the children population is a growing research field. The understanding of how the children react, or what are their preferences towards a robot with different styles of interaction is an important aspect to maximise the construction of a social bond between the robot and the child. In this paper we describe an experiment to assess these affective reactions and preferences based on an age-based stratification. A Wizard-Of-Oz scenario was used to design a playing mission where the child guides the robot, via voice commands, through a maze while collecting sweets and avoiding obstacles. Every participant interacted with two robots during the session. Two types of personalities were modelled in the robots with the same physical appearance: the agreeableness, and its opposite, disagreeableness personality trait. A total of 174 children between 6 and 11 years old took part in the experiment. The data about the affective reactions and preferences of the children towards the robots were collected through a multiple choice questionnaire. Significative statistical differences based on the age were found in the self-reported information provided by the children. These differences were noticeable in the youngest children (aged 6–7) with respect to the rest of the participants.

Modelling Personality-Based Individual Differences in the Use of Emotion Regulation Strategies

The modelling of the emotion regulation process is an important aspect that can contribute to the creation of more realistic intelligent virtual agents. The emotional reactions in a virtual agent, produced by the regulation process, can be useful to better adapt the agent’s behaviour to the particular requirements of a social interactive scenario. We propose a computational model of emotion regulation where the use of different strategies to down-regulate the negative emotions is based on personality-based individual differences. Our model implements a fuzzy mechanism that reproduce the correlation between different personality traits and the use of the specific emotion regulation strategies described in the literature. The validation of the model has been performed through a set of simulations where synthetic data have been generated to represent individuals with different personalities.

Using acoustic paralinguistic information to assess the interaction quality in speech-based systems for elderly users

Abstract At present, there is a growing need for assisting living systems for the elderly. This type of technology must meet certain accessibility and usability requirements to be adopted successfully. Speech-based interfaces represent a very attractive option for improving the usability and acceptability of assistive technologies for elderly users because they provide a very easily accessible interface and also facilitate paralinguistic analysis of the users voice to provide valuable information about their emotional and mental states while interacting with a computer system. This information can be used to improve the quality of the interaction between the assistive system and its user. In this study, we propose a method based on the automatic recognition of acoustic paralinguistic phenomena for estimating the quality of speech-based interactions. The proposed method automatically recognizes paralinguistic phenomena (e.g., shouting, hyper-articulation, and hesitation) during the interaction between the user and the system. The interaction is then characterized based on the occurrence of these phenomena. A model is trained using this characterization, which can estimate the quality of the interaction. Using this method, we obtained good recognition performance (F-measure around 70%) when classifying paralinguistic phenomena and there was also a strong correlation between the estimated quality of the interaction metrics and the quality of the interaction metrics reported by the users of an assistive system. The proposed method is useful for automatically estimating the quality of interactions perceived by users and it could support the construction of speech-based systems to adapt and personalize content as well as the style of the interaction with the user.

Recognition of Paralinguistic Information in Spoken Dialogue Systems for Elderly People

Different strategies are currently studied and applied with the objective to facilitate the acceptability and effective use of Ambient Assisted Living (AAL) applications. One of these strategies is the development of speech-based interfaces to facilitate the communication between the system and the user. In addition to the improvement of communication, the voice of the elder can be also used to automatically classify some paralinguistic phenomena associated with specific mental states and assess the quality of the interaction between the system and the target user. This paper presents our initial work in the construction of these classifiers using an existent spoken dialogue corpus. We present the performance obtained in our models using spoken dialogues from young and older users. We also discuss the further work to effectively integrate these models into AAL applications.