PRAIA Pedagogical Rational and Affective Inteligent Agents)

Universities Capes-Cofecub project between UFRGS and LIG-France
Duration 2007 a 2010
Role Collaborateur Researcher

PRAIA is an international project of cooperation between the researchers groups PPGC/UFRGS & PIPCA/UNISINOS in Brazil and LIG & LIMSI in France, which has received financial founds from Capes/Cofecub. We are interested in, with our research, responding the following questions:

  • How to model and to represent student emotions?
  • How to recognize student s emotions?
  • How to respond appropriately to student s emotions?
  • How to develop methods and techniques to evaluate the developed methods and techniques?

In order to respond to those questions, we defined some activities for our project, which are:

  1. Analyze and study the existing methods and techniques for recognizing, modeling and expressing emotions;
  2. Conception of new methods and techniques for recognizing, modeling and expressing emotions;
  3. Develop a prototype using the methods and techniques conceive by our researches groups.

This prototype is being developed and is a multi-user (collaborative) game, since we are mainly interested in emotions that arise between students in collaborative and competitive pedagogical activities.

Inference of Emotions by Face

University: PIPCA - UNISINOS
Duration: 2007-2008
Role: Superviser of Undergraduate work

This work aims at developping  an application that is able to infer the users’ emotions Happiness, Surprise, Fear, Anger, Sadness, Disgust, from their facial expression caught by a webcam. The emotional detection is based on FACs psychological face coding system and uses machine learning algorithms for the inference. The experiments’ results show an average success rate of 60%, arriving a 90% for the emotions happiness and sadness.

Uma Arquitetura Independente de Domínio e Plataforma para Apresentação de Comportamentos em Agentes Pedagógicos Animados.

University PIPCA - UNISINOS
Duration 2007
Role Superviser of Undergraduate work

This work aims ate developping the architecture and the implementation of a module responsible for the presentation of verbal (speech) and corporal (animation) behaviors of animated pedagogical agents. This module is independent of domain and application, being able to be inserted in any educational environment apart of its application domain, and independent of platform, making possible that it can be executed in different operational systems. In this way, it was implemented as a reactive agent in Java (what makes it independent of platform), called Body agent, that communicates with the agent’s mind using the agents communication language FIPA-ACL. This last capacity of the agent allows it to be inserted in other intelligent learning environments, if they are able to communicate in FIPA-ACL.
Availabe for download here (contém arquivo reademe,txt com guia de instalação)

Pat: An Animated Pedagogical Agent to Interact Affecivelly with the Student (PhD thesis)

University PGCC-UFRGS/LIG-France
Duration 2000-2004
Role PhD Student
Superviser Rosa Vicari

This work proposes an animated pedagogical agent, called Pat (Pedagogical and Affective Tutor), that has the role of providing emotional support to the student: motivating and encouraging him, making him believe in his self-ability, and promoting a positive mood in him, which fosters learning. This careful support of the agent, its affective tactics, is expressed through emotional behaviour and encouragement messages of the lifelike character. Due to human social tendency of anthropomorphising software, we believe that a software agent can accomplish this affective role.
In order to choose the adequate affective tactics, the agent should also know the student’s emotions. The proposed agent recognises the student’s emotions: joy/distress, satisfaction/disappointment, anger/gratitude, and shame, from the student’s observable behaviour, i. e. his actions in the interface of the educational system. The inference of emotions is psychologically grounded on the cognitive theory of emotions. More specifically, we use the OCC model which is based on the cognitive approach of emotion and can be computationally implemented.
Due to the dynamic nature of the student’s affective information, we adopted a BDI approach to implement the affective user model and the affective diagnosis. Besides, in our work we profit from the reasoning capacity of the BDI approach in order for the agent to deduce the student’s appraisal, which allows it to infer the student’s emotions.

Master Dissertation: Software Agents  for Analysis of Collaboration in a Virtual Classroom

University PUCRS
Duration 1998-2000
Role Master Student
Superviser Flavio de Oliveira

Trends in distance education show a growing emphasis in collaborative learning, stimulating students to exchange ideas and information. When the students interact among themselves, they feel more motivated and engaged and get better results in their studies. A collaborative environment, however, will demand a higher effort from the teacher, who will have to supervise all the discussions among the learners, so that they do not deviate from the intended topic for the lesson. Moreover, the information proceeding from the interactions among the students provide the teacher with insights useful for an individual evaluation of the students and the course. In this way, this work presents a Multi-agent architecture able to monitor the communication tools in a distance learning group. This system will analyze the discussions taking place in these tools (discussion list, chat and newsgroups), showing to the teacher statistical information (percentage of participation and number of messages) and identifying possible associations in the interactions - topics and subtopics that interest the students, groups of learners that interact intensively, etc.