6. Marco Teórico
6.3. Centrales Hidroeléctricas
example, the teacher could tell a joke in the classroom in order to change the current emotional states of the students, if they are bored or frustrated. Because, the student is bored, anxious, angry or depressed, she could not learn or think efficiently.
7.9
Emotional Ontology
To address this requirement, we are developing an Emotion Ontology (EMO). Requirements in computing and artificial intelligence have led to the development of ontology-like resources for emotions. Affective computing aims to integrate emotional responses into computer interfaces in order to produce more realistic systems which are able to respond to the emotional communication of their users. To facilitate affective computing, Lopez et al. propose a slim ontology schema for describing emotions in human-computer interfaces [189]. Also motivated by affective computing requirements, the W3C emotion markup language (EML, http://www.w3.org/TR/emotionml/) is an XML-based standard for markup of emotions in text or databases. The proposal presented in this work is based on a generic ontology for describing emotions and their detection and expression systems taking contextual and multimodal elements into account. The ontology is proposed as a way to develop a formal model that can be easily computerized. Moreover, it is based on a standard, the Web Ontology Language (OWL), which also makes ontologies easily shareable and extensible. Once formalized as an ontology, the knowledge about emotions is used in order to make computers more accessible, personalized and adapted to user needs.
Our model is based on the appraisal and coping theory. The produced ontology serves as a basis for the representation of emotions and it is stored in the student affective model. First, each affected event, has encent which could change the person emotion. Four types of effects are considered as a primary appraisal ( harmful, threatening, challenging, and benign). Second, according to the personality type of the person and the type of effect of the incoming event the
Emotional Appraisal T rigge r
Person
Personality type
Encent
Appraisal
Event
Encent
Emotion
Figure 7.9Emotional ontology
person appraises the event and trigger his/her emotion. Finally, the triggered emotion affects on a person’s behavior. However, the person behavior is affected by the person personality. Thus, the persons cope differently with the the incoming events according to a person’s personality type. The ovals represented classes in the ontology, while the arcs represent the relationships between these classes. In our system we consider this ontology to register an emotional state history for each student, which is used during the interaction with the student.
Chapter 8
Affective Planning
Personalization and adaptation in the educational process are important factors for providing an effective educational service on the Internet. We emphasize the fact that students perceive and process information in a very different way. The primary principle of individualized learning is that no single instructional strategy is best for all students. Therefore, it is necessary to define the model taking the whole context that surrounds the user into account. This chapter describes
PANDA.TUTOR, a model of Affective Intelligent Tutoring system that illustrates how a planning system supports a better treatment of emotional reasoning. In this model we introduce two types of planning modeling; dialog based planning and lesson generation.
8.1
PANDA.TUTOR
The process of arranging personalized adaptations is usually complex because people with differ- ent personalities cope with the problem by different ways. The current platforms usually do not provide more than a relatively simple way of personalization and adaptation. Also, many adaptive hypermedia systems have been created without considering personality types and different learning objects. For instance, [104, 190] have developed an adaptive system to the student that adjusts the
difficulty of the delivered exercises. Furthermore, most of the previous approaches are based on a model of generic student’s emotional reactions. Also, coping and regulation have been ignored in the course generation field.
It is a fact when the system appears to care about the student’s state, this may make the learning process more fun and help the student enjoy their learning process. In order to respond to the student, the system should consider the student’s emotion and personality correctly. Thus, if the student has a specific negative emotion and if we not consider the student’s state, this can have side effects on his/her learning state. We should try to change the negative emotion and motiva- tion to a positive one, or maintain the positive one. The foundation of emotional plan has been inspired by [191]. He represents emotional plan as a plan-based appraisal based on assessing the incoming event to goal dedication. Afterward, plan-based appraisal is ameliorated by consider- ing appraisal and coping theory [192]. Different systems consider the emotional planning during human-computer -interaction and conversation [177], [193] , and some consideration for appraisal and coping theory during the learning process [4, 5]. In this chapter we introduce a personalized planning system called PANDA.TUTOR, in which the system deals with students in different ways, according to their personality type, emotional and motivational state of student, aiming to improve the student’s emotional intelligence as well as intellectual intelligence. Moreover, the system tries to change the student’s emotional reasoning to increase the student’s internal motivation. Thus, the system attempts to put the student in a good and motivated particular mood that could be more receptive to the material being taught. For instance, the system could treat mistakes as less im- portant if the student is going through a particular bad emotion. As consequence, students will be able to achieve learning goals more efficiently when pedagogical procedures are adapted to their individual differences.
Indeed, the system is able to predict the emotional reaction according to a student’s appraisal for each specific personality types in various situations. That could help us to predict the student be-
8.1 PANDA.TUTOR 173
havior, which help us to correct the weakness in each personality type and strong the positive side. For example, suppose we have two personality type, the first one being a complicated student (ner- vous, conscientiousness and extroversion) and the second one being an impulsive student(nervous, conscientiousness, and introversion). The extroversion side in the first type could help him/her to be less depressed than the introversion side in the second one. Thus, if the student fails or gets a low score in the exam, the second type will blame his/her self, and lose his/her confidence easer than the first one, while the first type could be more nervous than the second one at the first impres- sion. Therefore, the system tries to reduce the level of nervousness in the personality character, or increase the conscientiousness or extroversion (as social factor)to increase the confidence level in student character. In contrast, the system could also try to help the high level of extroversion student to be more concentrated. That can be done by the emotional regulation. For Instance, by changing the student’s appraisal or understanding of the situation, (as ” consider the white side ”or ”see the glass as half full and think you’ll come through this difficult time”), being empathetic with the very sensitive person, or blaming the careless student and teaching him how to understand or value his/her responsibility.
The system could focus on different aspects of instruction depending on personality characteristics of the student. Thus, PANDA.TUTORhas the ability to adapt the learning environment taking into consideration the improvement of the emotional state through adaptive educational systems in the semantic web environment. We integrate different theories of instruction ,which are guided by the student model. That means our system can generate different presentation for the generated lessons from different scenarios according to the student’s personality and states.
Hybrid planning is used for modeling the the pedagogical module. The hybrid planning provides the generated plans with a hierarchical, temporal relationships and causal dependencies between tasks on both abstract and primitive levels. Hybrid planning can be used for synthesizing flexible and adapted plans. In this chapter we will explain how we used the hybrid planning approach in
modeling the dialog and course generation.
8.2
Pedagogical Module
The Pedagogical module makes decisions about the student model. The decisions are related to the selection of an appropriate topic for the student, selection of an appropriate learning method, curriculum sequencing, etc. through the course module.
It is a fact that learning environments that don’t consider motivational and emotional factors are not adequate. So, we should simulate a teacher’s teaching way and analyze the students’ emotion- al state to give a proper regulation for adjusting their negative emotions. In our system we aim
Pedagogical Module Course Module
Student Dialog Phase (Planning Domain1) Lesson generation Phase (Planning Domain 2) Interface Phase1 (Planning Problem) Lesson (Generated Plan) Interface Phase2
(Planning Problem) Student Module
(Student History)
Figure 8.1Phases of pedagogical module
to enhance the intellectual intelligence of student as well as the emotional intelligence of student. Therefore, the pedagogical module in PANDA.TUTOR has two phases; theDialog phase, and the
Lesson generation phase.
The goal of theDialog phase is to enhance the emotional intelligence of the student by teaching the student how to appraise his or her emotional state, by defining the current emotion and the reason behind this emotion. Note that, we ask the student about his/her emotion and the reason