EmoCulture: Towards an Ontology for Describing Cultural

Differences When Expressing, Handling and Regulating Emotions

Azza Labidi, Fadoua Ouamani and Narjès Bellamine Ben Saoud

Ecole Nationale des Sciences de l’Informatique, Université de Manouba, Manouba, Tunisia

Keywords: Emotion, Culture, Collaborative Learning, Ontology.

Abstract: Collaborative learning environments bring together learners from different sociocultural contexts, around a

common task. Besides, these environments are emotional places where learners frequently experience

emotions and bring emotions that concern events from outside the learning environment. Moreover, learners

express, handle and regulate their emotion differently according to the sociocultural context to which they

belong. And as it was proven by empirical research studies, emotions can have important effects on

students’ learning and achievement. Therefore, Detecting, understanding, handling and regulating the

learner emotion and understanding their cultural differences is a key issue that need to be tackled to enhance

collaborative learning. To do so, we propose the emoculture ontology, a domain ontology for representing

relevant aspects of affective phenomena and their culture differences in collaborative learning

environments. In this paper, we will discuss first the concept of emotion and its relations with learning,

collaborative learning and culture. Second, we will present a set of selected existing emotion ontologies

which will be compared in the same section according to criteria relevant to our study. Third, we will

describe the process upon which EmoCulture was built. Finally, we will discuss the quality of the proposed

ontology and how it will be used in future works to guide the building of an emotional and cultural aware

collaborative learning environment.

1 INTRODUCTION

During the last decade, emotions have been

acknowledged in humanities and social sciences (i.e.

psychology, sociology) as an important phenomenon

of human life (Kantzara, 2006). Similarly,

collaborative learning theory has acknowledged the

crucial role of emotions in social interaction. It has

been shown that emotions largely influence social

and behavioral engagement in face-to-face or

distance collaboration. The relationship between

emotions and learning was investigated by tracking

the emotions that learners experienced while

learning (Arthur and Sidney, 2012). In fact, Emotion

analysis allows extracting knowledge that will be

useful either to mediate collaboration by regulating

certain emotions and by encouraging others (Marta

et al., 2015)

However, emotions are not universal. In (Scollon

et al., 2004), the authors have shown that depending

on the culture in which they live, individuals will

experience certain emotions more or less frequently.

Moreover, as evidenced by (Kim-Prieto et al., 2004),

the very notions of positive and negative emotion

differ from one culture to another.

Therefore to consider this key concept while

developing emotionally aware applications,

researchers have proposed ontology based emotion

model. The use of Ontologies was justified by the

following argument: First, emotion is a semantically

rich concept that holds influence-links with concepts

like culture, cognition, motivation, and personality;

these influences need to be considered and

modelled. Second, the use of ontology will allow

reasoning based on these links in order to build

emotionally intelligent systems. Third, ontology

provides sharing and reuse of domain knowledge. In

fact, in ontology engineering, it is recommended to

reuse the existing ontologies as a whole or a part of a

new ontology depending on project needs (Hoekstra,

2010) to save, time and improve the quality and

maintainability of the new ontology (Fonou et al.,

2013). Fourth, the use of ontology modelling allows

refining the modelling without affecting the system

and its logic. Fifth, ontology allows then the

building of an affective knowledge base which

handles the knowledge and links between it, Finally,

298

Labidi, A., Ouamani, F. and Ben Saoud, N.

EmoCulture: Towards an Ontology for Describing Cultural Differences When Expressing, Handling and Regulating Emotions.

DOI: 10.5220/0008168002980304

In Proceedings of the 11th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2019), pages 298-304

ISBN: 978-989-758-382-7

ontology modelling ensure the interoperability

between systems (facilitate knowledge sharing) and

the usability of this knowledge to build other

systems (Fonou et al., 2013).

In this paper, we will first review (See Section 2)

the key emotion theories and models from learning

and education perspectives and we will study the

relation between emotion, culture and learning. In

Section 3, we will present and compare existing

emotion ontologies. Based on the comparison

findings, section 4 will be devoted for the

description of the proposed ontology and its building

process. Finally, the section 5 will be dedicated to

the discussion of the main contributions of the paper,

ongoing and future works.

2 EMOCULTURE FOUNDATION

2.1 What Is Emotion?

Despite its proven vital importance, emotion is a

difficult concept to define and model. There are two

basic approaches that have defined the concept of

emotion: cognitive approaches and physical

approaches. According to cognitive approaches,

emotions are important because they relate outer

events and other people to inner concerns. A

principle of this approach is that an emotion is a

judgment of value. It is an evaluation, an ‘appraisal’

(Scollon and al., 2004). The evaluation means here

to figure out the significances of everyday events

and of people with whom an individual interacts

(Scollon and al., 2004). However, the physical

approaches consider emotion as a physiological

reaction that follows from an event. For example as

a physiological reaction, the blood pressure should

go up when someone is angry or the heart rate

should rise when he is scared (Schachter and Singer,

1962).

These two approaches are complementary

definitions underlying the fact that emotion is not

only an "answer" felt as a result of an internal

perception of an event, but it is also manifested

corporally. These definitions also highlight the

strong relationship between emotion and cognition.

Emotion has a substantial influence on the

cognitive processes in humans, including perception,

attention, learning, memory, reasoning, and problem

solving. For this reason we will focus after defining

the term emotion on the relation between emotions

and learning (Boekaerts, 2010).

2.2 Emotion in Learning

The emotional awareness studies are nowadays at

the center of concern of researchers in various fields.

For example, Researchers in medicine try to identify

depression or stress to make clinical reasoning.

Emotional awareness is also one of the concern of

Customer Relationship Management as Emotion can

translate the customer satisfaction level with the

product or service. Emotions are considered when

developing serious games by the implementation of

specific affective and motivational features that can

enhance learning outcomes by exploiting the

interdependence between emotions and participatory

appropriation. Regulating learner emotions in

education is a learning success factor as emotion

impacts motivation (Boekaerts, 2010), creativity

(Boekaerts, 2010) and problem solving behavior

(Boekaerts, 2010).

(Scollon et al., 2004) have confirmed that

emotion depends on the culture in which they live.

For this reason, in the following section, we will

focus on the relation between emotion and culture.

2.3 Emotion and Cultural Differences

Emotions are not universal (Scollon and al., 2004).

For example, it has been shown that American rate

the same expressions of happiness, sadness and

surprise more intensely compared to the Japanese.

American participants, for instance, gave higher

ratings to the external appearance of emotions while

Japanese participants, assigned higher ratings to

internal experiences of emotions.

Research on the relationship between culture (A

selection of definitions and their meaning discussion

can be found (Ouamani et al., 2012)) and emotion

dates back to 1872 when Darwin (Darwin, 1998)

argued that emotions and the expression of emotions

are universal. Since that time, the universality of the

six basic emotions (Ekman, 1992) (i.e., happiness,

sadness, anger, fear, disgust, and surprise) has

ignited a discussion amongst psychologists,

anthropologists, and sociologists. While emotions

themselves are universal phenomena, they are

always influenced by culture (Batja, 2003).

3 EXISTING EMOTION

ONTOLOGIES

Basically, ontologies deal with knowledge

representation and can be defined as formal explicit

EmoCulture: Towards an Ontology for Describing Cultural Differences When Expressing, Handling and Regulating Emotions

299

descriptions of concepts in a domain of discourse

(named classes or concepts), properties of each

concept describing various features and attributes of

the concept (roles or properties), and restrictions on

property (role restrictions). Ontology together with a

set of individual instances of classes constitutes a

knowledge base.

As we focus on sentiment analysis, emotions

have been used throughout human existence to

enhance the expressiveness of language. Affective

computing is involved with understanding the

emotion and even creating emotion. However, this is

still a difficult task, because emotion is a mental

state that is difficult to describe and human emotion

changes easily and quickly due to the effects of

complex external stimuli. In order to have a deeper

understanding of emotion, especially in text,

emotion ontology is needed. Emotion ontology will

help in recognizing, classifying, and understanding

emotion (Marco et al., 2009).

Different ontologies have been proposed in

literature with the aim of modelling emotion and

affect related issues. These ontologies will be

discussed in the following sub-sections.

3.1 Semantic Lexicon (Mathieu, 2005)

Mathieu (2005) presented a semantic lexicon about

feelings and emotions composed of words labelled

as positive emotion, negative emotion or neutral.

The lexicon was represented by ontology.

It is an ontology that helps to give students

appropriate feedback in e-learning sessions (Marta el

al., 2015). The ontology is divided into two main

classes: Emotion Awareness and Affective

Feedback. The emotion awareness class allows the

analysis of, learner emotion while the affective

feedback class allows the teacher to provide the

learner with the appropriate feedback according to

his/her emotion. The emotion awareness class

includes the different types of emotions, moods

(bored, concentrated, motivated, and unsafe) and

learner behaviors in e-learning environment. The

emotion is detected during collaborative virtual

learning processes, including textual conversations,

debates and wikis.

3.2 An Ontology of Emotions and

Feeling (Yvette et al., 2005)

Ontology of Emotions and Feelings was proposed by

(Yvette et al., 2005) and it automatically annotates

emotion in texts and determines their intensity. This

ontology classifies 950 words (600 are verbs and

350 are nouns in French) into 38 semantic classes

according to their meanings. Fear, sadness, interest,

passion, astonishment are example of these classes.

It uses the discrete model and classifies emotions as

positive, negative and neutral.

3.3 An Ontology of Emotions and

Feeling in Chinese Text (Jiajun et

al., 2008)

To analyze Chinese text, Chinese emotion ontology

was created by (Jiajun et al., 2008). It was semi-

automatically created using HowNet

(http://www.keenage.com/). The ontology contains

113 emotion categories. A high-level ontology

named the Human Emotions Ontology (HEO) was

developed in order to annotate emotion in

multimedia data (Marco et al., 2009). The main class

in the ontology is Emotion which is expressed in

dimensional and categorical models. An emotion has

an intensity, appraisals and action tendencies, and it

can be expressed through face, text, voice and

gesture. Additionally, the ontology contains classes

for the multimedia content and the annotator of the

media. The Annotator class has two subclasses:

Human or Machine (automatically annotated). Since

the emotion is expressed by a person, HEO re-uses

the Friend Of A Friend (FOAF) ontology. A

subclass Observed Person of class person was

created in FOAF and connected to the Emotion class

of HEO. Moreover, some object properties were

added in FOAF that are relevant to emotion such as

age, language and education.

3.4 An Ontology about Emotion

Awareness and Affective Feedback

(Marta et al., 2015)

An emotive expression lexicon for Japanese

language was proposed by (Marta et al., 2015) to

distinguish emotion words. The ontology classes

represent emotion using «a collection of over two

thousand expressions describing emotional states

collected manually from a wide range of literature".

Emotion words were taken from websites such as

Twitter and are categorized into ten emotions: joy,

anger, sadness, fear, shame, like, disgust, exciting,

comforted and surprise. These emotions are also

classified as positive, negative or neutral emotions.

The intensity calculation is based on the number of

times an emotion word appears in a document.

KEOD 2019 - 11th International Conference on Knowledge Engineering and Ontology Development

300

Table 1: Comparative study of existing ontologies.

(Abaalkhail, 2017).

Ontology

name

Goal

Emotional

model

Reused

ontology

Language

(Marta et

al., 2015)

Represent the

emotional

aspects in

e-learning.

Discret -

WordNet,

NLP

Emotive

Detect and

analyze

emotions in text

of social

networks

Discret -

WordNet,

Dictionary.com,

Thesaurus.com,

Oxford Dictionary

(Yvette et

al., 2005)

Automatically

annotate

emotions in the

text

Discret - -

(Ptaszynski

et al., 2012)

Analyze

emotions from

text

Discret and

dimensional

Emotion

(Kunihiko

et al., 2013)

Define the words

of emotion and

their intensity

Discret and

dimensional

Japanese Emotion

Expression

Dictionary,

Emotion ML

(Jiajun et

al., 2008)

Analyze

emotions in the

text

Discret - Hownet

Onto

Emotion

Detect emotions

from English

and Spanish

texts

Discret - WordNet

HEO

Analyze

emotions from

text

Discret FOAF WordNet

We have conducted a comparative study of these

existing ontologies using the criteria: 1) Goal: Why

ontology was built? 3) Emotional model (discrete or

continuous) used ? 4) ontology reuse: indicated if

the ontology has integrated existing ontologies or

not. 5) Language: to indicate which dictionary was

used in the detection process.

After comparing existing ontologies, we have

concluding that none of them has integrated the

relationships between emotion and culture. In our

work, we have chosen Human Effective Ontology

(HEO) because it is very rich compared to other

ontology (Marco, 2009). It is more semantically rich

and finer grained ontology: It provides a

standardization of the knowledge of the emotion. It

allows the definition of a common vocabulary that

can be used in describing emotion (Marco, 2009).

4 EMOCULTURE BUILDING

PROCESS

To build our ontology, we have adopted the

construction process inspired by the Ontology

Engineering (OE) method proposed by Psyché

(2004). This process is composed of four steps

which are: 1) Feasibility study: We state why the

ontology is being built, what its users are and which

problems the ontology should solve. 2) The ontology

modeling phase is used to conceptualize the

ontology by collecting information, analyzing it and

extracting the terminology (concepts and relations).

3) ontology operationalization in which the ontology

will be implemented using a programming language.

4) The evaluation allows the verification and the

validation of the ontology.

4.1 Ontology Requirement

Specification

Ontology Requirements Specification refers to the

activity of collecting the requirements that the

ontology should fulfill (e.g., reasons to build the

ontology, target group, intended uses) and possibly

reach through a consensus process (Swati and

Kumar, 2018).

The purpose of building the EMOCULTURE

ontology is to provide a knowledge base able to

represent and store emotions and cultural differences

in emotion expression, handling and regulation

during a collaborative learning session (Emotional

data provided by the user, Emotional knowledge

inferred by the system from this data and Emotional

knowledge detected by the system during user-user

interaction or user-machine interaction).

4.2 Ontology Terminology Extraction

Our ontology Emoculture is built by using a merging

method of two ontology which are SOCUDO

(Ouamani et al., 2016) and HEO (Marco Grassi,

2009).

On the one hand, SOCUDO (Socio-Cultural

Domain Ontology) is generic core ontology. It

models the socio-cultural characteristics of any user

of any application. The purpose of building the

SOCUDO ontology is to provide a knowledge base

able to represent and store socio-cultural knowledge

(socio-cultural characteristics and characteristics that

are socio-culturally sensitive) about the user of any

software system:

‐ Socio-cultural data provided by the user

‐ Socio-cultural knowledge inferred by the system

from this data

‐ Socio-cultural knowledge detected by the system

during user-user interaction or user-machine

interaction.

On other hand, HEO named Human Emotions

Ontology was developed to annotate emotion in

multimedia data. The main class in the ontology is

EmoCulture: Towards an Ontology for Describing Cultural Differences When Expressing, Handling and Regulating Emotions

301

Emotion which is expressed in dimensional and

categorical models. An emotion has an intensity,

appraisals and action tendencies, and it can be

expressed through face, text, voice and gesture. All

the previous features were represented in the

ontology.

The two ontologies were merged together

following the process explained in Table 2. The

process of integrating two (or more) ontologies into

a single one consists in creating a new ontology

from two or more existing ontologies with

overlapping parts, which can be either virtual or

physical. To merge these two ontologies we have

used the algorithm introduced in (Swati and Kumar,

2018) where merging has been illustrated using

“refactor” option of protégé 5.2.0 where various

steps of merging have been illustrated and DLQuery

has been implemented to obtain significant results.

After merging the two ontologies HEO and

SOCUDO, we obtained our ontology EmoCulture

which will be enriched by adding the following

classes:

1- Event: Characterizes internal and external events

that trigger emotions.

2- System Reaction: This class is useful either to

mediate collaboration by regulating certain

emotions and by encouraging others or to

improve the design of emotionally learning

scenarios.

3- Duration: It is used to test if our system reaction

was reliable to regulate negative emotion.

In our case, we have defined for example the

property HasIntensityValue which relates the

concept emotionalState of the ontology SOCUDO

and the concept Individual of the ontology HEO. It

means that any detected emotion has an intensity

value and a category.

Table 2: Algorithm for merging two Ontologies (Swati

and Kumar, 2018).

Input: The two ontologies which will be merged

Output: Resultant MergedOntology

1. Open Second Ontology (Ontology to be merged) in

tool(Protégé 5.2.0).

2. Open First Ontology (Ontology in which merging has to be

done) in the same window in Protégé 5.2.0.

3. Check for similarities.

4. Select ‘Merge Ontologies’ in Refactor Menu.

5. Then, select ‘Merge into existing ontology’ radio button.

6. Select First Ontology as Target Ontology.

7. Resolve inconsistencies in Resultant Merged

Ontology by changing the Full URIs of conflicting classes and

individuals.

8. Check for consistency of Resultant Merged Ontology using

a reasoner.

4.3 Ontology Terminology Validation

As an attempt to validate the concepts and relations

extracted, we have tried to specialize our extracted

concepts from existing upper-level ontologies). To

do so, we have looked for specialization links with

existing upper-level ontologies such as DOLCE,

BFO, SUMO, CYC and YAMATO. We have

chosen DOLCE for the following reasons: 1) it fits

really well with the underlying cognitive aspects that

we have considered in order to build the conceptual

model. 2) It provides many generic concepts that

have been used in order to contextualise those in the

Emotions Ontology. 3) Unlike other, DOLCE

tackles concepts of culture and context and 4)

Compared to other, DOLCE is more semantically

rich and more finely grained. The specialization link

study was done manually based on concept

definitions in both ontologies (See figure 1)

Figure 1: Specialization links between EmoCulture and

DOLCE.

4.4 Ontology Formalisation

Our ontologies were encoded using the Protégé-

OWL editor. The modelling in this phase was guided

by the answers to competency questions described in

the ORSD. By using the integrated OWL plugin, we

have generated automatically the OWL ontologies.

The structure of the proposed Emoculture

ontology is defined in the table 3:

Table 3: EmoCulture classes.

Class Description Instantiation

Arousal

Valence

Dominance

The valence parameter

indicates whether the

emotion is positive or

negative.

Joy, pleasure are part of the

positive emotions. Anger,

frustration, are negative

emotions.

The arousal parameter

indicates the intensity of the

emotion detected.

Valence =

string (positive,

negative)

Arousal = int in

0, 1

KEOD 2019 - 11th International Conference on Knowledge Engineering and Ontology Development

302

Table 3: EmoCulture classes (cont.).

Class Description Instantiation

Community

Presents the actors who are

either learner or tutor.

Name= String

Age = int

Duration

Indicates the duration of

persistence of a negative

emotion

D= int

Event

Presents the internal and

external triggering events of

the emotion.

Event= String

(mourning,

divorce,

material)

Modality

Indicates the type of the

source of emotion detection:

Face, Gesture, text or Voice.

Modality=

String (text,

voice, picture)

Objects

Contains the class from

which we find the

collaborative writing session.

url= String

SystemReaction

Presents the reaction of the

system in case of detection

of negative or positive

emotions.

Reaction=

String

(lesson

reminder,

message,

music, timer)

Word

Presents the language and

the corresponding emotional

dictionary.

DictionaryLink

String

After merging, we have defined for example the

object properties hasIntensity which indicated that

any emotion detected has an intensity value. Also

the relation triggered by which shows that any

emotion is triggered by intern or extern event. The

property isTreatedWith indicates the system reaction

in case of negative emotion detection to regulate it.

Table 4: An excerpt of inference rules implemented within

the proposed ontology based system.

User inputs

EmoCulture

concept value

Adaptation rule

Age <18

Category =

negative

Emotions differ by age:

studies show that adolescents

experience an increase in

negative emotions.

Adolescents have not learned

to deal with (Amr, 2018). As

a reaction, the system will

display content for children

and adolescents in a different

way by applying content

adaptation. For example, the

course description will be

more detailed and questions

will be asked as multiple

choice questions.

Nationality

– African

and external

Event =

divorce

Category =

negative

In African culture, parental

divorce presents academic

difficulties (drop in school

performance and premature

school dropout) and a higher

rate of disruptive behavior

(Amr, 2018). The system

applies the adaptation of the

presentation to the preferred

language.

Nationality

= Japan

Category=

positive

Asian people feel bad during

positive events for example

feeling worried after winning

a victory (Amr, 2018).The

system displays the number of

questions to which the learner

responded rather than the

number of incorrect answers.

Also, the system launches a

relaxing music. Finally, the

system displays a timer at the

top of the page.

4.5 Ontology Operationalisation

First of all, EmoCulture ontology will be instantiated

in part based on the data of an xml file obtained as a

result of the process of registering the user via the

system and filling a form requesting these data.

(country, culturalModel, language, nationality,

religion, BiologicState, DemographicInfo, goal,

habit, need, age, kind, event, educational-Discourse,

virtual-spaces, Modality, Word). Over time, after the

analysis step, the class ArousalValenceDominance,

category and duration will be instantiated. By

detecting a negative emotion, the system applies the

adaptation rules based on the values of the

EmoCulture domain ontology instance concepts for

each user in order to generate the appropriate

recommendations.

5 CONCLUSION AND FUTURE

WORK

In this paper we have first discussed the relationship

between emotions, culture and learning. We have

concluded that emotions have an effect on

motivation, creativity and problem solving and it

differs from one person to another. Second, to model

emotion concept which is rich and who has relations

with different concepts like culture and learning we

have chosen ontologies. Ontology could be the

better solution to model all this influence mechanism

to provide emotionally sensitive intelligent based

ontology system.

Then, based on a comparative study of emotion

ontologies, we have proposed EmoCulture ontology

which was designed with the aim to create a

comprehensive model of emotion considering

cultural differences in emotion expression, emotion

interpretation, system reaction and events. The goal

of the proposed ontology is to help to build

emotionally intelligent system that resolves

emotional conflicts during a collaborative learning

session by applying inference rules based on

EmoCulture: Towards an Ontology for Describing Cultural Differences When Expressing, Handling and Regulating Emotions

303

ontology instantiation. This modeling effort is a first

attempt in performing emotion analysis based on

ontological reasoning. It helps gain better insight

into learner feelings, in order to regulate emotion

and resolve learning conflicts. EmoCulture can be

used for a wide number of applications for emotion

analysis especially those which are based on textual

inputs. It will be used to extract emotional

knowledge form collaborative writing session by

applying learning analytics algorithms in order to be

useful either to mediate collaboration by regulating

certain emotions and by encouraging others or to

improve the design of emotionally sensitive learning

scenarios.

REFERENCES

Ouamani, F., Bellamine Ben Saoud, N. & Hajjami Ben

ghézala, H., (2016a.) Socio-cultural adaptation

approach to enhance intercultural collaboration and

learning.

Ouamani, F., Bellamine Ben Saoud, N. & Hajjami Ben

ghézala, H., (2016b) SOCUDO-SCACLO ontology

ORSD.

C., Monique Boekaerts, (2010), Motivation and self-

regulation: two close friends.

SHELTON, C., (2000), Portraits in emotional awareness.

Educational Leadership.

Philip LeVine & Ron Scollon, (2004), Discourse and

Technology: Multimodal Discourse Analysis.

Georgetown University Press.

Kim-Prieto, C. & M. Eid, (2004), ‘Norms for experiencing

emotions in subSaharan Africa’, Journal of Happiness

Studies, this volume.

Tom Gruber, (1992), What is an Ontology?

D'Mello, S. K., Craig, S. D., Witherspoon A., Sullins J.,

McDaniel B., Gholson, B., & Graesser, A. C.,

Monique Boekaerts, (2010), Motivation and self-

regulation: two close friends.

Arthur C. Graesser and Sidney D'Mello, (2014), Emotions

During the Learning of Difficult Materi.

Jozefien De Leersnyder, Michael Boiger & Batja

Mesquita, (2015), Cultural Differences in Emotions.

Charles Darwin, Phillip Prodger, (1998) The Expression

of the Emotions in Man and Animals Oxford

University Press, 1998 - 472 pages.

Davou, B, (2000), Thought Processes in the Era of

Information: Issues on Cognitive Psychology and

Communication, Athens, Papazissis Publishers.

Ekman, P., (1992), An Argument for Basic Emotions.

Cognition and Emotion.

J. V. Fonou-Dombeu and M. Huisma, (2013) Adaptive

search and selection of domain ontologies for reuse on

the semantic web. Journal of Emerging Technologies

in Web Intelligence, 5(3):230–239.

R. Hoekstra., (2010), The knowledge reengineering

bottleneck. Semantic Web, 1(1, 2):111–115, 2010.

Jiajun Yan, David B. Bracewell, Fuji Ren & Shingo

Kuroiwa. (2008), The creation of a Chinese emotion

ontology based on HowNet.

Vasiliki Kantzara, (2006), Education in Conditions of

Crisis in Greece: An Empirical Exploration.

Kim-Prieto, C. & M. Eid, (2004), ‘Norms for experiencing

emotions in subSaharan Africa’, Journal of Happiness

Studies, this volume.

Marco Grassi. (2009), Developing HEO human emotions

ontology.

Marta Arguedas, Fatos Xhafa, & Thanasis Daradoumis.,

(2015) An ontology about emotion awareness and

affective feedback in elearning.

Yvette Yannick Mathieu., (2005), Annotation of emotions

and feelings in texts.

Stanley Schachter, (1964), The Interaction of Cognitive

and Physiological Determinants of Emotional State.

Swati Negi and Sanjay Kumar, (2018), International

Journal of Applied Engineering Research ISSN 0973-

4562 Volume 13, Number 12 (2018) pp. 10327-10.

Rana Abaalkhail, Benjamin Guthier and Abdulmottalab El

Saddik, Survey on ontologies of affective states and

their influences, 2017.

Michal Ptaszynski, Fumito Masui, Rafal Rzepka and Kenji

Araki, Detecting Emotive Sentences with Pattern-

based Language Modelling, 2012.

Kunihiko Kaneko and Yoshitaka Okada. Building of

Japanese emotion ontology from knowledge on the

web for realistic interactive CG characters, 2013.

KEOD 2019 - 11th International Conference on Knowledge Engineering and Ontology Development

304