Advances in Intelligent Systems Research, volume 166

7th Scientific Conference on Information Technologies for Intelligent Decision Making Support (ITIDS 2019)

Summarizing Emotions from Text Using Plutchik’s

Wheel of Emotions
Mohsin Manshad Abbasi Anatoly Beltiukov
Theoratical Foundation of Computer Sciences Theoratical Foundation of Computer Sciences
Udmurt State University Udmurt State University
Izhevsk, Russian Federation Izhevsk, Russian Federation
[email protected] [email protected]

Abstract—Text is an important and major source of shopping. We will analyze and summarize the emotions from
communication over Internet. It is analyzed to identify text using the concept of Plutchik’s wheel of emotions [1]. In
interesting information and trends of communication. Within 1980’s Robert Plutchik divided emotions into eight main
this work, we are analyzing emotions expressed by people on categories. Half of these emotions are positive emotions, and
Internet using Plutchik’s wheel of emotions. Plutchik’s wheel of the other half are negative ones. They are seen as opposite to
emotions is use as a tool to identify and summarize emotions to each other. We can observe this among secondary emotions,
their primary classes. To accomplish it, we allocate a weight to such as joy is opposite to sadness, surprise is opposite to
each emotion depending upon the class it belongs and its anticipation, trust is opposite to disgust, and anger is opposite
distance from the center of Plutchik’s wheel of emotions. These
to fear. He explained each emotion in detail and divided it
weights are then multiplied by the frequencies of emotions in
text to identify their intensity level. The intensity of each
into subgroups, treating them as secondary and tertiary
emotion is summed up with the intensity of its primary emotion emotions in a wheel-shaped mechanism. His work describes
while summarizing it. We observed that our mechanism a very interesting relationship between emotions, their
effectively summarize emotions from text. This paper is divided intensities and polarities. He also noticed that the intensity of
into several sections. The methodology, results and conclusions an emotion is high when lies in the center of the wheel and it
are discussed in detail in their respective sections. decreases as the distance from the center increases.

Keywords—Plutchik wheel, emotion, class, summarization,

internet, blogs, communication

I. INTRODUCTION
Text analysis is the subject of major research over the
recent decades. This is due to availability of online and
offline data. The online text is growing at a very high pace.
There is a need to develop tools and methods for solving text
analysis problems. Text analysis has a lot of applications. It
can be used to detect fraud in online transactions, predict the
future, identify groups and people, etc.
The theme of our research is to develop methods that can
be used to analyze emotions from a text efficiently and then
use them for classification of the text and its summarization.
Today we have various social networks, blogs, and groups
where people can express and convey their emotions to
others. This emotion transfer can be constructive if it is
Fig. 1. Plutchik wheel of emotions
positive, and can be destructive if it is negative. There are
still challenges in text analysis such as identifying a hate
group online using text analysis. This is a question that still II. HISTORY OF EMOTION ANALYSIS
requires efforts. Emotions from a text are used to determine
The analysis of emotions was began with the
public opinion about a topic and to predict a future event
development of the very first general survey system in 1966.
based on it, such as election results, reaction to governmental
It was started by Philip Stone at University of Harvard, and
policies, etc. The need of such system was seriously felt
it was probably the first stage of text analysis and
during the crisis of the Middle East, where governments
identification of emotions from it [2]. After that, text
failed to understand the people’s emotions. During these
analysis in different languages started gradually. Several
events people used the online text as the major source to
scientist made remarkable contributions in text processing.
organize protests against the government and its policies.
Important among them was the contribution of Jayne Wiebe,
Similarly, American elections were linked to online groups
Peter Turney and Vasileios Hatzivassiloglou in the early
and people who spread propaganda.
1990s. In 1990, Jane Wiebe defined the term “subjectivity”
In this paper, we will use a text taken from an online blog for researching information retrieval [3]. Later, in 1997
in which people expressed their emotions about online Hatzivassiloglou revealed semantic orientation of adjectives

Copyright © 2019, the Authors. Published by Atlantis Press.

This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). 291
 Advances in Intelligent Systems Research, volume 166

[4]. A few years later, in 2002, Peter Turney offered his 𝐼𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦 𝑜𝑓 𝐸𝑚𝑜𝑡𝑖𝑜𝑛 𝐼𝐸𝑖 = 𝐶𝑖 × 𝑊𝑖 ()
revolutionary approach of “Thumbs up” and “Thumbs
down” for positive and negative review of a text [5]. During 𝑊𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝐸𝑚𝑜𝑡𝑖𝑜𝑛 𝑊𝑖 𝑝𝑟𝑖𝑚𝑒𝑚𝑜 𝑢𝑝𝑑𝑎𝑡𝑒𝑑 =
the same year Pang proposed the mood lexicon manually for 𝑊𝑖 𝑝𝑟𝑖𝑚𝑒𝑚𝑜 + 𝐼𝐸𝑖 ()
text analysis [6]. In 2009, Denecke reported an interesting
study conducted in several areas to demonstrate usefulness V. EXPERIMENT & RESULTS
of previous polarity estimates from SentiWordNet [7].
Recently the scientists started using different data mining We applied our methodology on the text to identify
tools, statistical methods, supervised, semi-supervised emotions using Plutchik’s wheel of emotions. We observed
learning techniques and combinations to identify and a list of emotions from the text with different polarities.
analyze emotions from a text. First, we searched for positive emotions, counted their
frequencies and identified their weights. On multiplying the
III. RELATED WORK frequency of an emotion by its weight we observed the
Going beyond polarity in sentiment analysis is currently intensity of a particular emotion in the text. On summing up
not studied. However, several examples can be found where the intensities of all the positive emotions together we got an
methods are used to collect more information than just overall intensity of positive emotions in the text as described
polarity, such as the use of recursive auto-coder is to predict in table I below.
mood distributions in different dimensions [8]. The authors TABLE I. POSITIVE EMOTIONS IN TEXT
also promote efficient calculations using a structure they
called SenticNet [9,10]. Sentimental dimensions of this Positive emotions in the text
Emotions
structure are modeled in the hourglass model, which is a Frequency Weight Value
derivative of Plutchik’s emotion wheel [1]. In 2012 the Trust 15 3 45
author gathered a large collection of tweets and experimented
with self-signed hash tags of emotions [11]. Surprise 12 3 36

It is difficult to define standards for emotion analysis Interest 9 2 18

from a text because emotions are usually subjective and Amazement 5 4 20
cannot be clearly defined. The works of Parrott, Plutchik and
Schroder aimed at defining standards in this area by Acceptance 4 2 8
determining the minimum set of basic emotions from which Love 5 2 10
complex emotions can be constructed [12, 1, 13]. In 2012 the
authors developed methods for emotion reasoning [14]. In Awesome 7 1 7
our work we are using Plutchik’s wheel of emotions to
identify, analyze and summarize emotions from text.
From the above analysis, the most frequent positive
IV. METHODOLOGY emotions in the text can be observed. We can also sum up the
frequencies of all positive emotions to get an overall
In this work, we are purposing a new methodology that
frequency of positive emotions in text. The results of the
uses Plutchik’s wheel of emotions for identification and
above analysis are further elaborated using a graph plotted
analysis of emotions from a text. The work begins with the
below in Fig. 2. Positive emotions are represented with green
selection of a text from an online blog. In this blog people
colour whereas the blue colour represents the emotion
expressed their opinion about shopping online. Our system
‘Awesome’ that, according to Plutchik, is a feeling aroused
will be engaged in mining and identifying emotions from the
because of surprise and fear that makes it a neutral emotion.
text. Each emotion in Plutchik’s wheel of emotion will be
given a certain weight depending upon its intensity.
Positive Emotion Using Plutchik's Wheel
We will begin by identifying the most intense emotions
20
from the text that are close to the center of Plutchik’s wheel
and then broaden up the search for its secondary and tertiary
15
emotions along the radius of the wheel. The weight of the
emotions close to the center of the wheel is 4, which is also 10
the maximum level of intensity in Plutchik’s wheel of
emotions. For the emotions in second layer of Plutchik wheel 5
of emotions, the weight is 3 and it gradually decreases by l
from layer to layer of Plutchik’s wheel of emotions as we 0
move far from the center. We will count the total occurrences 1
of an emotion in text 𝐶𝑖 as in (1). Then by multiplying it
with the weight of an emotion𝑊𝑖 , we get the intensity 𝐼𝐸𝑖 of Trust surprise interest
each emotion in the text as in (2). The intensity of an emotion amazement acceptance love
is added to the weight of primary emotion 𝑊𝑖 𝑝𝑟𝑖𝑚𝑒𝑚𝑜 and
awesome
using this mechanism the weights of all primary emotions are
updated as in (3).
Fig. 2. Positive emotion of Plutchik’s wheel identified in the text
𝐸𝑚𝑜𝑡𝑖𝑜𝑛 𝐶𝑜𝑢𝑛𝑡 𝐶𝑖 = ∑𝑛𝑗=1 𝑂𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 𝑜𝑓 𝐸𝑚𝑜𝑡𝑖𝑜𝑛𝑖𝑗 ()

292
 Advances in Intelligent Systems Research, volume 166

Similarly, we observed negative emotions in the text, Plutchik’s wheel that lies close to the center as represented
counted their frequencies and identified their weights. On in the Table 3 below. The table represents the eight primary
multiplying the frequency of an emotion by its weight we emotions of Plutchik’s wheel of emotion and their weights
observed the intensity of a particular emotion in the text. On in the text.
summing up the intensities of all the negative emotions TABLE III. PRIMARY EMOTIONS OF PLUTCHIK’S
together we got an overall intensity of negative emotions in WHEEL OF EMOTIONS IN TEXT ALONGWITH THEIR
the text as described in Table II below WEIGHTS
TABLE II. NEGATIVE EMOTIONS IN TEXT

Primary Emotion Weight

Negative emotions in the text
Emotions Rage 12
Frequency Weight Values
Sadness 6 3 18 Vigilance 18
T
Anger 4 3 he 12 Ecstasy 0
resu Admiration 63
Rage 4 4 16 lt of
Boredorm 2 2 4 abo Terror 13
ve Amazement 43
Fear 2 3 6 anal
Awesome 7 1 7 ysis Grief 18
sho Loathing 4
ws the frequency of occurrence of negative emotions in the
text. By summing together the frequencies of all the negative
emotions from text, we can get an overall frequency of The analysis result done for the identification of primary
negative emotions in text. The same way we can calculate the emotions of Plutchik’s wheel of emotions from the text and
overall weights and values of negative emotions in text. In their respective weights are plotted below in Fig. 4.
general, it is observed that the frequency of negative
emotions in text is comparatively low to positive emotions.
These emotions are further explained using a graph plotted
below in Fig. 3. The red colour represents negative emotions Plutchik’s wheel of emotion
whereas blue colour represents a neutral feeling. 70
60
50
40
Negative emotion from Plutchik’s wheel of
30
emotion
20
16 10
14 0
12 1
10
Rage Vigilance Ecstasy
8
Admiration Terror Amazement
6
4 Grief Loathing
2
0
Fig. 4. Representing emotion of Plutchik’s wheel along with
1 their weights identified from text
Sadness Anger Rage Boredom Awe

The above analysis describes the summarized emotional

Fig. 3. Shows negative emotion and their frequencies in text contents in the text. It is observed that online shopping is an
admiration and amazement for the people. From this analysis
it can be concluded that in general people consider online
It is observable that the number of negative emotions, shopping a positive mechanism and trust it. However, the
their frequencies and weights are less than half of the weight feelings of rage, terror, grief and loathing can sometimes
of positive emotions in the text. Based on the analysis of the distract a customer while doing shopping online.
text according to Plutchik’s wheel of emotion it is observed VI. CONCLUSION
that in the text people mostly express positive emotions
about online shopping. To summarize the emotions to their We observed that our methodology affectively
primary or intensive emotion of Plutchik’s wheel we add the summarizes emotions in the text. Robert Plutchik’s work
weights of emotions to the most intense emotions of describes the main existing emotions and is still important

293
 Advances in Intelligent Systems Research, volume 166

while analyzing emotions in a text. It is a simple way of [7] K. Denecke,“Are SentiWordNet scores suited for multi-domain
identifying polarity of a document and summarize its sentiment classification?”, In: Proc. Of the fourth International
Conference on Digital Information Management, University of.
emotional contents. However, neutral emotions and feelings Michigan, USA, 1-4 November 2009.
described in Plutchik’s wheel of emotion complicate the [8] R. Socher, J. Pennington, E.H. Huang, A.Y. Ng, C.D. Manning,
process of summarization. In future, we will propose a “Semi-supervised recursive autoencoders for predicting sentiment
mechanism to avoid complications while summarizing distributions”, In Proceedings of the Conference on Empirical
neutral emotions. Methods in Natural Language Processing, EMNLP’11, USA, 2011,
pp. 151–161.
REFERENCES [9] E. Cambria, A. Hussain, “Sentic Computing: Techniques, Tools, and
Applications”, Berlin Heidelberg, Springer, 2012.
[1] P. Robert, Emotion: Theory, research, and experience, Theories of
emotion, vol. 1. New York: Academic, 1980. [10] E. Cambria, C. Havasi, A. Hussain, “Senticnet 2: A semantic and
affective resource for opinion mining and sentiment analysis”, In
[2] P. Stone, D. Dunphy, M. Smith, “The General Inquirer: A Computer Proceedings of the Twenty-Fifth International Florida Artificial
Approach to Content Analysis”. MIT Press - Cambridge, 1966. Intelligence Research Society Conference. AAAI Press, 2012.
[3] Wiebe, M. Janyce, “Identifying Subjectivity characters in Narrative”, [11] S. Mohammad. “#emotional tweets. In SEM 2012”, In the
In Proceedings of the 13th International Conference on Computational Proceedings of 1st Joint Conference on Lexical and Computational
Linguistics, Helsinki Morristown NJ: Association of Computational Semantics, Association for Computational Linguistic, 2012, pp. 246–
Linguistic, 1990, pp. 401-408. 255.
[4] Vasileios H., Kathleen R.M. “Predicting the Semantic Orientation of
[12] W.G. Parrott. Emotions in Social Psychology. Psychology Press,
Adjectives” EACL '97”, In Proceedings of the eighth conference on
Philadelphia, 2001.
European chapter of the Association for Computational Linguistics,
Madrid, Spain, July 07-12, 1997. [13] M. Schroder, P. Baggia, F. Burkhardt, C. Pelachaud, C. Peter, E.
Zovato, “ Emotional - an upcoming standard for representing
[5] P. Turney, “Thumbs Up or Thumbs Down? Semantic Orientation
emotions and related states”, In Proceedings of the 4th International
Applied to Unsupervised Classification of Reviews”, In Proceedings
Conference on Affective Computing and Intelligent Interaction, vol. 1
of the 40th Annual Meeting of the Association for Computational
ACII’11, , Berlin, Heidelberg. Springer-Verlag, 2011, pp. 316– 325.
Linguistics, Philadelphia, 2002, pp. 417-424.
[14] E. Cambria, C. Havasi, A. Hussain, “Senticnet 2: A semantic and
[6] B. Pang, L. Lee,“Thumbs up? Sentiment Classification using Machine
affective resource for opinion mining and sentiment analysis”, In
Learning Techniques”, In Proceedings of the Conference on Empirical
Proceedings of the Twenty-Fifth International Florida Artificial
Methods in Natural Language, Philadelphia, 2002, pp.79-86.
Intelligence Research Society Conference, AAAI Press, 2012.

294