Keyword-based Approach for Lyrics Emotion Variation Detection

Ricardo Malheiro

1,2

, Hugo Gonçalo Oliveira

, Paulo Gomes

and Rui Pedro Paiva

Center for Informatics and Systems of the University of Coimbra (CISUC), Pólo II, Pinhal de Marrocos 3030-290,

Coimbra, Portugal

Miguel Torga Higher Institute, Largo Cruz de Celas 1, 3000-132, Coimbra, Portugal

Keywords: Music Information Retrieval, Lyrics Music Emotion Recognition, Lyrics Music Emotion Variation Detection,

Keyword based Approach.

Abstract: This research addresses the role of the lyrics in the context of music emotion variation detection. To

accomplish this task we create a system to detect the predominant emotion expressed by each sentence (verse)

of the lyrics. The system employs Russell’s emotion model and contains 4 sets of emotions associated to each

quadrant. To detect the predominant emotion in each verse, we propose a novel keyword-based approach,

which receives a sentence (verse) and classifies it in the appropriate quadrant. To tune the system parameters,

we created a 129-sentence training dataset from 68 songs. To validate our system, we created a separate

ground-truth containing 239 sentences (verses) from 44 songs annotated manually with an average of 7

annotations per sentence. The system attains 67.4% F-Measure score.

1 INTRODUCTION

Music emotion recognition (MER) is gaining

significant attention in the Music Information

Retrieval (MIR) scientific community. In fact, the

search of music through emotions is one of the main

criteria utilized by users (Vignoli, 2004).

Real-world music databases from sites like

AllMusic or Last.fm grow larger and larger on a daily

basis, which requires a tremendous amount of manual

work for keeping them updated. Unfortunately,

manually annotating music with emotion tags is

normally a subjective, expensive and time-consuming

task. This should be overcome with the use of

automatic recognition systems (Hu and Downie,

2010).

Most of the early-stage automatic MER systems

were based on audio content analysis (e.g., (Lu et al.,

2006)). Later on, researchers started combining audio

and lyrics, leading to bi-modal MER systems with

improved accuracy (e.g., (Hu and Downie, 2010),

(Hu et al., 2009), (Laurier et al., 2008)). This does not

come as a surprise since it is evident that the

importance of each dimension (audio or lyrics)

depends on music style. For example, in dance music

audio is the most relevant dimension, while in poetic

music (like Jacques Brel) lyrics are key.

Several psychological studies confirm the

importance of lyrics to convey semantical

information. Namely, according to Juslin and Laukka

(2004), 29% of people mention that lyrics are an

important factor of how music expresses emotions.

Also, Besson et al. (1998) have shown that part of the

semantic information of songs resides exclusively in

the lyrics.

Each song is normally associated to a

predominant emotion (e.g., happiness, sadness),

which corresponds to the emotion perception of the

listeners concerning that song. Music Digital

Libraries (MDL) like AllMusic take this into account

to classify songs in their sites.

There are songs in which the predominant

emotion is easy to determine, i.e., for the majority of

listeners the perceived emotion is the same or almost

the same throughout the song, while in others the

perceived emotion varies significantly along the song.

The example below, from the song “Kim” by

Eminem, illustrates emotion variation:

Aw look at daddy's baby girl

That's daddy baby

Little sleepy head

Yesterday I changed your diaper

Wiped you and powdered you.

How did you get so big?

Can't believe it now you're two

Malheiro, R., Oliveira, H., Gomes, P. and Paiva, R.

Keyword-based Approach for Lyrics Emotion Variation Detection.

DOI: 10.5220/0006037300330044

In Proceedings of the 8th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2016) - Volume 1: KDIR, pages 33-44

ISBN: 978-989-758-203-5

Baby you're so precious

Daddy's so proud of you

Sit down bitch

If you move again I'll beat the shit out of you

Don't make me wake this baby

She don't need to see what I'm about to do

Quit crying bitch, why do you always make me shout at

you?

…

The lyric changes abruptly from emotions like serene

joy and relaxation to anger and tension.

Thus it is important to investigate the time-

varying relationship between music and emotion.

In the audio domain, there are a few studies

tackling Music Emotion Variation Detection

(MEVD), e.g., (Schubert, 1999), however, to the best

of our knowledge, we are not aware of any research

focused in this specific area of Lyrics Music Emotion

Variation Detection (LMEVD).

In this work, we propose a novel a keyword-based

approach (KBA) to classify song verses according to

Russell’s emotion model (Russell, 1980). To validate

our model, we create a new manually annotated

dataset with 239 sentences taken from 44 song lyrics.

This paper is organized as follows. In section 2,

the related work is described and discussed. Section 3

presents the methods employed in this work,

particularly the creation of the ground truth and the

description of the architecture of our KBA model.

The results attained by our system are presented and

discussed in Section 4. Finally, section 5 summarizes

the main conclusions of this work.

2 RELATED WORK

The relations between emotions and music have been

a subject of active research in music psychology for

many years. Different emotion paradigms (e.g.,

categorical or dimensional) and taxonomies (e.g.,

Hevner, Russell) have been defined (Hevner, 1936),

(Russell, 1980) and exploited in different

computational MER systems.

One of the most well-known dimensional models

is Russell’s circumplex model (Russell, 1980), where

emotions are positioned in a two-dimensional plane

comprising two axes, designated as valence (V) and

arousal (A), as illustrated in Figure 1. According to

Russell (Russell, 2003), valence and arousal are the

“core processes” of affect, forming the raw material

or primitive of emotional experience. We use in this

research a categorical version of this Russell’s model,

so we consider that a sentence belongs to quadrant 1

if both dimensions are positive; quadrant 2 if V is

smaller than 0 and A is bigger than 0; quadrant 3 if

both dimensions are negative and quadrant 4 if V is

bigger than 0 and A is smaller than 0. The main

emotions associated to each quadrant are illustrated in

Figure 1.

Figure 1: Main screen of the annotation platform.

Identification of musical emotions from lyrics is

still in an embryonic stage. Most of the previous

studies related to this subject used general text instead

of lyrics and polarity detection instead of emotion

detection. More recently, Lyrics MER (LMER) has

gained significant attention in the MIR scientific

community.

Each song is normally associated to a

predominant emotion (e.g., happiness, sadness),

which corresponds to the emotion-perception of the

listeners concerning to that song. Music Digital

Libraries (MDL) like AllMusic take this into account

to classify songs in their sites.

Human perception of the emotions expressed by a

song depends normally on several dimensions which

compose a song (e.g., audio, lyrics).

In the audio domain, a few works have addressed

MEVD. Namely, Schubert (Schubert, 1999) proposes

a time series analysis method and Korhonen

(Korhonen et al., 2006) tackles MEVD as a system

identification method, exploiting the temporal

information among the music segments. Other

authors, e.g., (Yang et al., 2006) do not integrate

temporal information and perform emotion prediction

independently for each music segment.

Concerning Lyrics MEVD, we are not aware of

any research of this type.

According to Chopade (2015), emotions may be

expressed in lyrics by one word or a bunch of words.

The sentence level emotion detection method plays a

crucial role to trace emotions or to search out the cues

for generating such emotions. Sentences are the basic

information units of any document. For that reason,

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval

the document level emotion detection method

depends on the emotion expressed by the individual

sentences of that document that successively relies on

the emotions expressed by the individual words.

According to Oxford Dictionaries (http://www.

oxforddictionaries.com), a verse is a group of lines

that form a unit in a poem or a song. Thus, regarding

the typical structure of a lyric (based on verses as in

poetry or based on sentences as in prose), we think

composers convey ideas and emotions having, as

basic unit of information, respectively the verses and

the sentences. In our work we use interchangeably the

terms sentence and verse.

There are basically three types of approaches to

work with the task of detection of emotions in text

(Binali et al., 2010):

 Learning-based approaches (LBA). LBA is

based on the use of a trained classifier to

categorize input text into emotion classes by

using keywords as features. To adapt to a new

domain we have to supply a large training set

to a machine learning algorithm to build a new

classification model. Thus, we use the features

extracted from the corpora. Here, the more

difficult step is normally acquiring the corpora,

e.g., (Yang et al., 2007).

 Keyword-based approaches (KBA). KBA is

based on the presence of keywords in text. It

typically involves steps such as pre-processing

with a parser and search based on an emotion

dictionary. This technique is domain specific,

relies on the presence of keywords for accurate

results and requires pre-processing for

improved accuracy results, e.g., (Chunling et

al., 2005), (Hancock et al., 2007) and (Li et al.,

2007).

 Hybrid approaches (HA). Hybrid approaches

are a combination of the previous methods.

These approaches can improve results from

training a combination of classifiers and adding

knowledge-rich linguistic information from

dictionaries and thesauri, e.g., (Aman and

Szpakowicz, 2007), (Binali et al., 2010) and

(Kao et al., 2009).

Some authors, e.g., (Chopade, 2015), consider

lexicon-based approaches (which counts the number

of words of a lexicon in the text) as a 4th independent

approach, while others, e.g., Binalli et al., (2010),

consider this approach as part of a KBA.

In our work, we use a KBA to detect emotions in

sentences and, then, to understand how the emotions

vary along the lyric.

There are some limitations associated normally to

KBA, which we attempt to mitigate.

1) Ambiguity in keyword definitions, i.e., the

meanings of keywords could be multiple and

vague, as most words could change their

meanings according to different usages and

contexts. Our system performs disambiguation

to some extent, since it retrieves the definitions

of the words from Wordnet (WN) (Miller,

1995) and counts on their words to the emotion

detection task. If we have for instance the word

"crush" in "he had a crush on her", applying

POS tags, "crush" is a noun and its definition

from WN is "temporary love of an adolescent".

If we have the same word in the sentence "He

crushed the car", crushed here is a verb and the

definition is "break into small pieces".

Probably this will not work in all situations,

even because WN may have more than one

definition for each grammatical class (e.g.,

noun). We consider the most common case.

Our system retrieves also from the WN

synonyms of the words and the same happens

here, i.e., depending on the grammatical class

the synonyms list is different.

2) Emotions are recognized only in the presence

of keywords. In our work, the retrieved

synonyms and definitions help to extent our

keyword list.

3 METHODS

3.1 Dataset Construction

To accomplish emotion variation detection based on

song lyrics, we need a ground-truth composed of

annotated sentences (verses). We consider the

sentence as the basic unit for the lyric. Hence, through

the variation of emotions along several consecutive

sentences, we can observe the way the emotions vary

along the lyric.

3.1.1 Validation Set

Data Collection and Pre-Processing

To construct our validation dataset, we collected 44

song lyrics, belonging to several genres. Musical

genres are distributed as follows: pop/rock (6 songs),

pop (18 songs), rock (8 songs), heavy-metal (3

songs), folk (2 songs), R&B (1 song), hip-hop (4

songs) and country (2 songs).

In the selection of the songs, we tried that the

songs were distributed uniformly for the 4 quadrants

Keyword-based Approach for Lyrics Emotion Variation Detection

of the Russell’s emotion model, according to our a

priori perception (11 for each quadrant).

The obtained lyrics were then pre-processed to

improve their quality. Namely, we performed the

following tasks:

 Correction of orthographic errors;

 Elimination of text not related with the lyric

(e.g., names of the artists, composers,

instruments);

 Elimination of common patterns in lyrics such

as [Chorus x2], [Vers1 x2], etc.;

 Complementation of the lyric according to the

corresponding audio (e.g., chorus repetitions in

the audio are added to the lyrics).

Annotation and Validation

To simplify the sentence annotation process, we

decided to create a web application in the Google App

Engine. This app was disclosed for the annotators

through direct invitations, mailing lists and social

networks.

Initially, the annotators have to register in the web

application and then confirm the email sent by the

application for their emails. The session starts after

authentication. The following items shows some

characteristics of the platform:

 The start-up screen shows information about

the goals of the research and instructions to

accomplish the task;

 The sentences are presented randomly to the

annotators;

 The same sentence does not appear twice for

the same annotator, even in different sessions;

 If a song has several repetitions of the same

sentence (e.g., chorus), the sentence only

appears once to the annotator;

 The annotator can continue his work in

different sessions;

 The annotator can classify any number of

sentences;

 If the annotator classifies all the sentences in

the database, the system shows, at the end, a

message saying that there are no more

sentences to annotate.

Figure 2 shows the application interface. The

annotator should read the sentence and then pick the

most appropriated choice with the mouse in the pie

chart.

If the user hovers with the mouse the several

regions in the pie chart (e.g., Q1, Q2, Q3, Q4), the

system shows the most predominant emotions from

that quadrants.

Finally, the application provides instructions on

how to correctly perform the task:

1. Read the sentence with attention;

2. Try to identify the basic predominant emotion

expressed by the sentence, according to the sets

of emotions (quadrants) in the figure (Figure

3);

3. If you think the sentence does not convey any

emotion, select the option Neutral.

To further improve the quality of the annotations,

the users were recommended not to use any known

previous knowledge about the lyric when they

recognized the song through the sentence, not to

search for information about the lyric neither the song

on the Internet or another place and to avoid tiredness

by taking a break and continuing later.

The 44 employed lyrics have a total of 330

sentences and we obtained an average of 7

annotations per sentence.

The classification of each sentence corresponds to

the most representative class among all the

annotations. In case of a draw the sentence is ignored.

This situation happened in 9 sentences.

Since our goal is to build a system to classify

sentences in 1 of the 4 possible quadrants, we ignore

the sentences annotated as neutral sentences, which

happened 18 times. In the future we intend to expand

our model to detect previously if a sentence is

emotional or non-emotional.

Additionally, we also ignore the repetitions of

verses and chorus, that is, we consider only one

occurrence of each repeated section. This excludes

more 64 sentences. So, at the end, we obtained 239

sentences in total (330 – 9 – 18 – 64).

The following examples illustrate the process of

annotation for some of these sentences: 1) the

sentence “I’ve got peace like a river, I’ve got peace

like a river in my soul” from the song “Peace like a

river” (Veggie Tales) has 7 annotations, all of them

in Q4; 2) the sentence “Well now she’s gone; even

though I hold her tight, I lost my love, my life, that

night” from the song “Last kiss” (Pearl Jam) has 6

annotations, all of them in Q3; 3) the sentence “At the

end of all this hatred lies even deeper hate, their

darkness has defeated you, your lifeline running

backwards” from the song “Blood on your hands”

(Arch Enemy) has 10 annotations, 9 on Q2 and 1 on

Q3, so the sentence was annotated in Q2; 4) the

sentence “You’re the light, you’re the night, you’re

the color of my blood, you’re the cure, you’re the

pain, you’re the only thing I wanna touch, never knew

that it could mean so much” from the song “Love me

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval

Figure 2: Main screen of the annotation platform.

Figure 3: Predominant emotions by quadrant.

like you do” (Ellie Goulding) has 7 annotations, 6 in

Q1 and 1 in Q2, so the sentence was annotated in Q1.

The consistency of the ground truth was evaluated

using Krippendorff’s alpha (Krippendorff, 2004), a

measure of inter-coder agreement. This measure

achieved, for the classes Q1, Q2, Q3, Q4 and N, a

value of 53%. This is considered a moderate

agreement among the annotators (Landis and Koch,

1977).

According to quadrants, the sentences are

distributed in the following way (Table 1).

As can be observed in Table 1, the final validation

dataset is not very balanced. Particularly, quadrants 3

Table 1: Distribution of the sentences by quadrant.

Quadrant # Sentences

Q1 86

Q2 67

Q3 47

Q4 39

Total 239

and 4 turned out to obtain a much lower number of

samples. However, as described below, the training

set is nearly balanced.

3.1.2 Training Set

As will be described later on, our system employs a

number of parameters that need to be tuned. To this

end, we have additionally created a training dataset.

This dataset was annotated according to Russell’s

model (4 quadrants) by 2 persons and we just

considered sentences in which there were unanimity.

We considered a total of 129 lyric sentences from 68

songs, distributed across the four quadrants according

to Table 2. As can be seen, this training is nearly

balanced.

Table 2: Distribution of the sentences by quadrant.

Quadrant # Sentences

Q1 35

Q2 36

Q3 27

Q4 31

Total 129

3.2 Sentence Emotion Recognition

Model (SERM)

We use a knowledge-based approach to create a

Sentence Emotion Recognition Model (SERM). This

model uses NLP techniques to assign to each sentence

an emotion quadrant in Russell’s plane, following an

unsupervised approach.

Figure 4 shows the architecture of our system.

Keyword-based Approach for Lyrics Emotion Variation Detection

We use two lexicons to retrieve the values of

valence and arousal from the words: Emotion

Dictionary (ED) and Dictionary of Affect in

Language (DAL) (Whissell, 1989).

To create de ED dictionary:

1. We define as seed words the emotion terms

defined for each quadrant and based on

Russell’s plane (see Figure 2).

2. From these terms, we consider for the

dictionary only the ones present in the DAL or

the ANEW (Bradley and Lang, 1999)

dictionaries. In the DAL, we assume that

pleasantness corresponds to valence, and

activation to arousal, based on (Fontaine,

2013). We employ the scale defined in the

DAL: arousal and valence (AV) values from 1

to 3. If the words are not in the DAL dictionary

but are present in ANEW, we still consider the

words and convert the arousal and valence

values from the ANEW scale to the DAL scale.

3. We then extend the seed words through

Wordnet Affect (Strapparava and Valitutti,

2004), where we collect the emotional

synonyms of the seed words (e.g., some

synonyms of joy are exuberance, happiness,

bonheur and gladness). The process of

assigning the AV values from DAL (or

ANEW) to these new words is performed as

described in step 2.

4. Finally, we search for synonyms of the

gazetteer’s current words in Wordnet and we

repeat the process described in step 2. Steps 2,

3 and 4 are repeated iteratively while we add at

least a word in an iteration.

Before the insertion of any word in the dictionary

(from step 1 on), each new proposed word is validated

or not by two persons, according to its emotional

value. There should be unanimity between the two

subjects. The two persons involved in the validation

were not linguistic scholars but were sufficiently

knowledgeable for the task.

Based on the procedure above, the emotion

dictionary ended up with 1246 words.

Next, we will explain in detail each one of the

modules.

After reading a directory containing the lyrics, the

lyrics are divided into sentences (verses) and the

system processes one sentence at a time.

Removal of Punctuation Marks

The punctuation marks of are first removed. For

example the sentence: “Martha, are you playing

cello?” is transformed in “Martha are you playing

cello”

Word Transformation

In this step, the words in the sentence are transformed

according to the rules below, if necessary:

 Verbs in gerund finished by the character “’”.

The character “’” is replaced by the character

“g” (e.g., sittin’ → sitting, sippin’ → sipping);

 Ended by the characters “’s”. These two

characters are removed from the word (e.g., the

sentence “my mother’s house” changes to “my

mother house”);

 Contraction of verbs or simplification of words

due to informal text or slang. These words are

corrected according to a dictionary (e.g., ain’t

→ am not, couldn’t → could not, won’t → will

not, they’re → they are, hadn’t → had not,

gonna → going to, gotta → got to, ‘cause →

because, ‘til → until, cuz → because, ‘em →

them).

VANA Detection

Several works such as (Lu et al., 2006) consider that

only verbs (V), adjectives (Adj), nouns (N) and

adverbs (A) can convey emotions or can help to

understand the emotions.

We follow the same assumption, so we applied a

POS tagger (Taylor et al., 2003) to identify the

VANA words.

For example, applying a POS tagger to the

sentence “Martha. Are you playing cello?” we obtain

“Martha/NNP are/VBP you/PRP playing/VBG

cello/NN”, so the VANA words are “Martha”, “are”,

“playing” and “cello”.

SVANA Detection (Selected VANA)

Among the VANA words from the original sentence,

we consider for the calculation of the emotion

conveyed by the sentence, the adjectives, the nouns

(except proper nouns) and the verbs (except auxiliary

verbs). So, from the sentence “Martha/NNP are/VBP

you/PRP playing/VBG cello/NN”, only two words

(playing and cello) are selected words to go to the

next level.

Modifiers Detection

In this step we will identify words that can change the

emotion of the other sentence’s words. In this class of

words (modifiers) we may include:

 Negations such as for example not, no, never;

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval

 Adverbs such as for example very, extremely,

little.

In these modifiers we have always a cause and an

object. The cause is the modifier and the object is the

word where we can apply the modifier (see Table 3).

Table 3: Example of modifiers in sentences.

Sentence Modifier Object

I’m not sad not sad

I’m very happy very happy

Our system detects automatically the modifiers

and the corresponding objects.

When the modifier is a negation, the object is not

considered anymore for the calculation of the

sentence’s emotion (Agrawal and An, 2012). In the

sentence “I’m not sad”, the emotion conveyed is not

necessarily an emotion from the 1

quadrant (e.g.,

happiness). It can be for example an emotion from the

quadrant (e.g., serene joy), i.e., the emotion

conveyed is not necessarily the antonym of the object.

So we have decided to not consider this kind of

words.

To the best of our knowledge, we did not find any

dictionary of adverbs classified by intensity. Hence,

we decided to create one, so the modifiers were

classified according to its intensity in a range between

-5 (minimum intensity) and 5 (maximum intensity) by

one person, who is not linguistic scholar but is

sufficiently knowledgeable for the task. The

dictionary has 102 adverbs.

Table 4 shows some examples of adverbs

classified according to its intensity.

Table 4: Examples to the weight of the word “happy” in

sentences with adverb modifiers.

Sentence Intensity

extremely

very

little

-3

rarely

-5

Assignment of Word Weights

These words get a specific weight (WL1), whose

value is set as described below (the same value for

each word). However, the weights can be modified if

they are objects of specific modifiers.

They may increase or decrease if the modifier is

an adverb or it may become zero if the modifier is a

negation. We have also other different possible

weights according to the provenience and the

emotional features of the words.

Therefore, we consider the following weights:

 WL1: Represents the weight of the SVANA

words – adjectives, nouns (except proper

nouns) and verbs (except auxiliary verbs) – that

belong to the original sentence.

 WL2: If the selected words from the original

sentence belong to the lexicon ED, then the

SVANA words of their definitions (see the

“retrieval of definitions” step, below) get a

weight with value WL2. Words that do not

belong neither to ED nor DAL, but their

synonyms belong to ED, also get a weight with

value WL2.

 WL3: If the selected words from the original

sentence do not belong to the lexicon ED then

the SVANA words of their definitions get a

weight with value WL3. Words that do not

belong neither to ED nor DAL, but their

synonyms do not belong to ED but belong to

DAL, get a weight with value WL3.

 WL4 and WL5: Represent weights to multiply

additionally by the initial weight of the words,

when these words belong to ED (WL4) and to

DAL (WL5).

After the assignment of the word weights, we

have to update the weights according the detection of

modifiers seen previously. If the selected word is

object of a modifier of the type negation then the word

will have the weight zero (e.g., the word “happy” in

the sentence “I’m not happy”).

When the modifier is an adverb, the weight of the

object, for the calculation of the emotion, can be

increased or decreased. Suppose for instance that the

word “happy” in the sentence “I’m happy” has an

initial weight of 10 and suppose that in our dictionary

the adverbs, extremely, very, little and rarely have

respectively the intensity values of 5, 3, -3 and -5. We

can see in Table 5 the weight of the object “happy”

for sentences using the previous adverbs as modifiers.

Observing the table, the weight of the object in the

first sentence (15) is obtained from the sum of the

weight associated to the word “happy” (10) by the

value associated to the modifier “extremely” (5).

Table 5: Examples to the weight of the word “happy” in

sentences with adverb modifiers.

Sentence Weight of the word happy

I’m extremely happy

I’m very happy

I’m happy

I’m little happy

I’m rarely happy

Keyword-based Approach for Lyrics Emotion Variation Detection

Figure 4: Architecture of the Sentence Emotion Recognition Model (SERM).

Retrieval of Definitions

The system retrieves the definition of the selected

words (adjectives, nouns (except proper nouns) and

verbs (except auxiliary verbs) taken from the original

sentence) from Wordnet. We then apply all the prior

steps to this definition (sentence): Remove

punctuation marks, word transformation, VANA

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval

detection, modifiers detection and word weight

update. The selected words from definition are then

added to database of selected words.

Emotion Keywords Detection

In this step, each one of the originally selected words,

as well as the selected words in the definitions, is

searched first in the ED, and if it not exists, searched

in the DAL.

If the word is in one of these two dictionaries, the

corresponding valence and arousal values will be

assigned to it.

If the word is not in any of the dictionaries, we

retrieve from Wordnet all of its synonyms and then

we search them on the ED and the DAL. If they are

in the dictionaries, we retrieve valence and arousal.

Emotion Calculator

At this point, the database of selected words contains

all the words found in the dictionaries. The

predominant emotion (valence and arousal) is then

calculated. The final emotion (valence and arousal) is

the weighted valence/arousal average of all the

selected words, taking into account the weight of each

word. The sentence is then classified in one quadrant

depending on the obtained valence and arousal

values.

4 RESULTS AND DISCUSSION

4.1 Discovering the Best Weights

To build our non-supervised model, we have to find

out the optimum values for the weights (WL1, WL2,

WL3, WL4 and WL5), which maximize the

performance (F-Measure) of the system, when this is

applied to new sentences.

To this end, we perform exhaustive tests with the

129 training sentences, combining different values for

the different weights in a specific range for each type

of weight.

First, we defined experimentally the allowed

range for each weight:

 WL1: between 10 e 1500.

 WL2: between 10 and 110.

 WL3: between 2 and 22.

 WL4: between 2 and 5

 WL5: between ½ and 1.

We then performed an iterative local search to

look for each optimum. We start with an initial large

granularity, which is decreased in the later iterations

until the possible minimum level, to find out the best

values for the 5 parameters. Illustrating, for WL1 in

the first iteration we went from 10 to 1500 in 50-unit

steps. Then, if the maximum performance were

achieved in the interval between 300 and 400, we

would test between 300 and 400 with 10-unit steps.

This was repeated until 1-unit granularity was

attained. We observed that our system has low

parameter sensitivity, as desired. In fact, the system

performance changed very slowly for different

parameters (see Table 8).

Tables 6 and 7 show respectively the best values

for each weight and the confusion matrix for these

parameters.

Table 6: Statistics for the best training model.

Weight Level Value

1 350

2 10

3 10

4 4

5 0.5

Table 7: Statistics for the best training model.

WL1 WL2 WL3 WL4 WL5

350 10 10 4 0.5

CM Q1 Q2 Q3 Q4

27 2 2 4

3 28 5 0

6 3 14 4

6 2 0 23

Precision Recall F-Measure

64.3% 77.1% 70.1%

80.0% 77.8% 78.9%

66.7% 51.2% 58.3%

74.2% 74.2% 74.2%

Average

71.3% 70.2%

70.3%

We can see that this combination of weights

achieved a performance of 70.38% (F-measure) in the

training set.

A possible cause for the lower results of quadrant

3 (13 sentences from quadrant 3 were incorrectly

classified in other quadrants) can be related to the fact

that this is a keyword-based approach. Quadrants 1, 2

and 4 are more influenced by keywords than quadrant

3, which is more influenced by ideas (e.g., he goes to

heaven), as discussed in another work by our team

(Malheiro et al., 2016).

We can see the comparison of results of the 10

best models in Table 8.

Keyword-based Approach for Lyrics Emotion Variation Detection

Table 8: Statistics for the best 10 training models.

1 2 3 4 5 Prec. Recall FM

350 10 10 4 0.5 71.29 70.24% 70.38%

250 10 10 6 1 70.38 69.55% 69.65%

450 10 2 2 0.5 69.77% 68.98% 69.14%

650 10 2 2 0.5 69.77% 68.98% 69.14%

450 10 2 4 1 69.77% 68.98% 69.14%

450 90 2 2 0.5 69.77% 68.98% 69.14%

550 10 2 2 0.5 69.77% 68.98% 69.14%

550 10 2 4 1 69.77% 68.98% 69.14%

650 10 2 4 1 69.77% 68.98% 69.14%

350 10 2 2 0.5 69.56% 68.98% 69.09%

4.2 Classification of Sentences

We applied SERM with the selected parameters to

our sentence validation dataset. The achieved results

are summarized in Table 9.

Table 9: Statistics for the validation model.

WL1 WL2 WL3 WL4 WL5

350 10 10 4 0.5

CM Q1 Q2 Q3 Q4

68 5 4 9

7 44 14 2

14 0 22 11

3 0 4 32

Precision Recall F-Measure

73.9% 79.1% 76.4%

89.8% 65.7% 75.9%

50.0% 46.8% 48.4%

59.3% 82.1% 68.8%

Average

68.2% 68.4%

67.4%

The average F-measure results (67.35%) are very

close to the results achieved in the training set

(70.82%).

In Table 9, we can also see the confusion matrix.

The validation dataset confirms the lower

performance of Q3 in comparison to the other

quadrants. This is shown by the amount of songs from

Q3 erroneously classified in other quadrants (recall is

48.35%) namely Q1 and Q4 (14 and 11 sentences

respectively). It is also shown by the amount of

sentences from Q2 (14) incorrectly classified in Q3.

This fact leads to a low precision for Q3 (50%). Q4

also has low precision (59.26%). This is due to the

sentences from Q1 and Q3 being erroneously

classified in Q4 (see example below).

At the end of section 3.1.1, we illustrated the

annotation results for 4 sentences of the dataset. Table

10 and the text below show the predicted classes for

these sentences and possible explanations for the

errors.

Table 10: Statistics for the validation model.

Sentences Actual Predicted

I’ve got peace like a river, I’ve

got peace like a river in my soul

Q4 Q4

Well now she’s gone, even

though I hold her tight, I lost my

love, my life, that night

Q3 Q1

At the end of all this hatred lies

even deeper hate, their

darkness has defeated you,

your lifeline running

backwards

Q2 Q2

You’re the light, you’re the

night, you’re the color of my

blood, you’re the cure, you’re

the pain, you’re the only thing I

wanna touch, never knew that

it could mean so much

Q1 Q2

Possible explanations for the wrong classify-

cations in the 2

and the 4

sentences are related to

the vocabulary used. In the 2

sentence, affective

words are almost absent. We can point out only the

word love, which is a word more related to Q1. This

confirms our conclusion that Q3 is more influenced

by ideas than keywords in comparison to the other

quadrants which are more influenced by the

keywords. We can see this typical behaviour in other

sentences like “Oh where, oh where, can my baby be?

The Lord took her away from me, she’s gone to

heaven, so I’ve got to be good so I can see my baby

when I leave this world” and “The stars are burning I

hear your voice in my mind, can’t you hear me

calling? My heart is yearning like the ocean that’s

running dry, catch me, I’m falling”, both of them have

essentially positive keywords (e.g., baby, heart,

ocean). The general idea conveyed by both sentences

is associated with Q3 (according to the annotators),

but our system classified them in Q1. An example

which explains the low recall from Q3 and low

precision from Q4 is the sentence “I lifted her head,

she looked at me and said – hold me darling just a

little while – I held her close, I kissed her our last kiss,

I found the love that I knew I had missed” from Q3

incorrectly classified in Q4. We can see the

predominance of words with positive valence, namely

kiss, darling, love, but the general idea for most

annotators was associated with Q3.

The 4

sentence belongs to Q1, but our system

classified it in Q2. This was probably due to the fact

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval

that the sentence uses antithesis and some of the

negative words are normally associated with Q2 (e.g.,

blood, pain).

Another example which can explain the amount

of sentences from Q2 erroneously classified in Q3

and consequently imply a low precision for Q3, is the

sentence “Shut up when I’m talking to you, shut up,

shut up, shut up, shut up when I’m talking to you, shut

up, shut up, shut up, I’m about to break”. This

sentence has a predominance of the word shut, and

our system has the limitation of not recognizing

phrasal verbs (e.g., shut up – more associated with

Q2) and the verb shut is associated with Q3,

according to DAL. We will address this issue in our

future work.

We cannot directly compare the results to other

works, because the datasets are different and ours is

only one composed by sentences from lyrics that we

are aware (the others are composed by other types of

text, such as children stories and less subjective text

such as journalistic text). Nevertheless the results

seem promising in comparison with approaches using

machine learning for complete song lyrics, e.g.,

73.6% F-measure in another work from our team

(Malheiro et al., 2016).

5 CONCLUSIONS

This research addresses the role of the lyrics in the

context of music emotion variation detection. To

accomplish this task we created a system to detect the

predominant emotion expressed by each sentence

(verse) of the lyrics, using a use a keyword-based

approach, which receives a sentence (verse) and

classifies it in the appropriate quadrant, according to

Russell’s emotion model. To validate our system, we

created a training set containing 129 verses and a

validation set with 239, annotated manually with an

average of 7 annotations per sentence. We attained

67.4% F-measure performance.

The main contributions of our work are the KBA

methodology proposed, as well as the ground-truth of

sentences created. In the future, we intend to improve

our methodology including the improvement of the

ED dictionary and a mechanism to detect beforehand

if the sentence is emotional or non-emotional.

Moreover, we intend to study emotion variation

detection along the lyric to understand the importance

of the different structures (e.g. chorus) along the lyric.

Additionally, we intend to make music emotion

variation detection in a bimodal scenario, including

audio and lyrics. This implies an audio-lyrics

alignment.

ACKNOWLEDGEMENTS

This work was supported by CISUC (Center for

Informatics and Systems of the University of

Coimbra).

REFERENCES

Agrawal, A., An, A., 2012. Unsupervised Emotion

Detection from Text using Semantic and Syntactic

Relations. In Proceedings of the 2012 IEEE/WIC/ACM

International Joint Conferences on Web Intelligence

and Intelligent Agent Technology, 346-353.

Aman, S., Szpakowicz, S. 2007. Identifying Expressions of

Emotion in Text. In Proceedings 10th International

Conference on Text, Speech and Dialogue TSD 2007,

Plzen, Czech Republic, Lecture Notes in Computer

Science 4629, Springer, pp. 196-205.

Besson, M., Faita, F., Peretz, I., Bonnel, A., Requin, J.

1998. Singing in the brain: Independence of lyrics and

tunes, Psychological Science, 9.

Binali, H., Wu, C., Potdar, V. 2010. Computational

Approaches for Emotion Detection in Text. 4th IEEE

International Conference on Digital Ecosystems and

Technologies.

Bradley, M., Lang, P. 1999. Affective Norms for English

Words (ANEW): Stimuli, Instruction Manual and

Affective Ratings. Technical report C-1, The Center for

Research in Psychophysiology, University of Florida.

Chopade, C. 2015. Text based Emotion Recognition.

International Journal of Science and Research (IJSR),

4(6), 409-414.

Chunling, M., Prendinger, H., Ishizuka, M. 2005. Emotion

Estimation and Reasoning Based on Affective Textual

Interaction. In Affective Computing and Intelligent

Interaction, Vol. 3784/2005: Springer Berlin /

Heidelberg, pp. 622-628.

Fontaine, J., Scherer, K., Soriano, C. 2013. Components of

Emotional Meaning. A Sourcebook. Oxford University

Press.

Hancock, J., Landrigan, C., Silver, C. 2007. Expressing

emotions in text-based communication. In Proceedings

of the SIGCHI conference on Human factors in

computing systems, pp. 929-932.

Hevner, K. 1936. Experimental studies of the elements of

expression in music. American Journal of Psychology,

48: 246-268.

Hu, Y., Chen, X., Yang, D. 2009. Lyric-Based Song

Emotion Detection with Affective Lexicon and Fuzzy

Clustering Method. Tenth Int. Society for Music

Information Retrieval Conference.

Hu, X., Downie, J. 2010. Improving mood classification in

music digital libraries by combining lyrics and audio.

Proc. Tenth Ann. joint conf. on Digital libraries, pp.

159-168.

Juslin, P., Laukka, P. 2004. Expression, Perception, and

Induction of Musical Emotions: A Review and a

Keyword-based Approach for Lyrics Emotion Variation Detection

Questionnaire Study of Everyday Listening. Journal of

New Music Research, 33 (3), 217–238.

Kao, E., Chun-Chieh, L., Ting-Hao, Y., Chang-Tai, H.,

Von-Wun, S. 2009. Towards Text-based Emotion

Detection. In International Conference on Information

Management and Engineering, pp. 70-74.

Korhonen, M., Clausi, D., Jernigan, M. 2006. Modeling

emotional content of music using system identification.

IEEE Transactions Systems Man Cyber, 36(3), 588-

599.

Krippendorff, K. 2004. Content Analysis: An Introduction

to its Methodology. 2nd edition, chapter 11. Sage,

Thousand Oaks, CA.

Landis, J., Koch, G. 1977. The measurement of observer

agreement for categorical data. Biometrics, 33:159–

174.

Laurier, C., Grivolla, J., Herrera, P. 2008. Multimodal

music mood classification using audio and lyrics. Proc.

of the Int. Conf. on Machine Learning and Applications.

Li, H., Pang, N., Guo, S. 2007. Research on Textual

Emotion Recognition Incorporating Personality Factor.

In International conference on Robotics and

Biomimetics, Sanya, China.

Lu, C., Hong, J-S., Cruz-Lara, S. 2006. Emotion Detection

in Textual Information by Semantic Role Labeling and

Web Mining Techniques. Third Taiwanese-French

Conf. on Information Technology.

Malheiro, R., Panda, R, Gomes, P., Paiva, R. 2016.

Emotionally-Relevant Features for Classification and

Regression of Music Lyrics. IEEE Transactions on

Journal Affective Computing, Vol 8.

Miller, G. 1995. WordNet: A Lexical Database for English

Communications of the ACM Vol. 38, No 11: 39-41.

Russell, J. 1980. A circumspect model of affect. Journal of

Psychology and Social Psychology, vol. 39, no. 6, p.

1161.

Russell, J. 2003. Core affect and the psychological

construction of emotion. Psychology Review, 110, 1,

145–172.

Schmidt, E., Turnbull, D., Kim, Y. 2010. Feature selection

for content-based, time-varying musical emotion

regression. In Proceedings of the ACM International

Conference on Multimedia Information Retrieval, 267-

274.

Schubert, E., 1999. Measurement and time series analysis

of emotion in music. Ph.D. dissertion, School of Music

Education, University of New South Wales, Sydney,

Australia.

Strapparava, C., Valitutti, A. 2004. Wordnet-affect: an

affective extension of wordnet. In Proceedings of the

4th International Conference on Language Resources

and Evaluation, pp. 1083-1086, Lisbon.

Taylor, A., Marcus, M., Santorini, B. 2003. The Penn

Treebank: an overview. Series Text, Speech and

Language Technology. Ch1. 20, 5-22.

Vignoli, F. 2004. Digital Music Interaction concepts: a user

study. Proc. of the 5th Int. Conference on Music

Information Retrieval.

Whissell, C., 1989. Dictionary of Affect in Language. In

Plutchik and Kellerman Emotion: Theory, Research

and Experience, vol 4, pp. 113–131, Academic Press,

NY.

Yang, Y-H., Liu, C., Chen, H. 2006. Music emotion

classification: A fuzzy approach. In Proceedings of the

ACM International Conference on Multimedia, 81-84.

Yang, C., Lin, K., Chen, H. 2007. Emotion Classification

Using Web Blog Corpora. In IEEE/WIC/ACM

International Conference on Web Intelligence.

Yang Y-H. Chen H. 2012. Machine recognition of music

emotion: a review. In ACM Transactions on Intelligent

Systems and Technology (TIST). Vol. 3, Issue 3.

KDIR 2016 - 8th International Conference on Knowledge Discovery and Information Retrieval