Lecture Three

by user

on 15 сентября 2016

Category: Documents

>> Downloads: 3

views

Report

Comments

Description

Download Lecture Three

Transcript

Lecture Three

Lecture Three
Stories of Complex Sociotechnical Systems:
Measurement, Mechanisms, and Meaning
Lipari Summer School, Summer, 2012
Complex
Sociotechnical
Systems
Complex
Sociotechnical
Systems
Press...
Measuring
Happiness
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
I
Analysis
Songs
Prof. Peter Dodds
Department of Mathematics & Statistics | Center for Complex Systems |
Vermont Advanced Computing Center | University of Vermont
Blogs
Tweets
“Social Scientists wade into the Tweet
stream” by Greg Miller,
Science, 333, 1814–1815, 2011 [15]
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
Positivity Bias
I
I
“Does a Nation’s Mood Lurk in Its Songs and
Blogs?” by Benedict Carey
New York Times, August 2009. ()
References
More here: http://www.uvm.edu/∼pdodds/research/ ()
Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License.
1 of 83
Complex
Sociotechnical
Systems
Outline
Measuring Happiness
Some motivation
Measuring emotional content
Data sets
4 of 83
Complex
Sociotechnical
Systems
Happiness:
Measuring
Happiness
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
Songs
Blogs
Blogs
Tweets
Analysis
Songs
Blogs
Tweets
Tweets
Positivity Bias
Positivity Bias
References
References
Positivity Bias
Socrates et al.:
eudaimonia [8]
References
Bentham:
hedonistic
calculus
Jefferson:
. . . the pursuit of
happiness
2 of 83
Papers and so on:
“Temporal patterns of happiness and
information in a global social network:
Hedonometrics and Twitter”
Dodds et al., PLoS ONE, 2011 [7]
Much better version here:
http://arxiv.org/abs/1101.5120 ()
Complex
Sociotechnical
Systems
Measuring
Happiness
“Positivity of the English Language”
Kloumann et al., PLoS ONE, 2012 [11]
I
“Measuring the Happiness of Large-Scale Written
Expression: Songs, Blogs, and Presidents”
Dodds and Danforth, Journal of Happiness Studies,
2009 [6]
I
language assessment by Mechanical Turk
(labMT 1.0)
I
http://www.onehappybird.com ()
Early drafts:
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
Songs
Blogs
Blogs
Tweets
I
6 of 83
Tweets
Positivity Bias
Positivity Bias
References
References
3 of 83
7 of 83
Desiring happiness—not just for boffins:
I
I
Average people routinely report being happy is what
they want most in life [12, 13, 5]
And it matters: “Happy people live longer:. . . ”
Survey by Diener and Chan. [5]
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
BREVIA
Emotional content
A Wandering Mind Is an
Unhappy Mind
Matthew A. Killingsworth* and Daniel T. Gilbert
So how does
one measure
U
2. levels of other emotional states?
nlike other animals, human beings spend
lot of time thinking about what is not
1. happiness?agoing
on around them, contemplating
Songs
Just ask people how happy they are.
Blogs
Tweets
Positivity Bias
I
Experience sampling
I
Day reconstruction
References
National indices of
well-being:
I
Bhutan
I
France
I
Australia
(Csikszentmihalyi et al.)
(Kahneman et al.)
But self-reporting has some drawbacks:
I
relies on memory and self-perception
I
induces misreporting
I
costly
8 of 83
Complex
Sociotechnical
Systems
An easy knock:
events that happened in the past, might happen
in the future, or will never happen at all. Indeed,
“stimulus-independent thought” or “mind wandering” appears to be the brain’s default mode
of operation (1–3). Although this ability is a remarkable evolutionary achievement that allows
people to learn, reason, and plan, it may have an
[2, 4,
emotional cost. Many philosophical
and3]
religious
traditions teach that happiness is to be found by
living in the moment, and[9]
practitioners are trained
to resist mind wandering and “to be here now.”
These traditions suggest that a wandering mind is
an unhappy mind. Are they right?
Laboratory experiments have revealed a great
deal about the cognitive and neural bases of mind
wandering (3–7), but little about its emotional
consequences in everyday life. The most reliable
method for investigating real-world emotion is experience sampling, which involves
contacting peo[14]
ple as they engage in their everyday activities and
asking them to report their thoughts, feelings, and
actions at that moment. Unfortunately, collecting
real-time reports from large numbers of people as
they go about their daily lives is so cumbersome
and expensive that experience sampling has rarely
been used to investigate the relationship between
mind wandering and happiness and has always
been limited to very small samples (8, 9).
We solved this problem by developing a Web
application for the iPhone (Apple Incorporated,
Cupertino, California), which we used to create
an unusually large database of real-time reports
of thoughts, feelings, and actions of a broad range
of people as they went about their daily activities. The application contacts participants through
their iPhones at random moments during their
waking hours, presents them with questions,
and records their answers to a database at www.
trackyourhappiness.org. The database currently
contains nearly a quarter of a million samples
from about 5000 people from 83 different countries who range in age from 18 to 88 and who
collectively represent every one of 86 major occupational categories.
To find out how often people’s minds wander,
what topics they wander to, and how those wanderings affect their happiness, we analyzed samples
from 2250 adults (58.8% male, 73.9% residing in
the United States, mean age of 34 years) who were
randomly assigned to answer a happiness question
(“How are you feeling right now?”) answered on a
continuous sliding scale from very bad (0) to very
good (100), an activity question (“What are you
doing right now?”) answered by endorsing one or
References and Notes
Happiness, attention, and doing:
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
Fig. 1. Mean happiness reported during each activity (top) and while mind wandering to unpleasant topics, neutral topics, pleasant topics or not
mind wandering (bottom). Dashed line indicates
mean of happiness across all samples. Bubble area
indicates the frequency of occurrence. The largest
bubble (“not mind wandering”) corresponds to
53.1% of the samples, and the smallest bubble
(“praying/worshipping/meditating”) corresponds to
0.1% of the samples.
1. M. E. Raichle et al., Proc. Natl. Acad. Sci. U.S.A
(2001).
2. K. Christoff, A. M. Gordon, J. Smallwood, R. Smith
J. W. Schooler, Proc. Natl. Acad. Sci. U.S.A. 10
12 of 83
(2009).
3. R. L. Buckner, J. R. Andrews-Hanna, D. L. Schac
Ann. N. Y. Acad. Sci. 1124, 1 (2008).
4. J. Smallwood, J. W. Schooler, Psychol. Bull. 132, 94
5. M. F. Mason et al., Science 315, 393 (2007).
6. J. Smallwood, E. Beach, J. W. Schooler, T. C. Ha
Complex
J. Cogn. Neurosci. 20, 458 (2008).
7.Sociotechnical
R. L. Buckner, D. C. Carroll, Trends Cogn. Sci. 11, 4
8. J. C. McVay, M. J. Kane, T. R. Kwapil, Psychon.
Systems
16, 857 (2009).
9. M. J. Kane et al., Psychol. Sci. 18, 614 (2007).
10. D. Kahneman, A. B. Krueger, D. A. Schkade, N.
A. A. Stone, Science 306, 1776 (2004).
11. A. B. Krueger, D. A. Schkade, J. Public Econ. 92, 183
Measuring
12. Materials and methods are available as support
Happiness
material on Science Online.
13. Some
J. Smallwood,
A. Fitzgerald, L. K. Miles, L. H. Ph
motivation
Emotion 9,emotional
271 (2009).
Measuring
14. content
We thank V. Pitiyanuvath for engineering www.
trackyourhappiness.org and R. Hackman, A. Jenkin
Data sets
W. Mendes, A. Oswald, and T. Wilson for helpful co
Supporting
Online Material
Analysis
www.sciencemag.org/cgi/content/full/330/6006/932/D
Songs
Materials and Methods
TableBlogs
S1
References
Tweets
18 May 2010; accepted 29 September 2010
Positivity Bias
10.1126/science.1192439
References
Harvard University, Cambridge, MA 02138, USA.
*To whom correspondence should be addressed
[email protected]
Killingsworth
and Gilbert,
VOL 330 SCIENCE
www.sciencemag.org
Science, 2010 [10]
932
Science = Orwell
more of 22 activities adapted from the day reconstruction method (10, 11), and a mind-wandering
question (“Are you thinking about something
other than what you’re currently doing?”) answered
with one of four options: no; yes, something pleasant; yes, something neutral; or yes, something unpleasant. Our analyses revealed three facts.
First, people’s minds wandered frequently, regardless of what they were doing. Mind wandering
occurred in 46.9% of the samples and in at least
30% of the samples taken during every activity
except making love. The frequency of mind wandering in our real-world sample was considerably
higher than is typically seen in laboratory experiments. Surprisingly, the nature of people’s activities had only a modest impact on whether their
minds wandered and had almost no impact on the
pleasantness of the topics to which their minds
wandered (12).
Second, multilevel regression revealed that people were less happy when their minds were wandering than when they were not [slope (b) = –8.79,
P < 0.001], and this was true during all activities,
Complex
Sociotechnical
including
the least enjoyable. Although p
minds
were more likely to wander to pleasan
Systems
(42.5% of samples) than to unpleasant
(26.5% of samples) or neutral topics (31%
ples), people were no happier when thinkin
Measuring
pleasant
topics than about their current activ
Happiness
–0.52,
not significant) and were considera
happier
thinking about neutral topi
Somewhen
motivation
–7.2,Measuring
P < 0.001)
or unpleasant topics (b =
emotional
P < content
0.001) than about their current activity
Data sets
bottom).
Although negative moods are
to cause mind wandering (13), time-lag a
Analysis
strongly
suggested that mind wandering
Songs
sample
was generally the cause, and not
the Blogs
consequence, of unhappiness (12).
Tweets
Third,
what people were thinking was
predictor of their happiness than was wh
Positivity Bias
were doing. The nature of people’s activi
plained
4.6% of the within-person variance
References
piness and 3.2% of the between-person vari
happiness, but mind wandering explained
of within-person variance in happiness and
of between-person variance in happiness. T
iance explained by mind wandering was
independent of the variance explained by
ture of activities, suggesting that the two w
dependent influences on happiness.
In conclusion, a human mind is a wan
mind, and a wandering mind is an unhappy
The ability to think about what is not hap
is a cognitive achievement that comes at a
tional cost.
12 NOVEMBER 2010
Policy = Brave New World
10 of 83
Complex
Sociotechnical
Systems
13 of 83
We’d like to build an ‘hedonometer’:
I
Measuring
Happiness
Some motivation
Measuring emotional
content
An instrument to ‘remotely-sense’
emotional states and levels, in real
time or post hoc.
Data sets
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Analysis
Songs
Songs
Ideally:
Blogs
Tweets
Blogs
Tweets
Positivity Bias
References
Positivity Bias
I
Transparent
I
Fast
I
Based on written
expression
I
Uses human evaluation
I
Non-reactive
I
Complementary to
self-reported measures
I
Improvable
References
Some possibilities:
I
See story here () for example [slate].
11 of 83
I
Natural language processing (e.g., OpinionFinder)
I
Declared mood levels in blogs (e.g., Livejournal) [16]
14 of 83
Complex
Sociotechnical
Systems
ANEW study
I
I
ANEW = “Affective Norms for English Words”
Study: participants shown lists of isolated words
Complex
Sociotechnical
Systems
Analysing text:
Measuring
Happiness
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
I
I
Asked to grade each word’s valence, arousal, and
dominance level
Integer scale of 1–9
I
N =1034 words—previously identified as bearing
emotional weight
I
Participants = College students (*cough*)
I
Results published by Bradley and Lang (1999) [1]
Blogs
Tweets
Positivity Bias
References
ANEW
words
Lyrics for
Michael Jackson’s Billie Jean
“She was more like a beauty queen
from a movie scene.
And mother always told me,
be careful who you love.
And be careful of what you do
’cause the lie becomes the truth.
Billie Jean is not my lover,
She’s just a girl who claims
that I am the one.
k=1. love
2. mother
3. baby
4. beauty
5. truth
6. people
7. strong
8. young
9. girl
10. movie
11. perfume
12. queen
13. name
14. lie
Songs
vk
8.72
8.39
8.22
7.82
7.80
7.33
7.11
6.89
6.87
6.86
6.76
6.44
5.55
2.79
fk
1
1
3
1
1
2
1
2
4
1
1
1
1
1
X
vtext = k
Blogs
vk fk
X
k
Tweets
Positivity Bias
fk
References
vBillie Jean = 7.1
vThriller = 6.3
vMichael = 6.4
Jackson
15 of 83
ANEW study—three 1–9 scales:
valence:
Complex
Sociotechnical
Systems
18 of 83
Complex
Sociotechnical
Systems
Data sets:
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Measuring
Happiness
Texts:
Some motivation
Measuring emotional
content
1. Song lyrics (1960–2007)
Data sets
Analysis
2. Song titles (1960–2008)
Songs
Songs
Blogs
arousal:
Blogs
3. State of the Union (SOTU) Addresses (1790–2008)
Tweets
Positivity Bias
References
dominance:
Tweets
Positivity Bias
References
Sources:
I
hotlyrics.com ()
I
freedb.com ()
I
American Presidency Project:
www.presidency.ucsb.edu ().
16 of 83
ANEW study words—examples
9
8
valence v
7
6
5
4
3
2
1
0
love/paradise/triumphant
glory/luxury/trophy
optimism/pancakes/church
Complex
Sociotechnical
Systems
Measuring
Happiness
derelict/neurotic/vanity
Complex
Sociotechnical
Systems
Data sets:
4. Blog phrases containing “I feel...”, “I am feeling”, etc.,
taken from wefeelfine.org () (API, 2005–2010)
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Thanks to ...
Songs
engine/paper/street
20 of 83
Blogs
Tweets
Positivity Bias
Kameron Harris
Songs
Blogs
Tweets
Isabel Kloumann
Catherine Bliss
References
Positivity Bias
References
fault/corrupt/lawsuit
trauma/hostage/disgusted
funeral/rape/suicide
50 100 150 200
frequency
I
ANEW = “Affective Norms for English Words” [1]
17 of 83
Jonathan Harris & Sep Kamvar
wefeelfine.org
Created by
Jonathan Harris
& Sep Kamvar
21 of 83
Complex
Sociotechnical
Systems
wefeelfine.org:
Song Lyrics—average happiness (valence)
Measuring
Happiness
Measuring
Happiness
6.8
Some motivation
Some motivation
6.7
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
mean valence vavg
Measuring emotional
content
Data sets
Complex
Sociotechnical
Systems
Measuring emotional
content
Data sets
6.6
Analysis
Songs
6.5
Blogs
Tweets
6.4
Positivity Bias
6.3
References
6.2
6.1
6
5.9
1960
1970
1980
1990
2000
2010
year
22 of 83
Complex
Sociotechnical
Systems
More data sets:
26 of 83
Song Lyrics—average happiness of genres:
Measuring
Happiness
Measuring
Happiness
Some motivation
Some motivation
7
Measuring emotional
content
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
6. New York Times (20 years)
7. Gutenberg.org
8. Google Books: http://ngrams.googlelabs.com/ ()
mean valence vavg
Data sets
5.
Complex
Sociotechnical
Systems
Analysis
6.5
Songs
Blogs
Tweets
6
Positivity Bias
Gospel/Soul (6.91)
Pop (6.69)
Reggae (6.40)
Rock (6.27)
Rap/Hip−Hop (6.01)
Punk (5.61)
Metal/Industrial (5.10)
5.5
5
9. . . .
4.5
1960
1970
References
1980
1990
2000
2010
year
23 of 83
Complex
Sociotechnical
Systems
Some numbers:
Counts
All words
ANEW words
Individuals
Song lyrics
58,610,849
3,477,575 (5.9%)
∼ 20,000
Song titles
60,867,223
5,612,708 (9.2%)
∼ 632,000
27 of 83
Measuring
Happiness
Per word drop in valence of lyrics from 1980−2007 relative to valence of lyrics from 1960−1979:Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Counts
All words
Tweets
Individuals
blogs
155,667,394
8,581,226 (5.5%)
∼ 2,335,000
SOTU
1,796,763
61,926 (3.5%)
43
Tweets
Positivity Bias
References
Twitter
∼ 100 billion
∼ 10 billion
∼ 100 million
Word number i
Counts
All words
ANEW words
Individuals
Complex
Sociotechnical
Systems
Happiness Word Shift Graph:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
hate ↑
pain ↑
baby ↓
death ↑
dead ↑
home ↓
sick ↑
fear ↑
hit ↑
hell ↑
fall ↑
sin ↑
lost ↑
sad ↓
burn ↑
lie ↑
scared ↑
afraid ↑
music ↓
life ↑
god ↑
trouble ↓
loneliness ↓
−20
−10
Some motivation
love ↓
lonely ↓
0
Key:
Measuring emotional
content
Data sets
love ↓
baby ↓
home ↓
music ↓
good ↓
Decreases in relatively
high valence words
contribute to drop
in average valence
hate ↑
pain ↑
death ↑
dead ↑
sick ↑
Increases in relatively
low valence words
contribute to drop
in average valence
life ↑
god ↑
truth ↑
party ↑
sex ↑
Increases in relatively
high valence words
contribute to increase
in average valence
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
lonely ↓
sad ↓
trouble ↓
loneliness ↓
devil ↓
Decreases in relatively
low valence words
contribute to increase
in average valence
10
Per word valence shift ∆i
24 of 83
28 of 83
Complex
Sociotechnical
Systems
Top 16 of ' 20,000 artists:
Artist
All-4-One
Luther Vandross
S Club 7
K Ci & JoJo
Perry Como
Diana Ross & The Supremes
Buddy Holly
Faith Evans
The Beach Boys
Jon B
Dru Hill
Earth Wind & Fire
Ashanti
Otis Redding
Faith Hill
NSync
Valence
7.15
7.12
7.05
7.04
7.04
7.03
7.02
7.01
7.01
6.98
6.96
6.95
6.95
6.93
6.93
6.93
Measuring
Happiness
Some motivation
Measuring emotional
content
Songs
Blogs
Tweets
Positivity Bias
6.4
State of the Union
Messages [6]
New York Times
(1987–2007) [17]
Blogs [6]
Dante’s Inferno
6.1
church (6.28), tree (6.32), air (6.34)
clouds (6.18), alert (6.20), computer
(6.24)
grass (6.12), idol (6.12), bottle (6.15)
6.0
hotel (6.00), tennis (6.02), wonder (6.03)
5.8
5.5
Heavy Metal
lyrics [6]
5.4
owl (5.80), whistle (5.81), humble (5.86)
glacier (5.50), repentant (5.53), mischief
(5.57)
lamp (5.41), elevator (5.44), truck (5.47)
6.3
6.2
References
29 of 83
Complex
Sociotechnical
Systems
Bottom 16 of ' 20,000 artists:
Artist
Slayer
Misfits
Staind
Slipknot
Darkthrone
Death
Black Label Society
Pig
Voivod
Fear Factory
Iced Earth
Simple Plan
Machine Head
Metallica
Dimmu Borgir
Mudvayne
Tweets, 9/9/2008
to 12/31/2010
Rock lyrics [6]
Enron Emails ()
6.7
6.5
Data sets
Analysis
(criteria: ≥ 50 songs and ≥ 1000 ANEW words)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
havg
6.9
Valence
4.80
4.88
4.93
4.98
4.98
5.02
5.05
5.08
5.14
5.15
5.16
5.16
5.17
5.19
5.20
5.21
Words with a similar score:
chocolate (6.88), leisurely (6.88),
penthouse (6.81)
dream (6.73), honey (6.73), sugar (6.74)
muffin (6.57), rabbit (6.57), smooth
(6.58)
thought (6.39), face (6.39), blond (6.42)
Complex
Sociotechnical
Systems
Blogs
Measuring
Happiness
Measuring
Happiness
6.1
Some motivation
valence (v)
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Text:
Soul/Gospel
lyrics [6]
Pop lyrics [6]
Dante’s Paradise
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
Some motivation
6
Measuring emotional
content
Data sets
5.9
Analysis
5.8
Songs
Blogs
Tweets
5.7
Positivity Bias
5.6
5.5
References
13 20 30 40 50 60 70 80
blogger age
I
Average happiness as a function of the age bloggers
report they will turn in the year of their posting.
(criteria: ≥ 50 songs and ≥ 1000 ANEW words)
30 of 83
34 of 83
Tref : born in 1960-1969 (havg =5.96)
Tcomp: 14 years old (havg =5.55)
Blogs—Overall trend
1 −↑sick
−↑hate
−↑stupid
−↑sad
5
6.4
US Inauguration 1/20
♥
♥
6.1
♥
6
5.9
♥
5.8
5.7
♥
10
0
10
9/10
9/11
Data sets
Songs
Blogs
Text size:
Tref Tcomp
loved +↑
1
10
fun +↑
20
−↑dead
−↑scared
−↑terrible
2
10
2008
2009
Tweets
Positivity Bias
References
Balance:
−169 : +69
+↓
+↑
friend +↑
people +↑
3
2007
Measuring emotional
content
Analysis
A S OND J FMAM J J A S OND J FMAM J J A S OND J FMAM J J A S OND J FMAM J J A S OND J FMA
2005 2006
Some motivation
happy +↑
love +↑
−↑upset
−↑fat
15
Michael
Jackson
9/10
9/10
Word rank r
average happiness havg
6.3
6.2
Measuring
Happiness
−↑depressed
−↑bored
−↑lonely
−↑alone
−↑mad
−↑pain
+↓life
US Election 11/4
Complex
Sociotechnical
Systems
10
−100
2010
25
Pr
i=1
−20
−↑confused
time −↓
0
δhavg,i
−↑hurt
−10
0
10
−↑
−↓
20
Per word average happiness shift δhavg,r (%)
35 of 83
Tref : Male (havg =5.91)
Tcomp: Female (havg =5.89)
Complex
Sociotechnical
Systems
love +↑
1
−↑hurt
−↑hate
−↑sad
+↓good
−↑alone
5
Measuring
Happiness
Some motivation
Measuring emotional
content
Word rank r
baby +↑
loved +↑
happy +↑
−↑stupid
−↑guilty
−↑sick
10
0
heart +↑
−↑scared
−↑lost
+↓music
+↓free
1
10
20
Analysis
Songs
Blogs
10
15
Data sets
Text size:
Tref Tcomp
Tweets
Positivity Bias
References
Balance:
death −↓ −607 : +507
life +↑
+↓
+↑
family +↑
2
10
+↓christmas
cold −↓
−↑upset
3
10
25
−100
P
r
i=1
friend +↑
dead −↓
0
δhavg,i
−100
−50
0
50
−↑
−↓
100
Per word average happiness shift δhavg,r (%)
Complex
Sociotechnical
Systems
Twitter—living in the now:
Measuring
Happiness
0.16
count fraction
36 of 83
breakfast
0.14
lunch
0.12
dinner
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
0.1
Tweets
Positivity Bias
0.08
References
0.06
0.04
0.02
0
0
2
4
6
8
10 12 14 16 18 20 22 24
hour of day (local time)
38 of 83
Twitter—living in the now:
Complex
Sociotechnical
Systems
Measuring
Happiness
0.07
Some motivation
Word
stomach
mcdonalds
hungry
wings
ham
starving
spaghetti
ihop
noodles
ketchup
fat
sprite
cookin
heartburn
sugar
kool-aid
miller
honey
candy
havg
5.40
5.98
3.38
6.52
5.66
2.58
0.00
0.00
0.00
0.00
3.24
0.00
0.00
0.00
6.74
0.00
5.36
7.44
7.52
rs
0.37
0.30
0.27
0.25
0.24
0.22
0.20
0.19
0.18
0.18
0.18
0.17
0.17
0.17
0.15
0.15
0.15
0.15
0.15
p-value
1.98894e-07
2.60824e-05
0.000206297
0.000388915
0.000763101
0.00272286
0.00689403
0.0100034
0.0106139
0.0145088
0.0148845
0.0175705
0.0182976
0.0200551
0.0329359
0.0354226
0.036325
0.0395531
0.0398618
Words most anti-correlated with obesity
levels in cities:
brunch
bar
banana
barista
delicious
dinner
coffee
espresso
cocktails
booze
mimosa
spiced
veggie
sushi
wines
tofu
panini
gnocchi
clams
caffeine
cocktailin
Twitter—living
bento
huevos
mojitos
vegan
6.32 -0.41
5.82 -0.35
6.86 -0.35
0.00 -0.35
7.92 -0.34
7.40 -0.34
7.18 -0.34
0.00 -0.33
0.00 -0.32
0.00 -0.32
0.00 -0.31
0.00 -0.31
0.00 -0.31
5.40 -0.31
6.28 -0.31
0.00 -0.31
0.00 -0.31
0.00 -0.30
0.00 -0.30
5.80 -0.30
-0.30
0.00 now:
the
0.00 -0.30
0.00 -0.30
0.00 -0.30
4.82 -0.30
6.37431e-09
5.54374e-07
5.67492e-07
7.29324e-07
1.09807e-06
1.35413e-06
2.04145e-06
4.45903e-06
4.96518e-06
6.38461e-06
1.24472e-05
1.52074e-05
1.60439e-05
1.71997e-05
1.7432e-05
1.86278e-05
1.86719e-05
2.51419e-05
2.52124e-05
2.6263e-05
2.63227e-05
2.6349e-05
3.19903e-05
3.52542e-05
3.5502e-05
Measuring emotional
content
0.06
count fraction
Words most correlated with obesity levels in
cities:
0.05
0.03
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
40 of 83
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
41 of 83
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
Songs
Blogs
Blogs
Tweets
0.04
Complex
Sociotechnical
Systems
Tweets
Positivity Bias
Positivity Bias
References
References
hungry
starving
0.02
food
0.01
0
0
eat
2
4
6
8
10 12 14 16 18 20 22 24
Tweeting the Superbowl () [NY Times]
hour of day (local time)
39 of 83
42 of 83
a
hah
6
aha
hah
ha
10
a
hah
aha
Frequency
hah
hah
aha
aha
4
10
ah
hah
ha
aha
hah
a
hah
2
10
0
10
2
4
6
10
26
hahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahahaha
The happiest distribution:
100 140
Number of Letters (String Length)
labMT 1.0:
language assessment by Mechanical Turk
http:/www.onehappybird.com ()
http://flowingdata.com/2011/10/27/language-communities-of-twitter/ ()
valence
rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
.
.
.
word
valence
std dev
twitter
rank
g-books
rank
nyt
rank
lyrics
rank
laughter
happiness
love
happy
laughed
laugh
laughing
excellent
laughs
joy
successful
win
rainbow
smile
won
pleasure
smiled
rainbows
winning
celebration
enjoyed
healthy
music
celebrating
congratulations
weekend
celebrate
comedy
jokes
rich
.
.
.
8.50
8.44
8.42
8.30
8.26
8.22
8.20
8.18
8.18
8.16
8.16
8.12
8.10
8.10
8.10
8.08
8.08
8.06
8.04
8.02
8.02
8.02
8.02
8.00
8.00
8.00
7.98
7.98
7.98
7.98
.
.
.
0.93
0.97
1.11
0.99
1.16
1.37
1.11
1.10
1.16
1.06
1.08
1.08
0.99
1.02
1.22
0.97
1.07
1.36
1.05
1.53
1.53
1.06
1.12
1.14
1.63
1.29
1.15
1.15
0.98
1.32
.
.
.
3600
1853
25
65
3334
1002
1579
1496
3554
988
2176
154
2726
925
810
1497
–
–
1876
3306
1530
1393
132
2550
2246
317
1606
1444
2812
1625
.
.
.
–
2458
317
1372
3542
3998
–
1756
–
2336
1198
3031
–
2666
1167
1526
3537
–
–
–
2908
3200
875
–
–
–
–
–
–
1221
.
.
.
–
–
328
1313
–
4488
–
3155
–
2723
1565
776
–
2898
439
4253
–
–
1426
2762
3502
3292
167
–
–
833
3574
2566
–
1469
.
.
.
1728
1230
23
375
2332
647
1122
–
2856
809
–
694
1723
349
1493
1398
2248
4216
3646
4070
–
4619
374
–
–
2256
2108
–
3808
890
.
.
.
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
48 of 83
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
.
.
.
g-books
rank
nyt
rank
lyrics
rank
.
.
.
violence
cruel
cry
failed
sickness
abused
tortured
fatal
killings
murdered
war
kills
jail
terror
die
killing
arrested
deaths
raped
torture
died
kill
killed
cancer
death
murder
terrorism
rape
suicide
terrorist
.
.
.
1.86
1.84
1.84
1.84
1.84
1.83
1.82
1.80
1.80
1.80
1.80
1.78
1.76
1.76
1.74
1.70
1.64
1.64
1.64
1.58
1.56
1.56
1.56
1.54
1.54
1.48
1.48
1.44
1.30
1.30
.
.
.
1.05
1.15
1.28
1.00
1.18
1.31
1.42
1.53
1.54
1.63
1.41
1.23
1.02
1.00
1.19
1.36
1.01
1.14
1.43
1.05
1.20
1.05
1.23
1.07
1.28
1.01
0.91
0.79
0.84
0.91
.
.
.
4299
2963
1028
2645
4735
–
–
–
–
–
468
2459
1642
4625
418
1507
2435
–
–
3175
1223
798
1137
946
509
2762
–
3133
2124
3576
.
.
.
1724
–
3075
1618
–
–
–
4089
–
–
175
–
–
4117
730
4428
4474
–
–
–
866
2727
1603
1884
307
3110
–
–
4707
–
.
.
.
1238
–
–
1276
–
–
–
–
4914
–
291
–
2573
4048
2605
1672
1435
2974
–
–
208
2572
814
796
373
1541
3192
4115
3319
3026
.
.
.
2016
1447
226
2920
3782
4589
4693
3724
–
4796
462
2857
1619
2370
143
998
–
–
4528
3126
826
430
1273
3802
433
1059
–
2977
2107
–
Complex
Sociotechnical
Systems
Twitter—overall time series:
6.4
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
average happiness havg
twitter
rank
2008—12/25 2009—
A
References
valence
std dev
twitter
rank
g-books
rank
nyt
rank
lyrics
rank
fE@king
fKKkin
fKKked
pussy
whiskey
slut
cigarettes
fKKk
mortality
cigarette
motherfKKkers
churches
motherfKKking
capitalism
porn
summer
beer
execution
wines
zombies
aids
capitalist
revenge
mcdonalds
beatles
islam
pay
alcohol
muthafKKkin
christ
.
.
.
4.64
3.86
3.56
4.80
5.72
3.57
3.31
4.14
4.38
3.09
2.51
5.70
2.64
5.16
4.18
6.40
5.92
3.10
6.28
4.00
4.28
4.84
3.71
5.98
6.44
4.68
5.30
5.20
3.00
6.16
.
.
.
2.93
2.74
2.71
2.66
2.64
2.63
2.60
2.58
2.55
2.52
2.47
2.46
2.46
2.45
2.43
2.39
2.39
2.39
2.37
2.37
2.35
2.34
2.34
2.33
2.33
2.33
2.32
2.32
2.31
2.31
.
.
.
448
1077
1840
2019
–
–
–
322
–
–
–
–
–
–
1801
896
839
–
–
4708
2983
–
–
3831
3797
–
627
2787
–
2509
.
.
.
–
–
–
–
–
–
–
–
3960
–
–
2281
–
4648
–
1226
4924
2975
–
–
3996
4694
–
–
–
4514
769
2617
–
909
.
.
.
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
721
3960
–
3316
–
1197
–
–
–
–
–
460
3752
–
4238
.
.
.
620
688
904
949
2208
4071
3279
185
–
2678
1466
–
2910
–
–
590
1413
–
–
–
–
–
2766
–
–
–
499
3600
4107
1526
.
.
.
Complex
Sociotechnical
Systems
Data sets
Analysis
Songs
Blogs
Tweets
09/14
02/27
08/06
Positivity Bias
03/11
09/1109/21
07/23 08/0808/23
700
B
600
500
400
300
Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
4
C
3
2
1
Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Measuring societal happiness through tweets:
6.4
2008—
2009—
12/25
12/25
12/24
01/01
04/12
11/27
C
0.8
1000
0.6
2
600
0.4
0.2
1
400
0
0.5
200
2
3
4
5
havg
6
7
8
9
0
0
D
1
2
3
∆havg
1.5
−0.2
1
2
∆havg
3
% coverage
∆havg
800
12/25
Saturday
11/25
12/24
Sunday
04/29 05/08
11/24
06/19
04/24
12/31
10/31
06/25
04/27
09/29
09/14
02/27
08/06
05/24
06/27
03/11
07/05
07/23
5.8
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
08/08
Jul
08/23 09/11
Aug
Sep
09/21
Oct
Nov
Dec
date
I
Global happiness spikes = predictable rituals.
I
Global sadness spikes = unpredictable, exogeneous
shocks.
I
No accidental happiness outbreaks.
A
Bailout of the U.S. financial system:
Tref : 7 days before and after (havg =6.00)
Tcomp: Monday, 2008/09/29 (havg =5.95)
Royal Wedding of Prince William & Catherine Middleton:
T : 7 days before and after (havg =5.98)
B Tref
comp: Friday, 2011/04/29 (havg =6.04)
−↑no
−↑fail
−↑fails
−↑blame
+↓love
−↑failure
−↑bad
−↑don’t
−↑against
−↑die
−↑rejected
−↑depression
−↑crisis
15
money +↑
weekend +↑
25
+↓won
old −↓
−↑worst
+↓great
−↑panic
+↓awesome
−↑didn’t
+↓google
+↓saturday
2
billion +↑
∆havg
cancer −↓
10
40
−↑sick
−↑problem
−↑crash
+↓friday
−↑falling
1
10
2
10
45
house +↑
10
15000
20000
50
Pr
i=1
−20
married +↑
+↓game
35
Text size:
Tref Tcomp
0
10
Balance:
−165 : +65
+↓
+↑
0
δhavg,i
−10
10
2
+↓lol
10
−↓
20
Per word average happiness shift δhavg,r (%)
50
live +↑
bad −↓
congrats +↑
amazing +↑
kill −↓
ill −↓
nigga −↓
Text size:
wow +↑
Tref Tcomp
died −↓
30
ass −↓
hell −↓
gorgeous +↑ Balance:
congratulations
+↑
−68 : +168
40
couple +↑
3
10
celebrating +↑
ill −↓
−↑hell
celebrate +↑
+↓mom
−↑loss
−↑mourn
35
america +↑
+↓you
−↑deaths
0
10
+↓
+↑
0
P
r
δhavg,i
−30
−20
i=1
kissed +↑
she +↑
killing −↓
100
+↓haha
−10
0
10
10
+↓beautiful
peace +↑
2
+↓fun
−↑evil
+↓good
−↑attack
3
10
20
−↓
30
Per word average happiness shift δhavg,r (%)
10
−100
50
Pr
i=1
−20
Balance:
−143 : +43
+↓
+↑
we +↑
4
−↑
Text size:
Tref Tcomp
1st +↑
sea +↑
usa +↑
1
10
45
+↓friends
4
−↑
0
1
10
−↑gossip
−↑poor
25
+↓life
10
45
miss −↓
4
10
−100
10000
word rank
+↓chocolate
+↓love
+↓win
25
40
+↓home
−↑killed
−↑fear
3
no −↓
−↑hussein
+↓hahaha
−↑terror
+↓like
+↓wedding
−↑terrorism
−↑enemy
+↓lol
−↑shot
20
weekend +↑
friday +↑
party +↑
30
0
3
15
+↓me
+↓good
+↓you
20
+↓sunday
+↓fun
+↓party
+↓game
−↑hurt
+↓win
1
10
no −↓
princess +↑
never −↓
shit −↓
killed −↓
not −↓
dress +↑
real +↑
15
20
−↑killed
+↓love
−↑not
−↑kill
−↑died
−↑killing
+↓happy
−↑saddam
+↓haha
+↓me
−↑war
+↓mothers
−↑terrorist
−↑bad
−↑pakistan
−↑buried
5
+↓easter
+↓happy
10
35
0
dead −↓
dont −↓
death −↓
beautiful +↑
hate −↓
kiss +↑
prince +↑
5
Death of Osama Bin Laden:
Tref : 7 days before and after (havg =5.98)
Tcomp: Monday, 2011/05/02 (havg =5.89)
C
1 −↑dead
−↑death
wedding +↑
1
−↑bill
−↑down
−↑failed
−↑not
10
F
5000
02/14
01/01
06/20
6
5.9
30
E
100
80
60
40
20
0
0
Wednesday
Thursday
07/04
04/04
10/31
Word rank r
01/01/11
100
80
60
40
20
0
% coverage
2.5
Monday
Tuesday
12/31
Friday
05/09
01/01
10/31
10/31
6.1
Word rank r
10/01/10
# words
2∆havg
2011—
12/25
12/24
12/31
02/14
07/04
06/21
12/31
last −↓
Date
B
12/24
11/26
02/14
6.2
5
5.4
07/01/10
2010—
6.3
50 of 83
6
10000
8000
6000
4000
2000
0
0
07/05
References
5.8
1
05/24 06/27
Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1
Average happiness havg
06/25
04/27
09/29
05/02
6.2
04/01/10
11/24
06/19
Twitter—overall time series:
Measuring emotional
content
6.4
01/01/10
12/31
12/24
04/2905/08
Some motivation
6.6
10/01/09
02/14
01/01
11/25
6
5.9
1−↑bailout
3
07/04
06/20
Measuring
Happiness
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
07/01/09
05/09
04/04
04/24
∆havg :
04/01/09
02/14
01/01
10/31
49 of 83
5.6
Frequency
06/21 07/04
12/31
12/31
date
6.8
0
1
02/14
11/27
0
word
01/01/09
Monday
Tuesday
Wednesday
Thursday
12/25
Friday
Saturday
Sunday
2011—
12/25
12/24
05/02
Positivity Bias
A
10/01/08
2010—
12/24 12/31
11/26
04/12
10/31
6.1
5.8
7.2
1200
12/24 01/01
6.2
The very surprising tunable hedonometer:
7
12/25
6.3
Tweets
Simpson lexical size N S
std dev
Word rank r
std dev
rank
valence
word count (x10 7 )
.
.
.
10193
10194
10195
10196
10197
10198
10199
10200
10201
10202
10203
10204
10205
10206
10207
10208
10209
10210
10211
10212
10213
10214
10215
10216
10217
10218
10219
10220
10221
10222
word
average happiness havg
valence
rank
0
δhavg,i
−↑breaking
−↑burial
−10
−↑
0
10
−↓
20
Per word average happiness shift δhavg,r (%)
Some motivation
Measuring emotional
content
havg
Data sets
Analysis
6.35
Songs
Tweets
F
S
S
M
T
W
T
F
S
S
Positivity Bias
M
day of week
References
8
7
havg
5
likelove
all
new
we
will
da
goo
day
know
tod
more
hah
nt
ay
see
think
make
people
vid
she
th
our
eo
right
well
us
an
be
please
che
hahaha
eat
stgrks
ck
ho
free
me
show
live
life
hey
happy
firs
bet
thop
ter
school
wor
blog
ver
ld
nigh
big
y giv
help
always
t yes
emu
morn
find
rea
nic
girl
tha
try
ing
lcool
e
sure
ha
face
lun
egame
youtube
sic
online
most
book
ever
ch
go
amo
bab
done
start
yone
son
dynk
iphone
r top
phon
house
gwin
wismoney
awe
guys
tonight
hehe
etv
wow
h
frien
som
tomorr
ready
super
ma
soon
read
talk
wit
thoug
eds
ow
feliz
play
wee
buy
inte
am
son
birthday
kend
pretty
socia
rnet
azifun
l hthi true
food
cacon
nt
last
illbad
sh
it never
down
die hate
t
wait
sdidn
stop
damn mis
ass
old
sorry
mea
nothing
doesn
n t
hell
nigga
bitch
without
lostwrong
waitingisnt
gone
late
sickd sad
mad
least
dea
dtired
sin bore
stupid
ugh
forget
wasnt
problem
hungry
against
cutmissed
niggas
fire bye
killwar hurt fight
alone
wouldnt
fat
lose sucksforgot
bill
low
couldn
shut
tfuckin
death
boo
#fail
killedr
cance
died
own
dark
brokeugly
shout
court
crap leaving
taxdumblazy
stress
kick
problems
hang
worse
annoyin
jealous
scared
broken
hurts
ends
g dirty
breaking
mess
issues
fool caught
falling
drunk
fell
closedhomework
attack trouble smoke
fucked
fear
killingevil horrible
quit
zero
sore doubt lame
zit bug
waste
nothin none
angrypissed
blame
nasty
humidity
shame
afraid
confused
error
force
tough
spam
drug monster
crying
fighting
apart
dying
losing
stopped
terrible
battle
riskcrack
burn
bomb
scary
headache
lies
dope
pressure
rob
bullshit
dies
gun
shootin
twit
tears
rude
goodbye
crash
debt
g weak
smoking
avoid shots
hanging
gang
bloody
ow
dropped
haters
flu
jailwarshates worried
killer
nooo
difficult
faultgossip
lonely idiot
upset
lack
ridiculous
falls
ban
rent
delete offline
trial
crime
crisis
mistake
accident
lying
surgery
nervous
pisscuts
arrested
oops
expensive
impossible
failed
empty
mistakes
horrorignore
warning
endedloose
addicted
blind
strike
devil
insane
drugsbills
dentist awkward
murder kills failure
fired
dangerous
argh
hitting
grossmortgage
negative
freezing
ouch
sosalarm
bite
awful complain
steal guiltyfever
burning
beast
loser
−4
hugs
winning
earn
happiness
deliciou
angel
brilliant
hero
sjoy
successful
indeed
england
offers
thanx
speed
florida
aha
feed
career
mum
campus
xoxo
value
kitchen
large
queen
animal
skype
boyfriend
nights
tiger
dollar
france
faithmoon
ityhoney
helps
celebrity
deals
e−mail
daughte
jeans
truly
australia
rise
xp halloween
vip
bigger
worlds
tip
milk
father
male
samsung
simply
learned
rgolden
soup
phones
hollywood
deserve
keeping
buddy
fitness
professional
sandwich
yummy
yum
pie
launches
exclusive
sport
hug
easier
salad
basketball
sharing
indian
december
modern
guest
wave
surprise
trading
designer
cousin
likely
growing
burger
oscar
plane
cooking
created
lake
bread
unlocked
mario
coast
certain
rose
increase
bird
husband
leader
silver
planet
picked
europe
halo
anniversary
ultimate
flying
language
knowing
memory
blessed
score
higher
agreed
heaven
host
beginning
instant
fixed
favourite
disco
lights
walked
promote
talent
raise
nature
honor
accept
entertainme
nation
prince
bella
discount
icon
surprised
girlfriend
freedom
photoshop
bbq
incredible
sunshin
sugar
soccer
candy
purple
kidding
legs
spirit
river
#music
japanese
receive
holidays
wifi
arrived
amen
faster
enjoyed
valentines
billion
acting
pure
nt e
artists
engineer
pregnant
slept
honestly
romance
received
fantasy
express
musical
babies
aid
juice
prize
prefer
invite
networking
teams
rice
solar
spain
asian
cats
interest
females
upcoming
ups
honest
winners
lyrics
rights
boots
plaza
smiles
breath
vista
mp3
numbers
meat
clearly
purchase
xmas
grown
wireless
mmmm
playin
jazz
plays
woods
conversation
gods
photography
mountain
relax
bright
package
lips
marriage
cookies
dates
dressed
turkey
drinks
income
adult
wishes
menu
hiring
solution
safety
meal
partner
stores
customers
bath
presents
youth
americans
kisses
calm
inspiratio
paint
union
valley
cheers
allow
sisters
stream
easily
faces
wings
justice
fruit
productive
exercise
dollars
donate
unique
benefits
arts
suggestions
skills
vintage
draw
comedy
boat
singer
leads
records
improve
memories
nhelped
teachers
amsterdam
bringing
strategy
leading
gallery
genius
emails
yoga
horse
chips
growth
survey
coupon
images
pleasure
inspired
fancy
cares
premiere
hotels
journey
voting
sexual
gifts
fabulous
soft
mommy
charity
published
correct
goals
answers
upgrade
recipe
powerful
riding
discover
delivery
t−shirt
jokes
fully
taco
expert
chief
wise
premier
potential
rare
hearts
sweetie
fave
liverpool
fab
saved
italy
built
birds
easter
useful
production
chances
banana
comic
midnight
adorable
chef
perform
relief
launched
bonus
teaching
diamond
spa
uncle
options
knowledge
recovery
marry
saints
culture
legend
moral
museum
owner
progress
willing
twin
newest
animals
princess
highly
butter
models
meaning
creating
goodness
jacket
effective
protect
italian
bedroom
provide
eggs
debut
shine
foods
profit
piano
passion
massage
casino
saving
goodmorning
finale
friendly
moments
rising
shares
cure
nike
lovers
benefit
strength
robot
universal
lounge
vision
cookie
mirror
suite
bff
boost
wishing
attend
solutions
desire
facts
minds
opened
offering
farm
fund
fri
result
b−day victory
films
fiesta
tuned
sauce
parties
purpose
amore
signing
premium
websites
confirmed
millions
puppy
swimming
rofl
performing
resources
christ
prayer
cheer
pasta
plenty
lover
pride
wood
birth
mothers
cruise
homes
reached
woohoo
fest
praying
forgive
breathe
parade
slim
shirts
balance
ocean
grace
smiling
colors
kindle
tennis
hills
universe
pub
actor
firefox
curious
flowers
organic
bands
independent
toys
writers
thanksgiving
breast
talented
drawing
thousands
degree
discovered
grandma
activity
trees
asia
theater
bible
relationships
grill
fame
wisdom
village
blonde
plant
coke
racing
prime
painting
cds
reader
adventure
zoo
hire
pal
celebrating
beatles
active
thai
gmail
toast
virgin
surely
awareness
largest
laughs
kitty
momma
grad
couch
leadership
invited
designs
completed
keyboard
religion
certainly
lovin
theatre
swim
advance
clothing
hopes
ideal
costume
worldwide
berry
prayers
monkey
votes
snowing
hubby
crystal
bass
jewelry
opportunitie
swing
relaxing
holland
dads
darling
computers
supporting
fries
babes
captain
gaming
survive
precious
cultural
afford
angels
bucks
visiting
thankful
decent
listened
australian
rescue
patience
pays
sfunniestcelebratio
perfume
supreme
raised
soundtrack
beans
downloads
shore
ability
clue
flag
unlimited
browser
younger
virginia
personality
understanding
advantage
gr8
motivation
programs
markets
novel
academy
familiar
prepared
resort
nurse
replied
essential
luxury glory
fishing
impressed
playlist
western
joining
raises
doll
designed
revealed
recipes
albums
superior
tribute
highest
hawaii
heroes
grant
newspaper
n
acededicated
steak
greek
rockin
families
flower
rainbow
bubble
partners
sang
exact
liking
downloaded
grocery
designers
dime
cherry
colour
vehicle
maps
haircut
discovery
greater
stronger
grammy
remembered
shared
housing
lifetime
champions
friendship
pearl
twins
karma
produced
basket
tops
duck
arrive
actress
rocking
comfortable
grateful
forest
herself
tournament
batman
praise
waves
expecting
confidence
promotion
cousins
aunt
advanced
romantic
goodnight
funds
80s
stunning
dvds
focused
hip
downloading
experts
television
clouds
enterprise
communication
smoothreunion
offered
brazilian
included
deliver
furniture
interactive
studies
legit
sight
americas
prizes
cameras
roses
encore
olympics
finishing
headphones
newly
kingdom
electronic
dreaming
wonderland
graduate
laughed
properly
helpful
weekends
queens
experienced
pleased
olympic
shops
rings
peanut
contacts
castle
champion
coat
soda
strawberry
choices
steady
courage
cupcakes
appreciated
attending
cowboys
delivered
savings
acoustic
seasons
tasty
specialist
scored
scotland
comfort
pacific
victor
protection
readers
lemon
develop
allows
cartoon
soap
cooler
produce
provides
sons
miracle
grass
globe
saint
cutie
volunteer
necklace
cricket
brave
chili
corn
tacos
quest
connected
courtesy
potato
dresses
spice
attractive
handsome
fortune
kissing
eaten
lion
solve
pancakes
wealth
inn
dawn
coolest
earned
pilot
environment
inspiring
sponsor
outdoor
perfectly
wink
belle
jumping
pets
muscle
freelance
cereal
membership
handy
earrings
clubs
sings
tunes
bunny
clever
houses
photographer
tastes
bowling
ooooh
electronics
trail
sweetheart
vibe
masters
accepted
promised
fridge
scientists
supplies
cutest
intelligence
answered
recognize
strategies
invest
homemade
equal
entertaining
emotions
polish
impressive
blessings
elvis
valentine
succeed
pillow
childhood
wonders
promises
documentary
hint
penny
expand
bake
chemistry
shining libertyblessing
perspective
creation
vanilla
surf
comics
paradise
neighborhood
resource
invented
secure
mint
innocent
toward
proposal
dish
achievement
powers
melbourne
fathers
cheaper
raising
revenue
cupcake
automatic
traveling
naturally
profits
nephew
adore fridays
presence
swiss
evolution
supports
sooner
karaoke
achieve
realise
developing
mercedes
fiction
plants
livin
punch
virus
gunsscream
costspakistan
ghost exams
emergency
slap
shock
niggaz
creepy
regretcough
deleted
trashcancelled
rough
cheat
worries
stole
madnessnoise
gripe closing
unfortunately
loans
starving
zombie
tear
terror
asshole
disasterdisappointed
wasted
hating
deadline
ughhbum
earthquake
refuse blocked
taxes
dammit
removemessed
nightmare
annoyed
screamingkickedwhip
illegal
suicide
prison sadly
badly desperate
whorenuclear
locked
violence
hater
grave
hangover
injury
jerk obsessed
afghanistan
abuse
damage
dire anti hardest
false
shitty
panic
charges
banned
forced
disease
angerattacks
ghetto
forgotten
criminal
delay
ruin
disgusting
owe
dislike
homeless
racist
complaining
envy feeprotest
cheating
flood
lied
threat
hardly
cried
kicking
arrest
mafia
fails
complicated
funeral
critical
wicked
injured
theft
painful
retarded accused suspect
beating
killintrapfallenshocked
liar
hated
useless
patients junk iraq
stolen
hurting
divorce
cruel
rage
recession
destroy
idiots screwed
hunger
addiction
stressed
enemy
suffering
struggle
monstersdustrevenge
danger bitter
bald
ignorant
fools
pimplawyer
depression
dirt
ruined
terminal
assault
depressed
suckedslut fees
pathetic crappy
victimmiserable
slip
suffer
curse
argument
scare
spitdelayed
exhausted
depressing
hurricane
fears
victims
minus
rape
eagles worship
traditional
vitamin
seas
adults
clap
broadway
eats
creates
restaurants
nations
native
oven
rum
fairy
popcorn
brunch
activities
guaranteed
elizabeth
niece
bounce
horses
nursing
automatically
guests
knight
approved
oregon
couples
rocket
confident
panties
resolution
engaged
wheels
creek
fixing
hats
springs classy
caring
snack
updating
sundays merry
mornings
poetry
sunset
breathing
healingparent
visitors
intelligent
cakes
temple
recommended
alliance
installed
trains
pumpkin
destination
incredibly
technologies
pregnancy
neighbor
lottery
touched
vagina
graphics
deserved
inspire
ruby
cooked
baked
chattinghumans
exhibition
persons
cotton
variety
breezetrips
excitement kissed
fireworks
cleaned
writes
atlantic
jelly
parks hooray
awsome
farmers
magical
whale
sausage
believes
environmental
craft
rabbit
gains
goods
instantly
earnings
skirtdestiny
eagle
superstar
improved
rains
selected
www
tours
agreement
sofa
treats
vocal
volunteers
crown
mercy
mature
champ
superman
olive
seed
embrace
spiritual
platinum
rome
concerts
flavor
charm
citizens
grows
importance
remembering
brains
lighting
ease
magazines
valid
medal
reward
employment
transport
dining
relevant
imagination
reception
heal
diamonds
certificate
waters
visits
humble
showers
investing
participate
successfully
champagne
resolve
tasting
gardens
finest
noble
genuine
visited
purchased
carnival
jamaica
bells
baking
styles
starring
protein
conversations
actors
ceremony
honored
treasure
roast
stereo
supporters
amazed
jenny
greetshiny
pies
leaf
rocked
granted
lawn
explore
worthy
skillcurrency
professionals
daughters
appropriate
futures
recommendations
rises
rides
bride
owns
saves
dew
improvement
upside
saturdays butterfly
productivity
palace
guarantee
significant
increased
generate
skating
experiences
juicy
partnership
themes
blanket
relations
wondered
neat
gained
attract
islands
shakespeare
seeds
retirecombination
shelter
shade
baker
accomplished
potatoes
heating
spotlight
practical
cowboy
autumn
feeds
proposed
loyal
donations
gathering
cuddle pleasant
logic
introduced
understood
providing
determined
approve
candle
informed
haven
encourage
safely valuable
dessert
assist
bliss
backyard
tale
electricity
painted
independence
approval
receiving
christians
sweden
correctly
appreciation
donated
groove
luckily
commitment
diploma
egypt
poem
dancer
sweetest
glorious
hired
oakinsight
drums
touching
feeding
recover
developed
sparks
mountains
cambridge
believing
supported
fairly
reliable
vehicles
maximum
alumnireceives
kindness
philosophy
attraction
committed
restore
improving
extend
invitation
lamb
loyalty perfection
initiative
increases
efficient
earning
faithful
survived
performances
circus surprising
portrait
introduction
dictionary
surprisingly
outstanding
recovering
provided
velvet
scientific
garlic
values
gardening
skies
adoption
comfy
increasing
venus
architecture
believed
performed
lipstick
ensure
athletes
reputation
relate
shoulders
beds
humanity
sleeps
democracy
honesty
sheep
possibility
authors
applying
satisfied
musicians
expression
belongs
spark
colours
connections
salary
expansion
belovedpeaceful
belief
kin
educational
divine
adopt tales
horizon
muscles
scientist
draws
accurate
rider
dove
allowing
achieved
capable
millionaire
beers
orchestra
cheek
architect
costumes
democratic
scholarship
sunrise
aims
overcome
wished
muffin
wildlife
fruits
organized
kitten
mamma
climbing
coin
romeo
prevention
neon
heritage
choir
acceptable
pony
indians
transportation
grants
exists
musician
wheat
abroad
overseas
deposit
singers
floating progressive
engage
newspapers
impress
eternal
greatly
mutual
glowawarded
arrival
colored
tourist
sketch
grin
acquire
reflection
unite
pioneer
dynamic
principles
exercises
ethics lime
mentor
bloom
gentle
cheeks
qualified
butterflies
constitution
collective
compassion
dances
teaches
reasonable
harmony
harbor
languages
emerging
weddings relaxed
prospect
eden
popularitygentlemen
landscape
engineers
enters
bridges
integrated
dancers
established
infinite
mansion
elected
rays
stable colleges
promoted
delight
bentley
promising
passionate
colleagues
gentleman
contribute
literature
extraordinary
recognized
wines
touches
trusted
spirits
asset
communicate
challenging
brighter
grandmother
graduated
preparation
nursesbuyer
recognition
silk
athletic
salvation
dough
understands
symphony
leap
tradition
tender
satisfaction
float
protected
lifted
uptown
switzerland
elegant summertime
essencerevolutionary
jingle
voyage
equally
generous
fountain
optimistic
candles
supporterdreamed
atmosphere
instrumental
miracles
phdregards
libraries
brewtourists
rhythm
waving
strongly
classical
dine
educated
rivers terrific
vocals
protecting
remedyguidance
assets
acquired
mellow
forgivenesscozy
springfield
pension
prospects
reflect
practically
proceeds
occasion
attracted
exclusively
chess softly
porch
dedication
convenient sailingunity
efficiency
sensation
possibilities
advantages
feast
wealthy
enthusiasm
riverside
sail favour
copper
equality
principle boulevard
sanity rhymes
daylight
doggy photographs
acceptance
celebrated
eligible
effectively
tasted
melody
sentimental
chickens
interests
travelers
chorus
mineral
preferred
integrity paintings
intellectual
farms
aircraft
streams
sung
resolved
proved
boats
contributions
photograph
encouraging
visible patiently
quicker
instrument
compensation
crop
participants
relative invention novelsfarmer
scoring
outdoors
conscious
yacht
sands
magnetic residence
riches
romans
recovered
theaters
v−day
investments
sailor
instruments
shines
integral
hallelujah
smiled
valued
functional
desires cared
flowing
scent
beings mirrors
edison
heavens
studied
aboard
wage
cheddar
balanced
earns
whistle
poet
triumph
concepts
honour
eager
gentlyheavenly
heartbeat infant
gravy literary
dearest championships
nomination sciences
relatives
survivorsoxygen
interaction
madame
universities
ambitious
aviation
restored
outlets
artistic
manners
drawings
numerous
intimate
satisfy
affection
clownshopedexistedfarmingherb
remarkable
increasinglyparticipation
archives
pearls eternity
grandfathercherish
sunlight
thrill
surviving
danced
relation
prosperity
determination
towns
ethical
proposals
legitimate
rhyme
dynamics
contribution conservation
privilege
observe
guitars
argue
struggling
misses
confusing
zombies
ratbeg wasting gangsta
delays
fraudenemies
fist
fights
operation
bothered
ruins
stealing
frustrated
disagree
bastard loses
accidentally
darkness
snake flies
deadly
yuck
garbage
busted
messy
removed
thirsty
ignorance
selfish burnsburned
attacked
suspended
yelling
chaos
anxiety
hatin
failing
excuses
fatal
cage
bullet
forgetting
violent
missin
losers
offense
dull burnt
canceled
limits
brutal
unemployme
mob
severe
combat
slam
lowest poop
nt
slave
killers
thug
yell
ashamed
ache infection
defeat
deny
poverty
cigarettes
politicians
obsession
cigarette
ignored
cheated
offensive
tornado
rival sink
poison
diss
syndrome
trapped
lawsuit
witch
devils
cracked
injuries
disabled
destroyed
crashedstroke
scandal weapon warned
misery
dump
risks
explosion deficit
creep
addict
symptoms
bleeding
annoy
unhappy
sucker
ambulance
stink
weapons
penalty
armed
aging
runaway
begging
disappear
arguing
fought
fucker
acid
unfair
pity
bastards
confusion
dim
decline
abortion
cries greedy
crimes
smashed
denied
struck
rejected
unable
cracks
robbed
buried
complaints
cocaine
weakness
torn opposition
hospitals
errors
conflict damned
harm deaf
breakdown
riot
pale
conspiracy
alleged
tragedy
bankruptcy
sins
anxious
threats
foul
wreck
foolish
dizzy
stranded goddamn
disorder chokeshouting
unlikely
tsunami
bleed complaint
crowded
skull
swine
sickness
punishment
farewell demands
murdered
corruption
paranoid
unnecessary
bombs
jealousy
psycho
eliminate
insanity
reckless
critics
beaten
robbery
satan
thief
lynch rats
whackinvasion
suspected controversy
creeping
drown
tragic
sucka
tension regrets
troubles
nightmares
fewer
doom
cryin
demon
criticism
aggressive
disrespect
unfortunate
artificialdefensive
bugs
torture
terrorist
deaths
raped
terrorism
screams
slapped rot
inflation
haunted
insecure gangster
illness
disappeared
filthy
downs
stab
losses threatened
opposed
destruction
stingdrunken
refused
collapse
motherfucker
chronic
sadness
hatred
battles
suspicious
crashing
bailout
denialbrat
abandoned
savagewhores
baddest
ghosts
burden buryarguments
absence
killa
nothings
rotten
frustration
snakes
sentenced
swollen
captured
suffered
graves convicted
damaged
crushed
expenses tobacco
nigger
protesters
heartbreak
thugs
guilt
dread doubts
griefwounded
scars
defeated
hangsvain
screamed
crudedemanding
shattered
lacking
thieves drowning
faggot
worthless
slipped
racial
voodoo
dispute
hopeless
demons
fooled
tore
bullets suspension vulnerable
pollution
darker
mistaken
difficulties
bore
restrictions
void
restless
disturbed
tearing
slavery
wound
scar critic
suspects
dusty
rejection
woe
arrests
forbidden
difficulty
oppose absent
wounds
raging
aching
abused
crooked
execution
withdrawal
sorrow bombing
rustyfragile
shotgun
expense
diseases
tumor
shove
discrimination
snatch
wrath
greed
despise
pistol
slaves
holocaust
agony
drowned
weep
heartache
tortured
killings
ownership
providence
suitable
nearest
acknowledge imagined
companion
sincere
beliefs
virtue
leisure
oceans
fluid
transformation
enable
sparkle
warmth
amour secured
precise
dignity improvements
employed
profession
generationspoems
logical
senses
qualities sweeter
trinity railway
cosmic
illustrated believer illustration
gladly
beamproceed
civilization
revival
consciousness
landmark
traveler
likewise
commonwealth
conscience
encouraged
fulfill
definitecurves
firefighters
produces
sculpture museums
observation
founded
forgiven
bedtime
acres woof
simplicity
approaches
prayed
agriculture
elder expressions
righteous
preference
visions stability
nephews
preserve
shores
immortal
brook
solving
automobile metropolitan hearted
significantly
strollcharms
performers
forthcoming behold
respected
poetsfeatherspecialty
volumes
almighty
developments
glanceconsensus
tremendous
advances
foxyrevenues
glowing
savior
accuracy
sophisticatedcontinuous
sixteen illustrations
mankind
devoted
advocates
moonlight
dreamer
sights extending
rational
branches
merit
fireplace
consideration
x−mas
contributed
nieces
freely
phenomenon
theoriescathedral
innocence
spontaneous
completion
furnished
audiences
mend seventeen
jewels
foundations
groovy
wages
crimson
contributing
painter
restoration
shopper
occasions
transformed
profound
altogether
everlasting
dearly
fantasies
composition lyrical
enabled
traditions
climbed
voluntary
cooperation
representation
treaty
traveled
justified
cradle
advocacy experiments
yields
purely
avid
monetary
ecstasy
significance
prominent
composer
shells
attachment
respects
negotiate
crops
reasoning
throne
functioning
clearer
continent
effectiveness
desired
narrative
assured
dimes
outcomes
prospective
metals
poetic employ princescultures salaries
goodman
antiques
surplus
cattle
roam
allies
velocity
praised
quantities
arise calcium
therapeutic
scholar
composed
architecturalsufficient
devotion
succeeded
pianist
villages
preservation
favorable
influences
raindrops
lullaby
reindeer
grandson
screenplay
centuries
1970s
scholars
comprehend
honorary
expedition
upward
1980s
deposits lifelong
1960s
carriage
distinction
starry
obtained
neighborhoodsinteractions
liberation
favored
objectives
reasonably
heartfelt
jukebox
ideals
infants
associations
philosophical
sails
agreements
distinct
prevail
gaze
saviour
starlight
agricultural
maintained
generosity
peacefully
acknowledged
competentallright
abundant
technological
greeks
meanings
victories
telecommunications
majesty
elevated
guiding
distinguished
writings
rudolph
liberties
nearer
eighteen
accompanied
iraqi
allegations
radiation
sinking
opponent
slipping
foe
troubled
stomp
cursedinferior fury
cemetery
prisoner
gutter
vicious
screamin injection
bacteria
damages
stain
incorrect
condolences
strain
disgraceneglected
misunderstood
prisoners
avoided
sanctions
mold
roughly
choking
phony
frown
separation
bruised
dismissed missile
darkest denying
despair
spite uncertainty
violation
motherfuckers
suckers muthafucka
prejudice
bruise
slaughter
snitch
urine
demise
disorders
havoc
violated
opponentsflee
feared
diagnosis
loneliness
muthafuckin
hustler
seized
separated
stainsbaghdad
resignedthirstuncertain
declined
sinner
restricted
haunt
neglect violations
executed
nixon
litigation
complained
infections
plead
coffinresignation
sleepless
inmates sirens
widow
hostile
erased
hauntingisolated
motherfucking
vengeance blinded
frightened
suspicion limitations
punks
betrayed
corpse
outdated
tomb
fades
penalties
strapped
misspelled
refugees
slug
unconscious
abnormal
heartbreaker
lawsuits
desperation
deception
scarred eliminated
shiver stained
fiend
distress
servants
thorn
mourn
impose
argues
pimps
helpless
conflicts
gangstas
condemned
fright
complications
tensions
decayaccusations
weeping
cowards
unclear
isolationdefect
saddened
dagger
reluctant
gloom
rotting
burial
sinners
slumdog
casket
inadequate
lonesome
missiles
cavity
deficiency
argued
motherfuckin
criticized incorrectly
discontinued
sorrows
unholy
obituary
remorse
turmoil
pressures
hussein
indicted
emptiness shatter deceive
confrontation
excluded
bleeds
foes
prosecution
combinations
respective
successor
begged
jaded
pleaded
confined
respectively
adapted
unwind
accompanying
initiated
treasurer
considerations proceeded
thorns
chained
defendant
noose
bookstores
stimulation
certainty
varieties
novelist
supermarkets
virtues
thrills
fled
subjected
fiends
rainbows
admiration
desirable
belonging
eternally
adored
succeeding
caress
exercised
belonged
readily
acquainted
artillery
sorely
mourns
sweetness
splendid
forests
grandchildren
cooperative
accurately
sympathetic
privileges
sufficiently
proteins
defects
tumors
laughter
continuity
aesthetic
distinctive
succession
fists
decreased
bled
conceived
applicable
justification
historian
rags
untruetaxation
forsaken
betray
saddam
car google
part
hahahaha
friend
funny
y ng
bea
utiful
t
slee
eat
making
girls
care
hair
p family
face
yea
open
boy
tempo
movie
found
okay
bien
book
full
high
play
album
heart
mind
believe
hello
fans
team
listeni
suppo
kids
myse
ing
fan
white
ng
rtr welc
remember
lflady
apple
websi
summe
ipad
rock
cute
talkin
thinkin
te
join
jam
mom
swee
ome
rain
sex
meet
g g email
bed
story
radio
idea
yay
sale
chance
favor
friday
download
women
excited
favor
mine
glad
power
enjoy
info
bro
dream
plan
learn
ite
bueno
dear
health
yours
easy
air
luck
latest
deal
elf
started
lmfao
toget
save
water
sexy
film
wonder
green
service
kind
photo
eyes
woma
hahah
dog
readin
chat
her
star
shopping
snow
plus
walk
photos
fine
perfect
avatar
game
n1st
future
gspecial
everybody
tour
park
availab
visit
design
eating
young
cd
giving
coffee
fast
worth
drive
amigo
words
jobs
london
rest
hehehe
sun
congsbday
le
smile
interesting
art
tips
tickets
boys
forward
mobile
starting
pictur
ipod
dance
christmas
saturd
rats
dvd
ate
mucho
beac
videos
shows
college
mama
agree
date
light
gift
totally
event
esongs
hotel
ay
band
pics
h
foreve
digital
kid
trip
sea
market
understand
south
chocolate
american
windows
usa
brother
lots
touch
rshare
loves
computer
holida
jesus
answer
shop
dad
million
voted
living
drink
king
nap
bought
y won
lovely
lucky
huge
loved
couple
writing
study
chicken
yours
gave
able
english
ladies
seeing
sunday
importa
country
definitely
hahaha
hands
personal
message
finished
release
network
dinner
tea
april
pizza
sister
trust
quick
clean
act
knows
softwa
fashion
liked
luv
haha
nt
simple
energ
awards
spring
travel
exactly
football
fresh
re
ride
extra
shower
y gold
blackberry
america
laptop
june
plans
movies
truth
concert
itunes
success
males
camer
studio
pictures
dress
ideas
festival
books
kiss
eye
style
loving
sports
proud
hopefu
experience
walking
driving
guide
ahaha
shoes
cat
parents
babe
hahaha
dreams
asafe
possible
child
teacher
international
mother
finish
wife
artist
xbox
fly
sing
mail
lly
strong
breakfast
children
pink
sis
fix
lord
hwonde
rful laugh
choice
self
straight
wedding
sense
create
cash
fair
offer
commu
paid
original
human
santa
india
shirt
complete
vegas
peace
informatio
nity
joined
security
biggest
smart
color
thoughts
gives
warm
alright
grow
played
global
contact
stars
earth
ncream
cheese
build
interested
brothers
ball
clothes
thinks
cake
fit
soul
paris
beaut
tech
students
quality
especially
research
lead
updated
favorited
winner
female
natural
planning
singing
ahead
library
launch
healthy
wine
y
holy
joke
united
pray
lives
brazil
student
sky
opening
miami
brain
popular
flight
ring
learning
aah
buying
imagine
respect
absolutely
cafe
pool
wed
folks
camp
technolog
toy
focus
enjoyin
relationship
november
stories
cook
hoping
performance
gorgeous
advice
science
taste
sites
washington
giveaway
classic
wondering
alive
wins
born
grand
hilarious
vacation
disney
magic
japan
appreciate
g
y
dogs
amazon
released
married
bike
award
clear
likes exciting
papa
dating
california
content
bless
greatest
mega
university
professor
cinema
excellent
positive
fantastic
celebrate
island
daddy
goal
begin
s
points
places
developme
trade
aaah
fish
sleeping
tune
yahoo
promise
epic
collection
completely
asleep
cards
events
sunny
mate
congratulation
estate
september
updates
present
rich
drinking
christian
wii
sony
land
nt
orange
educati
famous
window
dancing
gain
pet
tree
theme
guitar
restaurant
laughing
moms
garden
evening
creativehumor
teach
starbucks
helping
cars
opportun
on
glee
treat
boring
missing
slowbang
fake suck
lie
painpoor
worst
cryfail
traffic
shot
nobody
blood rip
bitches
shoot
stuck
cannot
loss
worry
hospital
−5
6
ymm
oyue
up lol
no
donot
−2
−3
Blogs
T
log10 normalized word frequency f (∆havg )
Measuring
Happiness
6.4
W
∆havg = 1.00
−1
2009−05−21 to 2010−12−31:
6.45
6.3
T
0
Complex
Sociotechnical
Systems
Twitter—weekly time series:
assembled
considerable
equilibrium
cherished
attained
slugs
dividends
granddaughter
trembling
servings
teardrops
ceased
sailed
arises
preserved
adoring
satisfactory
tenderness
comforted
unkind
honky
indictment
−6
tremble
darkened
4
T
W
T
F
S
S
M
T
W
T
F
S
S
M
day of week
Hour 13, 2010
−7
−4
−2
0
2
Average happiness havg - 5
55 of 83
Complex
Sociotechnical
Systems
Tref : Tuesdays (havg =6.03)
Tcomp: Saturdays (havg =6.06)
love +↑
no −↓
haha +↑
party +↑
fun +↑
saturday +↑
0.2
Measuring
Happiness
+↓new
weekend +↑
not −↓
happy +↑
dont −↓
10
−↑last
Word rank r
Analysis
live +↑
die −↓
friends +↑
game +↑
con −↓
movie +↑
cant −↓
:)
!
Blogs
Tweets
−0.8
References
Positivity Bias
References
−1
−1.2
09/09/08
12/01/08
03/01/09
sick −↓
shopping +↑
playing +↑
B
−↑don’t
Text size:
Tcomp
amazing +↑T
ref
bad −↓
awesome +↑
homework −↓
wedding +↑
−↑hangover
−↑miss
+↓free
2
10
45
3
10
4
10
50
Balance:
−87 : +187
+↓
+↑
shit −↓
court −↓
nice +↑
won +↑
0
P
r
i=1
100
−10
movies +↑
−5
0
06/01/09
09/01/09
−1
−2
−3
−4
−5
09/09/08
+↓school
δhavg,i
09/01/09
10
1
10
log rel freq
40
06/01/09
date
+↓lunch
0
Analysis
Songs
Positivity Bias
birthday +↑
10
Data sets
Sad
Tweets
great +↑
sunday +↑
family +↑
beautiful +↑
beach +↑
home +↑
35
:(
Afghanistan
−0.6
Blogs
+↓google
30
Happy
Tea Party
−0.4
Songs
−↑fight
25
Measuring emotional
content
−0.2
Data sets
hahaha +↑
20
Some motivation
Measuring emotional
content
−↑bored
−↑drunk
15
Measuring
Happiness
0
Some motivation
avg
5
Complex
Sociotechnical
Systems
0.4
A
h(amb)
1
4
−↑
12/01/08
03/01/09
date
−↓
5
56 of 83
10
59 of 83
Per word average happiness shift δhavg,r (%)
The daily unravelling of the human mind:
A
B
Tiger Woods
0.2
0
(amb)
avg
h(amb)
avg
0
−0.2
h
−0.4
−1
8
12 16 20 24
4
8
0.03
0
0
2
4
6
8
10 12 14 16 18 20 22 24
C
hour of day (local time)
12/01/09
6
5.95
noon
Wednesday
midnight
noon
Thursday
midnight
noon
05/01/10
07/01/10
Some motivation
date
03/01/10
Data sets
−4
Analysis
−5
05/31/10
Measuring emotional
content
−3
12/01/09
03/01/10
Songs
05/31/10
Blogs
date
T re f: All Tweets (h avg=6.06)
T co m p: Tiger Woods (h avg=5.74)
T ref: All Tweets (h avg=6.04)
T c o m p: BP (h avg=5.57)
Friday
midnight
noon
Saturday
midnight
noon
Sunday
midnight
noon
Monday
midnight
noon
5
wife +↑
car +↑
+↓me
0
santa +↑
10
−↑cheated
15
−↑cheat
2
sex +↑
+↓haha
−↑alone
−↑not
+↓happy
−↑breaking
−↑divorce
10
20
3
10
4
10
−100
25
Pr
movie +↑
0
i =1 δ h av g, i
−20
Text size:
T r e f T c om p
ill −↓
1
10
midnight
Local time
−↑disaster
1
−↑alleged
10
D
no −↓
−↑scandal
+↓love
−↑hospital
−↑cheating
5
6.05
Tuesday
03/01/10
date
−↑accident
−↑crash
−↑injured
1
2010
midnight
01/01/10
Tweets
12 16 20 24
Word rank r
Average happiness havg
09/01/09
6.1
noon
Measuring
Happiness
−1
06/01/10
−4
−5
0.02
6.15
Monday
04/01/10
−3
0.01
4
02/01/10
log10rel freq
10
0.04
hour of day (local time)
5.9
midnight
12/01/09
+↓good
0
Ba la nce:
−188 : +88
+↓
+↑
−↑
−↓
20
Per word average happi ness shi ft δ h av g, r (%)
Word rank r
6
log rel freq
count (%)
havg
6.05
0
10/01/09
date
0.05
6.1
−0.4
−0.8
−0.8
0.06
2009−05−21 to 2010−12−31:
−0.2
−0.6
−0.6
6.15
Complex
Sociotechnical
Systems
BP
0.2
10
0
dont −↓
1
−↑died
billion +↑
con −↓
2
10
−↑blame
−↑criminal
20 3
10
+↓happy
−↑damage
4
−↑costs
10
−100
0
−↑fails
Pr
−↑crisis
25
i =1 δ h av g, i
−20
−10
Text size:
T r e f T c om p
ill −↓
+↓hahaha
10
References
no −↓
−↑down
−↑shut
−↑kill
+↓love
−↑fake
−↑stop
+↓me
+↓haha
−↑not
+↓lol
10
15
Positivity Bias
Ba la nce:
−167 : +67
+↓
+↑
−↑
0
10
−↓
20
Per word average happi ness shi ft δ h av g, r (%)
60 of 83
A
B
Pope
0.2
0
(amb)
avg
h(amb)
avg
−0.2
h
−0.4
−1
02/01/10
04/01/10
06/01/10
Measuring
Happiness
−1
08/01/10
11/01/08
01/01/09
log10rel freq
10
−4
−5
03/01/10
03/01/09
05/01/09
date
date
−3
−3
−4
−5
09/09/08
05/31/10
12/01/08
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
03/01/09
Songs
Blogs
date
date
Blogs
Tweets
T ref: All Tweets (h avg=6.04)
T c o m p: Pope (h avg=5.40)
−↑abuse
1
Word rank r
sex +↑
15
child +↑
−↑not
−↑resignation
+↓me
+↓haha
−↑crisis
+↓lol
0
10
1
10
Text size:
T r e f T c om p
dont −↓
2
−↑crimes
−↑attacks
−↑accused
−↑deaf
−↑down
4
+↓hahaha
10
−100
0
−↑evil
Pr
−↑rape
25
i =1 δ h av g, i
10
20
5
no −↓
−↑against
+↓love
−↑arrested
10
Ba la nce:
−157 : +57
+↓
+↑
3
10
−50
−↑war
1
−↑scandal
−↑victims
−↑arrest
5
T ref: All Tweets (h avg=6.05)
T c o m p: Israel (h avg=5.30)
D
−↑
0
−↓
50
Per word average happi ness shi ft δ h av g, r (%)
Word rank r
C
Complex
Sociotechnical
Systems
−0.4
−0.8
−0.8
Happiness in Manhattan (just for fun):
−0.2
−0.6
−0.6
log rel freq
Complex
Sociotechnical
Systems
Israel
0.2
0
−↑fired
−↑against
−↑killed
−↑attacks
−↑attack
−↑offensive
−↑fire
−↑crimes
Positivity Bias
References
References
peace +↑
10
−↑no
−↑not
−↑die
−↑killing
−↑stop
15 101
+↓love
−↑bombs
2
−↑fighting
10
−↑kill
−↑kills
20 3
10
−↑conflict
−↑death
4
−↑terrorist
10
−100
0−↑weapons
Pr
25
i
i =1 δ h av g,−↑bombing
Text size:
T r e f T c om p
0
10
−10
Tweets
Positivity Bias
Ba la nce:
−131 : +31
+↓
+↑
−↑
0
See Blog post on onehappybird ()
−↓
61 of 83
10
64 of 83
Per word average happi ness shi ft δ h av g, r (%)
13
Word
1. love
2. happy
3. win
4. kiss
5. cash
6. vacation
7. Christmas
8. God
9. party
10. sex
11. Valentine
12. family
13. sun
14. life
15. hope
16. heaven
17. :)
18. income
19. friends
20. snow
21. :-)
22. night
23. vegan
24. Jesus
25. girl
26. USA
27. you
28. our
29. ;)
30. health
31. tomorrow
32. !
33. summer
34. we
35. today
36. man
37. woman
38. Stephen Colbert
39. ;-)
40. RT
41. coﬀee
42. church
43. work
44. I
45. yes
46. them
47. hot
48. boy
49. yesterday
50. Michael Jackson
(amb)
havg
+1.42
+1.32
+1.26
+1.21
+1.21
+1.11
+1.03
+0.95
+0.93
+0.89
+0.85
+0.79
+0.65
+0.50
+0.48
+0.43
+0.42
+0.36
+0.33
+0.32
+0.32
+0.29
+0.28
+0.27
+0.25
+0.23
+0.22
+0.21
+0.20
+0.20
+0.20
+0.16
+0.13
+0.13
+0.13
+0.12
+0.10
+0.10
+0.10
+0.06
+0.04
+0.03
+0.02
+0.02
+0.02
0.00
-0.01
-0.01
-0.01
-0.02
Total Tweets
46,687,476 (6)
16,541,968 (13)
7,981,856 (26)
1,697,405 (59)
1,279,236 (63)
934,501 (67)
4,887,968 (35)
8,576,364 (25)
6,438,886 (29)
3,551,767 (39)
247,288 (84)
5,014,816 (32)
2,385,348 (52)
14,006,454 (17)
11,833,337 (18)
741,878 (71)
10,470,483 (20)
510,425 (76)
7,669,719 (27)
2,596,165 (49)
1,680,165 (60)
17,089,505 (12)
183,889 (90)
2,027,720 (56)
10,070,132 (22)
2,157,172 (54)
173,276,993 (3)
14,062,465 (16)
2,618,940 (48)
2,575,543 (50)
10,379,637 (21)
3,463,257 (40)
2,998,785 (43)
39,132,934 (7)
25,588,506 (9)
15,856,341 (14)
2,543,036 (51)
23,778 (99)
943,413 (66)
339,055,724 (1)
2,800,972 (46)
1,812,251 (58)
18,415,618 (11)
307,960,343 (2)
11,593,356 (19)
15,352,295 (15)
7,122,144 (28)
4,933,333 (33)
3,077,761 (42)
825,979 (70)
Total ANEW
85,269,499 (5)
32,442,529 (8)
14,640,728 (20)
3,162,330 (48)
2,468,496 (51)
1,783,270 (56)
10,645,630 (25)
17,867,768 (16)
12,090,597 (23)
7,087,972 (31)
464,914 (75)
10,629,361 (26)
4,602,627 (44)
27,770,768 (10)
22,952,366 (13)
1,485,702 (59)
6,787,678 (35)
418,161 (77)
7,541,106 (29)
5,011,785 (40)
1,102,512 (67)
17,606,796 (17)
178,676 (86)
1,673,992 (58)
19,886,691 (14)
1,204,585 (65)
145,464,084 (2)
14,437,899 (21)
1,475,221 (60)
4,950,202 (41)
8,899,406 (28)
1,385,072 (62)
2,554,459 (50)
34,513,587 (7)
23,619,518 (12)
29,558,118 (9)
5,603,347 (39)
14,697 (99)
516,171 (73)
142,219,359 (3)
2,399,867 (52)
3,452,171 (45)
16,191,802 (18)
282,865,043 (1)
7,499,840 (30)
14,398,889 (22)
6,286,163 (37)
9,670,512 (27)
2,852,623 (49)
571,442 (71)
Complex
Sociotechnical
Systems
Twitter—location:
(amb)
Total Tweets Total ANEW
Word
havg
51. me
-0.06 144,342,098 (4) 88,088,051 (4)
52. ?
-0.07 2,333,283 (53)
674,679 (69)
53. commute
-0.09
90,126 (94)
90,092 (92)
54. gay
-0.09 2,727,309 (47) 1,697,177 (57)
55. right
-0.10 19,166,480 (10) 15,850,283 (19)
56. school
-0.11 9,264,217 (24) 6,924,193 (34)
-0.13
229,773 (86)
188,338 (85)
57. Republican
58. they
-0.16 27,442,360 (8) 27,150,189 (11)
59. winter
-0.19 1,255,945 (64) 1,217,225 (64)
60. lose
-0.19 2,056,468 (55) 2,091,540 (53)
61. Jon Stewart
-0.20
52,084 (97)
33,086 (96)
62. gas
-0.22 1,022,879 (65)
812,029 (68)
63. no
-0.22 95,129,093 (5) 38,894,616 (6)
64. Democrat
-0.23
93,193 (93)
75,450 (93)
65. left
-0.27 4,893,634 (34) 4,611,878 (43)
66. Senate
-0.29
447,732 (78)
316,835 (80)
67. election
-0.30
560,184 (75)
375,055 (78)
68. Sarah Palin
-0.34
225,577 (87)
150,096 (88)
69. Obama
-0.35 2,981,150 (44) 1,998,326 (54)
70. economy
-0.36
608,878 (73)
460,834 (76)
-0.36
391,510 (79)
279,695 (81)
71. Congress
72. drugs
-0.39
509,606 (77)
469,091 (74)
73. Muslim
-0.42
215,300 (88)
146,506 (89)
74. George Bush
-0.43
32,341 (98)
23,102 (98)
75. climate
-0.44
364,177 (80)
229,129 (83)
76. Pope
-0.51
152,320 (91)
135,955 (90)
77. oil
-0.53 1,377,355 (62) 1,148,990 (66)
78. I feel
-0.54 5,173,513 (31) 4,702,352 (42)
79. Glenn Beck
-0.54
113,991 (92)
101,090 (91)
80. Islam
-0.54
187,223 (89)
70,311 (94)
81. :-(
-0.65
341,141 (81)
244,215 (82)
82. :(
-0.70 2,907,145 (45) 1,891,225 (55)
83. flu
-0.75
901,403 (68)
639,000 (70)
84. rain
-0.78 3,233,464 (41) 5,959,903 (38)
-0.78
582,167 (74)
326,100 (79)
85. BP
86. mosque
-0.79
69,812 (95)
46,736 (95)
87. dark
-0.95 1,577,553 (61) 3,233,911 (47)
88. Lehman Brothers -1.08
8,500 (100)
4,280 (100)
89. Goldman Sachs
-1.08
52,703 (96)
30,769 (97)
90. Afghanistan
-1.15
273,519 (83)
172,637 (87)
91. Iraq
-1.37
238,931 (85)
213,425 (84)
92. cold
-1.39 3,670,447 (36) 7,015,518 (32)
93. gun
-1.81
680,903 (72) 1,263,217 (63)
94. hate
-2.43 9,652,881 (23) 18,158,870 (15)
95. hell
-2.49 6,266,162 (30) 11,056,735 (24)
96. sick
-2.55 3,576,058 (37) 6,783,395 (36)
97. sad
-2.56 3,563,745 (38) 6,951,686 (33)
98. war
-2.63 1,955,901 (57) 3,417,588 (46)
99. depressed
-2.64
280,872 (82)
541,394 (72)
100. headache
-2.83
856,600 (69) 1,446,064 (61)
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
65 of 83
(amb)
invite the reader to explore the tables beyond the observations we detail here. We begin with the highest
(amb)
and lowest rankings of ambient happiness havg , for
our list, finding them to be reassuring sensible. The
three top ranked words are ‘love’ (havg =+1.42), ‘happy’ (havg =+1.32), and ‘win’ (havg =+1.22), and the
last five, in reverse order, are ‘headache’ (havg =−2.55),
‘depressed’ (havg =−2.56), ‘war’ (havg =−2.63), ‘sad’
Text element and context correlate in happiness
scores:
I
I
I
I
Compare ambient happiness with text element
happiness.
Spearman correlation coefficient:
rs ' 0.79, p-value < 10−10 .
Complex
Sociotechnical
Systems
Complex
Sociotechnical
Systems
Twitter—location:
Tref : NY (havg =6.32)
Tcomp: CA (havg =6.38)
Measuring
Happiness
mad −↓
1
Some motivation
fun +↑
beach +↑
home +↑
hell −↓
dead −↓
free +↑
hate −↓
Measuring emotional
content
Data sets
5
Analysis
Songs
Blogs
Positivity Bias
References
An on-average result: says nothing about any
individual sentence.
Data sets
Songs
Blogs
Tweets
10
+↓snow
−↑time
cold −↓
10
Text size:
Tref Tcomp
Positivity Bias
References
−↑rain
fall −↓
15
+↓music
+↓sex
1
10
Extra random piece: stemming fails.
20
bus −↓
cut −↓
2
10
−↑bomb
−↑ambulance
10
25
0
P
r
i=1
−20
100
δhavg,i
−10
Balance:
−79 : +179
+↓
+↑
square −↓
hit −↓
accident −↓
−↑
3
63 of 83
Some motivation
Analysis
car +↑
0
Measuring
Happiness
Measuring emotional
content
+↓wit
Tweets
Word rank r
TABLE III: A selection of 100 keywords and text elements ordered by average ambient happiness havg . We define ambient
happiness of a keyword as the average happiness of tweets containing that keyword, relative to the overall happiness of tweets,
havg � 6.37. The number of tweets and total number of ANEW study words are listed in the third and fourth columns, with
the ranking of the keyword according to these quantities shown in brackets. Note that all pattern matches with tweets were
case-insensitive.
−↑fire
0
10
−↓
20
Per word average happiness shift δhavg,r (%)
66 of 83
Twitter—popularity based on follower count:
09/09/08 − 06/30/09
9
# Words
# ANEW
8
10
7
10
6
10
Measuring
Happiness
8
10
6
5
10
0
10
1
10
2
3
10
4
10
10
5
10
Complex
Sociotechnical
Systems
Twitter—interactions:
09/09/08 − 06/30/09
10
10
10
Complex
Sociotechnical
Systems
0
10
10
1
10
2
10
3
10
4
10
5
10
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
Songs
Blogs
Ambient Diversity (N(amb)
)
S
Happiness (havg)
6.5
6.45
6.4
6.35
6.3
6.25
0
10
1
10
2
3
10
4
10
Positivity Bias
Positivity Bias
40
References
References
20
0
−20
−40
−60
Background: 336.96
−80
5
10
10
Tweets
60
0
10
1
10
Follower Count
I
Blogs
Tweets
6.55
2
10
3
10
4
10
5
10
Follower Count
I
Decay in happiness correlation in social network.
I
ρ = Spearman’s correlation coefficient.
Dunbar’s number ' 150.
67 of 83
70 of 83
free +↑
1
0.15
+↓home
hate −↓
sick −↓
bored −↓
good +↑
5
Complex
Sociotechnical
Systems
Positive bias in the English language:
T ref: ≤ 10 2 followers (h avg=6.29)
T c o m p: ≥ 10 3 followers (h avg=6.44)
Measuring
Happiness
Some motivation
0.125
Measuring emotional
content
Data sets
people +↑
hell −↓
stupid −↓
love +↑
cold −↓ Text size:
10
0
Tweets
1
Positivity Bias
References
0.05
+↓god
money +↑ Ba la nce:
20
0.075
social +↑
success +↑
sad −↓
10
Songs
Blogs
T r e f T c om p
+↓sleep
10
15
Analysis
0.1
N
Word rank r
+↓bed
+↓christmas
2
0.025
−93 : +193
+↑
10
headache +↓
−↓
−↑news
lost −↓
hungry −↓
−↑
war −↓
3
10
25
0
P
r
100
i =1 δ h av g, i
−20
0
1
−↓
2
3
4
5
−10
0
10
6
7
8
9
havg
−↑failure
20
Per word average happi ness shi ft δ h av g, r (%)
71 of 83
6000
6.1
4000
6
2000
5.9
1
45
90
135
180
225
day number
270
315
0
0.04
1
3
5 7
havg
100
75
50
25
0
9
positive
0.08
negative
0
percentile
0.12
percentage
A. Twitter
100
75
50
25
0
1 2 3
|havg − 5|
4
0.04
Complex
Sociotechnical
Systems
B. Books
1
3
5 7
havg
Measuring
Happiness
9
Some motivation
positive
Measuring emotional
content
negative
0
1 2 3
|havg − 5|
Data sets
4
Analysis
Songs
Blogs
Tweets
0
1
0.16
0.12
0.08
0.04
0
1
2
100
75
50
25
0
100
75
50
25
0
3
4
5
6
7
8
9
0
1
0.16
C. New York Times
0.12
1
3
5 7
havg
9
positive
0.08
negative
0
1 2 3
|havg − 5|
2
3
percentile
6.2
100
75
50
25
0
0.16
4
4
percentage
8000
6.3
percentile
10000
6.4
0.08
percentage
12000
percentile
6.5
0.12
100
75
50
25
0
percentage
14000
Normalized frequency P
6.6
average follower number
average valence v
0.16
0.04
5
6
7
8
9
0
1
2
100
75
50
25
0
100
75
50
25
0
3
4
5
6
7
8
9
Positivity Bias
References
D. Music Lyrics
1−1000
1
3
5 7
havg
1001−2000
9
2001−3000
3001−4000
positive
4001−5000
negative
0
1 2 3
|havg − 5|
2
3
4
4
5
6
7
8
9
Average happiness havg
72 of 83
Complex
Sociotechnical
Systems
New York Times
Complex
Sociotechnical
Systems
1
ot
i we new illion
the he
no n
ms
m
ing e
x
g
lov
t
ly l
l
aol methin ook union ive womenparty onlov
rta recotspita eas
o
s
g
l
l
r
r
w
sm
er nco h
a
i
y
ts
e
ers bi ion
nr
es l
cer
bt
doca paren peacwin
nc tria arm offic olum act
can
de
c
lon
ole
k
e ves re
nttack n ond d
r
b
t
o
o
l
y
a
s
i
it
sa
e
r
vis go
ma cut figu netw
pnrice mou ecclin low te
d
d
e
g
viorle
ol
nin l
ex ed
e
s g
s
ey
ks e
rfcuel ssfu
rd g
e
ffr pas soccersong rtiensdwein
u
ne
ac los
u
e
t
k
s
e
o
ir
i
j
t
l
u
n
s
m killi
a
a y str req i
pawo suc
e
l
g
u
ly l
e s
in
ge enc
ng
re ama erg dispute eavybox hed twic site
gh gir
so
hi
h
ilu d em
c
r
ls
fa e
s
on
ua o
se ght dia uly erfect
us
t
ti
i
a
d
s
s
b
fl
v
u
ine
a
p
tr
fee en
tau in
arr
un r
ad
ve
dy
ion p
uto
si
er
jail ry
r e tion
ed c
te n
uslyrat ma attend deals ond sucpoem
wasoffe warn rose seriofede
ra
inju
w
p
leb
t ths
e
e
r
y
s
ly
e
g
c
h
i
t
e
p
p grap terary oura ins
dg
ous dly ar
rodr ea
en
li enc n w
ndean
hits sohra bio
one har sh
ter
nyell
wou ndt err
tio
at
n funexc
a
r
o
y
e
t
t
ti
ts
s
s
b
o
ane
b pil
is
n spira
d
ym ars
ent al fendag
in bo
ners ye
pe ind
plo
e ictm riv de
m
win enjo
m
r
em s r f ind k
ly ad on
de lkco hase que nks o
al
udnisaesste
i ha her
nesacand
at houl
low
o
n
r
c
f
n
o
r
s
i
w
d
t
r
u
a es
rs pu ts
ss
ss
sa
cle nc ation eting oye
c
ce
e
re aine
p
fer king ies ure
uen ue
pe
rap
pl
os pre coo ictor
s ens eq estigcom
sick coc
as
v
em pr ion ed
suitsoff ons inv
w
ing s dpeled
la ge actes
ss el
ip
ngs
re
ov
am
es ch impre rav
pr southcece
dbilli
s sadd dama inm
r
le
r
f
m
p
t
i
m
wa
A. Twitter
al
Word Usage Frequency Rank r
500
1000
1500
2000
2500
3000
3500
4000
4500
wafur ner
5000
1
2
3
4
5
6
7
8
Measuring
Happiness
Some motivation
B. Books
1000
1000
2000
2000
3000
3000
Measuring
Happiness
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
Word Usage Frequency Rank r
1
1
4000
Songs
4000
Blogs
Tweets
5000
0
1
2
3
5000
0
1
2
3
Positivity Bias
References
1
1
C. New York
Times
1000
2000
3000
3000
4000
4000
1
Average happiness havg
2
D. Music
Lyrics
1000
2000
5000
0
9
Data sets
Analysis
3
5000
0
1
2
3
Standard deviation of happiness hσ
73 of 83
1
1
A. Twitter
Word Usage Frequency Rank r
Complex
Sociotechnical
Systems
1000
2000
2000
3000
3000
Measuring
Happiness
Some motivation
Measuring emotional
content
4000
Songs
4000
Blogs
Tweets
3
4
5
6
7
8
5000
9
1
Gross National Happiness Index, hedonometer.org
(in development)
I
Prediction . . .
I
Scores for letters, phonemes, as a function of tense.
2
3
4
5
6
7
8
9
Positivity Bias
I
References
1
Fifteen additional languages being scored on
Mechanical Turk
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
1
I
How does happiness vary with proximity to nature?
to Walmart?
2000
I
Emotional contagion.
3000
3000
I
Quantifying metaphor and narrative and stories . . .
4000
4000
C. New York
Times
1000
1000
2000
5000
1
2
3
4
5
6
7
8
9
5000
1
D. Music
Lyrics
2
3
4
5
6
7
8
9
74 of 83
Average happiness havg
Complex
Sociotechnical
Systems
Books
1
500
Word Usage Frequency Rank
Complex
Sociotechnical
Systems
Measuring
Happiness
I
Data sets
Analysis
2
Random other things (now and next):
B. Books
1000
5000
1
76 of 83
1000
1500
2000
2500
3000
3500
4000
4500
5000
0
g
on
Measuring
Happiness
e
st r al
0.5
References I
Complex
Sociotechnical
Systems
h
his wit
e
i
on us own work
e
nc ugh
d
r
ng rie
thi xpe eno nea lor
g
my
e
din
n ear a
es etc rstan aining tratioratur indi rical pay
e
to
typ
t
tr
is
e
d
his
t min li
r
n
e
un ed
e ad vy
rn
b
er
sitlee ow reem hea outhe cem
g
s
mm
ica
de
sh a
su
afr
h
san
s h try inter
st
ht arc
e
ig
a
i
s
w
co stor fis poe
se
ium l
ly
ed a
m ion pre les eight tural bar
’s
g
h
apgeod
c
n
na
s
here’s
o
s
ce ti e
es
k
ow nan ump
d
c
ch
te
d
d
e
o
i
n
te s
ur
n
le
m qu
wi ain on bib
chol
w
s
m
c
o
h
s
d
r
l
nssn lco n
h ommile un extracoppe ora
a tio
tia
inc ro s
so
is
a
n
s
e io
hr e adliist
liamcreativ flect trip bondlespchon cria
il
n
w
te
so ecall
an
re
r
e nts ke
il
n
h
t
a
y
a
t
w
ot ean rticipa ist
we org bus
ne scm
cy
sts
ml
pa
an
edrie
ar p
gn
pa
ici mantiecxcitedha ppe cat pre
n
ty
u ro
sa l
ds ids rtali
m
edid ca
s lor
s
a mo
ral
hs
ledg gi
ve
ente s ine atom
est
now eolo elati
u
k
q
c
sev topicm
a s th
r
con
n
d
t
y
rs tio
r
cove fini sooneaime loyalt buersvotionislam lism
d
ita
de
cap
se are
the
77 of 83
1
1.5
2
2.5
[1]
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
M. Bradley and P. Lang.
Affective norms for english words (anew): Stimuli,
instruction manual and affective ratings.
Technical report c-1, University of Florida,
Gainesville, FL, 1999. pdf ()
Measuring
Happiness
T. Conner Christensen, L. Feldman Barrett,
E. Bliss-Moreau, K. Lebo, and C. Kaschub.
A practical guide to experience-sampling
procedures.
Journal of Happiness Studies, 4:53–78, 2003.
Positivity Bias
Tweets
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
[2]
References
[3]
References
M. Csikszentmihalyi.
Flow.
Harper & Row, New York, 1990.
3
Standard deviation of happiness hσ
75 of 83
78 of 83
References II
[4]
[5]
[6]
M. Csikszentmihalyi, R. Larson, and S. Prescott.
The ecology of adolescent activity and experience.
Journal of Youth and Adolescence, 6:281–294,
1977.
E. Diener and M. Y. Chan.
Happy people live longer: Subjective well-being
contributes to health and longevity.
Applied Psychhology: Health and Well-Being,
3:1–43, 2011. pdf ()
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
References V
[13] S. Lyubomirsky.
The How of Happiness.
The Penguin Press, New York, 2007.
Data sets
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
[14] C. Martinelli and S. W. Parker.
Deception and misreporting in a social program.
forthcoming in Journal of the European Economic
Association, 2007. pdf ()
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
[15] G. Miller.
Social Scientists wade into the Tweet stream.
Science Magazine, 333:1814–1815, 2011. pdf ()
P. S. Dodds and C. M. Danforth.
Measuring the happiness of large-scale written
expression: Songs, blogs, and presidents.
Journal of Happiness Studies, 2009.
doi:10.1007/s10902-009-9150-9. pdf ()
[16] G. Mishne and M. de Rijke.
Capturing global mood levels using blog posts.
AAAI 2006 Spring Symposium on Computational
Approaches to Analysing Weblogs, 2005.
79 of 83
References III
[7]
P. S. Dodds, K. D. Harris, I. M. Kloumann, C. A.
Bliss, and C. M. Danforth.
Temporal patterns of happiness and information in a
global social network: Hedonometrics and Twitter.
PLoS ONE, 6:e26752, 2011.
Draft version available at
http://arxiv.org/abs/1101.5120v4.
Accessed October 24, 2011. pdf ()
Complex
Sociotechnical
Systems
W. T. Jones.
The Classical Mind.
Harcourt, Brace, Jovanovich, New York, 1970.
[9]
D. Kahneman, A. B. Krueger, D. A. Schkade,
N. Schwarz, and A. A. Stone.
A survey method for characterizing daily life
experience: The day reconstruction method.
Science, 306(5702):1776–1780, 2004. pdf ()
[10] M. A. Killingsworth and D. T. Gilbert.
A wondering mind is an unhappy mind.
Science Magazine, 330:932, 2010. pdf ()
[11] I. M. Kloumann, C. M. Danforth, K. D. Harris, C. A.
Bliss, and P. S. Dodds.
Positivity of the English language.
PLoS ONE, 7:e29484, 2012.
Draft version available at
http://arxiv.org/abs/1108.5192. Accessed
October 24, 2011. pdf ()
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Some motivation
Measuring emotional
content
Measuring emotional
content
Data sets
Data sets
Analysis
Analysis
Songs
Songs
Blogs
Tweets
Positivity Bias
80 of 83
References IV
References VI
Measuring
Happiness
References
[8]
82 of 83
Complex
Sociotechnical
Systems
Measuring
Happiness
Some motivation
Measuring emotional
content
Data sets
Analysis
Songs
Blogs
Tweets
Positivity Bias
References
[12] R. Layard.
Happiness.
The Penguin Press, London, 2005.
81 of 83
[17] E. Sandhaus.
The New York Times Annotated Corpus.
Linguistic Data Consortium, Philadelphia, 2008.
Blogs
Tweets
Positivity Bias
References
83 of 83