VIRUSES structure and classification ppt by Dr.Prince C P
AINL 2016: Panicheva, Ledovaya
1. Lexical, Morphological and
Semantic Correlates
of the Dark Triad Personality Traits
in Russian Facebook Texts
Polina Panicheva
ppolin86@gmail.com
Olga Bogolyubova
Yanina Ledovaya
St. Petersburg State
University
Clarkson University
Supported by SPBU research grant 8.38.351.2015:
"A cross-cultural study of the markers of stress,
health and well-being in social networks".
2. Linguistic correlates of the Dark Traits
◎Motivation
○ English background (Sumner 2012), (Schwartz
2013)
○ Dark traits
○ Russian LIWC
◎Data collection
○ Questionnaire
◎Statistics and interpretation
○ Text volume features
○ Lexical
○ Morphological
○ Content
◎Significance issues and future work
3. Background:
Author Profiling in English
◎Personality profiling
○ Big Five
○ Dark traits (Sumner 2012), Psychopathy (Hancock
2013)
○ Mental health issues: PTSD, depression (Harman
2014)
◎Data
○ Twitter, Facebook
○ Self-report narratives
◎Word-count statistics
○ Linguistic Inquiry and Word Count
5. The Dark Traits
◎Narcissism, Machiavellianism, Psychopathy
○ Related but distinct, lack of empathy – common feature
○ Subclinical, continuous scales
◎The Short Dark Trait Survey by Jones 2014
○ Russian adaptation by Egorova (2014)
○ Likert 5-item agreement scales (9 q. per trait)
6. The Dark Traits
◎Narcissism
○ self-focus, grandiosity
○ ‘People see me as a natural leader’
◎Machiavellianism
○ manipulating
○ ‘It’s not wise to tell your secrets‘
◎Psychopathy
○ impulsiveness, aggression and asocial behavior
○ ‘People often say I’m out of control’
7. The Dark Traits: structure
Narcissism Machiavellianism Psychopathy
Motivation Ego-identity
goals
Instrumental goals,
material gain
Temporal
focus
Distant future,
strategic
Immediate
Lack of empathy,
interpersonal manipulation
Facets Exploitative/
entitlement
Leadership/
authority
Cynical worldview
Machiavellian tactics
Manipulation
Callous affect
Erratic lifestyle
Antisocial
behavior
(Jones 2014)
8. Russian Linguistic Inquiry and Word Count
◎English LIWC – dictionary (Pennebaker 2001)
○ ~80 categories
◎Translated to Russian directly
○ Preserving the English-based morphological and
semantic structure
○ E.g content/function words – doesn’t hold in Russian
○ Different language type: affixation & fusion
◎Why impose the foreign top-down category
structure?
◎Induce native language-specific categories
in a bottom-up way
9. Task
◎Obtain significant linguistic correlates
of the Dark Traits
○ Bottom-up lexical approach
○ Morphological analysis-based features
○ Derive content features by semantic clustering
11. Text length features
Text feature Na Ma Ps
Sentence length 0.022 -0.057 0.006
Post length, sentences 0.054 -0.109 -0.04
Post length, tokens 0.045 -0.101 -0.02
p < 0.05 p < 0.01
Machiavellianism: guard personal image, avoid oversharing (Rauthmann
2012)
Narcissism: expansive, attract attention (Raskin 1988)
12. Morphological and lexical features
◎PyMorphy analyzer [Korobov 2015]
◎Morphological gramemes:
○ POS
○ person, number (in verb and pronoun)
○ verb modality: tense, voice, mood, reflexivity
○ NE: names, organization, geo-location, abbr
○ adjective features: short/full, qualitative, superlative
○ possessive pronouns
○ style:
◉vernacular, slang
13. Semantic clustering
◎Words in >5 authors’ texts -> 3.7K words
◎k-means over a RNC-trained word-embeddings
semantic space, [Panicheva 2016]
◎182 thematic clusters
○ fully automatic
○ interpretable
Cluster Contents
Authority выборы дума комитет
парламент рада рф
election parliament
committee Rada Russia
Friend близкий друг незнакомый
приятель родные
folks friend stranger pal
relative
Female_na
me
алиса анна вера виктория
елена ирина мария
Alice Anna Vera Victoria
Elena Irina Maria
Passion безумие веселье желание
любовь страсть
madness joy desire love
passion
14. Results: Narcissism
Feature Lexical Clustering Morphology
Self focus I, my Perfection,
High_low
1st
pers sing,
1st
pers poss
Social
involveme
nt
you, gratitude,
thank, company,
invitation
Appeal,
Take_give, Want,
Feeling
imperative,
2nd
pers plur
Positive
emotion
honorable,
important
Awful, Passion,
Pos_quality,
Perfection
Goal focus become, vocation,
skill, solution
Reasoning, Goal,
Achievement
Speech
involvement
interj, punct,
prons, verbs
p < 0.05
Facets of Narcissism: Leadership/Authority, Exploitativeness/Entitlement
(Jones 2014)
15. Results: Machiavellianism
Feature Lexical Clustering Morphology
Social
involvement
friend, we Affirm, Friend, Feeling_vb,
Tender_adj, male/female
names
1st
pers plur,
names, 2nd
, 3rd
pers
Positive affect love, heart Wellbeing, Impress
Mental
processing
Faith, Religion, Perception,
Sensation, Appearance
Politics russia, isis,
president
Personal
detachment
prons, verbs
Common daily
topics
Neg_action, Trouble,
Face_part, Body_situation,
Number, Age, Being
Facets of Machiavellianism: Cynicism, Manipulative Tactics
(Jones 2014)
17. Significance discussion
◎Significance filtering:
○ Bonferroni correction – too strict
○ False Rate Discovery
◎Very modest correlation values
○ r ~ 0.1, but p < 0.01!
○ Sumner (2012), Schwartz (2013) report similar values
for English
○ Much more data needed, pref. more texts p.person
○ Are r’s supposed to be very high?
◎Prediction: just above baseline (cf. Sumner)
18. Future work
◎More data
○ Identify meaningful n-grams
○ Technical features: TTR, POS ratio
◎Prediction experiment – are high results
possible in this setting?
◎Identify ‘repost language’ correlations
◎Covariates: investigate difference in language
caused by sex, age, …
◎Project goal: language of well-being and
online aggression
◎More remote goal: compare Russian VS US
well-being and Dark traits
20. References
◎ M. Egorova and M. Sitnikova, “The dark triad,” psystudy.ru, 2014
◎ J. T. Hancock, M. T. Woodworth, and S. Porter, “Hungry like the wolf: A word-
pattern analysis of the language of psychopaths,” 2013.
◎ G. Harman, M. Coppersmith, and C. Dredze, “Quantifying mental health signals in
twitter,” ACL 2014, p. 51, 2014.
◎ D. N. Jones and D. L. Paulhus, “Introducing the short dark triad (sd3) a brief
measure of dark personality traits,” Assessment, 2014
◎ M. Korobov, “Morphological analyzer and generator for russian and ukrainian
languages,” AIST, 2015
◎ P. Panicheva, Y. Ledovaya, and O. Bogoliubova, “Revealing interpetable content
correlates of the dark triad personality traits,” RuSSIR, 2016
◎ J. W. Pennebaker, M. E. Francis, and R. J. Booth, “Linguistic inquiry and word
count: Liwc 2001,”
◎ R. Raskin and H. Terry, “A principal-components analysis of the narcissistic
personality inventory and further evidence of its construct validity,” 1988.
◎ J. F. Rauthmann and G. P. Kolar, “How “dark” are the dark triad traits? Examining
the perceived darkness of narcissism, machiavellianism, and psychopathy,” 2012
◎ H. A. Schwartz, J. C. Eichstaedt, M. L. Kern, L. Dziurzynski, S. M. Ramones, M.
Agrawal, A. Shah, M. Kosinski, D. Stillwell, M. E. Seligman et al., “Personality,
gender, and age in the language of social media: The open-vocabulary approach,”
2013.
◎ C. Sumner, A. Byers, R. Boochever, and G. Park, “Predicting dark triad personality
traits from twitter usage and a linguistic analysis of tweets,” ICMLA, 2012