Upload
job-kir
View
227
Download
0
Embed Size (px)
Citation preview
8/13/2019 Gill Oberland Er 2
1/16
Taking CareOf the Linguistic Features of Extraversion
Alastair Gill and Jon Oberlander
University of Edinburgh
8/13/2019 Gill Oberland Er 2
2/16
12/2/2013 Gill & Oberlander 2
Jim Carrey: Introvert? Albert Einstein: Extravert?
Which are they ... and where are you?
Extravert or Introvert?
8/13/2019 Gill Oberland Er 2
3/16
12/2/2013 Gill & Oberlander 4
Personality meets HCI
Reeves and Nass had subjects solve a problem and thendiscuss solution with the computer (via a textual interface)
Language variables manipulated to provide different systempersonalities:
Dominant:
Always goes first
Strong language Uses assertions and commands
Indicates high confidence
Submissive:
Goes second
Unassertive language Uses questions and suggestions
Indicates low confidence
8/13/2019 Gill Oberland Er 2
4/16
12/2/2013 Gill & Oberlander 5
Eysenck's Three Factor PEN model
Extraversion social interest and positive affect "arousal level in the cortex"
(lower levels = more extravert)
Neuroticism Response to stressors and negative affect
"activation thresholds in the limbic system"(lower thresholds = more neurotic)
Psychoticism Aggressiveness, individuality "testosterone levels?"
(higher levels = more psychotic) What are the implications for language?
8/13/2019 Gill Oberland Er 2
5/16
12/2/2013 Gill & Oberlander 6
Personality and Language ... so far
Work has focussed mainly on Extraversion in speech
What has this found? Talk more in discussions (Carment et al., 1965)
Distinguished by speech cues (Scherer, 1979)
Specific (eg. syntactic category) and less specific (style)features related to lexical choice proposed (Furnham, 1990)
LIWC content analysis program used with transcriptions to look atwarmth and dominance facets of Extraversion (Berry et al., 1997)
8/13/2019 Gill Oberland Er 2
6/16
12/2/2013 Gill & Oberlander 7
Textual Personality
Relatively little work on written linguistic personality Written diary entries analysed using LIWC (Pennebaker and King,
1999)
Can we replicate Pennebaker and King's findings?
How else is personality embodied in text Which linguistic features are most important for personality?
What can more sophisticated computational linguistics techniquesreveal?
8/13/2019 Gill Oberland Er 2
7/16
12/2/2013 Gill & Oberlander 8
Our personality corpus
105 subjects generating two texts each
Each completed the Eysenck Personality Questionnaire
Each then composed two emails: "To a good friend whom they hadn't seen for quite some time"
One concerned past activities over the previous week
The other concerned planned activities over the next week. Each message took around 10 minutes to compose and submit by
HTML form.
The resulting 210 texts contain 65,000 words.
8/13/2019 Gill Oberland Er 2
8/16
12/2/2013 Gill & Oberlander 9
Experiment 1: replication and text analysis
Factor analysis Selection of LIWC items for principal components analysis used the same
criteria as Pennebaker and King
Four factors derived which are essentially similar to theirs, with minorvariations in factor loadings
Multiple regression analysis Variables which showing a small correlation with personality type and topicindependence were entered for:
LIWC
MRC Psycholinguistic Database (used to compute mean scores for verbalfrequency, written frequency, concreteness, age of acquisition, etc.)
8/13/2019 Gill Oberland Er 2
9/16
12/2/2013 Gill & Oberlander 10
Interpretation
Extraverts do write more This echoes earlier findings on speech
And they write more loosely (with fewer exact numberexpressions) For eg. several vs five
Supports Furnhams hypothesis And we also find they write less concretely
Eg. Furniture vs chair
But: The variance explained is not as great as for P and N
Correlations not as strong as found by Pennebaker and King
8/13/2019 Gill Oberland Er 2
10/16
12/2/2013 Gill & Oberlander 11
Experiment 2: NLP techniques
Use simple bigram techniques to take more of a word's context into account
Select high-E and low-E subcorpora by taking texts fromsubjects with E-score > 1 s.d from mean (cf. Dewaele andPavlenko, 2002): 21 High E versus 17 Low E
12,000 words versus 8,000
Generate bigram profiles ranked by Dunning's log-likelihoodstatistic (top 50 bigrams with frequency >= 2, p < 0.001).
Calculate relative frequency ratios (Damerau 1993) for bigrams
common to both subcorpora
8/13/2019 Gill Oberland Er 2
11/16
12/2/2013 Gill & Oberlander 12
On the surface
The gross features are perhaps the most intuitive in theirrepresentation of the Extraverts or Introverts.
For example, [ hi], the marker followed byhi, was unique to Extravert texts; message-initial hi.
By contrast the more formal [ hello] was found solelyin Introvert texts
Use of punctuation also differs between the two groups: Extraverts preferring multiple exclamation marks [! !], and solely
using multiple full stops [. .] as in the elliptical (...)
features of informal style, and looser use of language.
8/13/2019 Gill Oberland Er 2
12/16
12/2/2013 Gill & Oberlander 13
Quantification
Introverts tend to show preference for a greater use ofquantifiers [a lot], [a few] and uniquely [all the], [one of], [lots of] and [loads
of]
Extraverts show a preference for [a bit] and uniquely use[couple of].
Not only does this demonstrate an Extravert tendency to belooser and less specific, it also apparently reveals a tendencytowards exaggeration on the part of the Introvert.
8/13/2019 Gill Oberland Er 2
13/16
12/2/2013 Gill & Oberlander 14
Valence
Bigrams containing negations were used significantly only byIntroverts, as in [i dont] and [dont know] (indeed [i dont] is the bigram with most frequent use of i)
Similarly, the Extravert preferences suggest a more positive,relaxed disposition: [looking forward] and [forward to]
(presumably as in looking forward to)
[a good]
[catch up]
[take care]
8/13/2019 Gill Oberland Er 2
14/16
12/2/2013 Gill & Oberlander 15
Ability and Modality
Personal views on capability are suggested by the different
collocations with infinitival to. Extraverts demonstrate ability with want, need, and able (to) Introverts more timidly and tentatively are [trying to] or [going to]
Similarly, collocations with the verb be show a distinction inuse of modal auxiliaries which has an effect on the projection
of certainty. For example, Introverts uniquely use the weaker [should be] Extraverts prefer the stronger predictive [will be], and
contracted form [ill be] (i will be).
8/13/2019 Gill Oberland Er 2
15/16
12/2/2013 Gill & Oberlander 16
What about the P and N dimensions? Bigrams in 3D
High-E use ... and !! Both associated with High-N;
But one goes with High-P; the other with Low-P
High-E use [take care] This is Low-N, Low-P
Low-E use [, but] and [, because] One is High-N, High-P; The other is Mid-N, Mid-P.
8/13/2019 Gill Oberland Er 2
16/16
12/2/2013 Gill & Oberlander 17
Take care!
Summary
CASA tells us that linguistic personality might matter
Simple techniques can confirm known linguistic features ofextraversion and uncover new ones
Applications: Interface agents
Personality Language Checker Future work:
Test sensitivity of readers to personality features
Investigate feature generalisability