Expert answer:critical essay

Solved by verified expert:Based on the reading and write a 300-word essay on it. This essay must be a critical reflection on the reading you have chosen.Note that the term critical as used here does not mean to exclusively find fault or to point out something negative in your chosen reading. Instead, the term criticism here is used in its broader sense. You are encouraged to explore both the merits and faults of the article and to position the article in its larger, scientific and social context.Here are a number of prompts to serve as guidelines for the writing of your thought essay:What are some of the strengths of the article? Why should these strengths be considered positively?What are some of the weaknesses of the article? Why do these weaknesses hurt the article?What is the overall premise of the article? Do you agree with it? Why or why not?In your opinion, why should this article exist? What purpose does it serve, and what questions or issues does it address?What is the context of this article? Why is it the logical thing to do given existing research in this area? What methodological or theoretical problems does it address? What are the broader implications of this article? Whether you believe this article makes a valid conclusion or is incorrect, what does this imply for future research in this area?These prompts are meant to help you get started, but we discourage you from simply going down these bullet points and answering these questions in the order they’re presented.During your reflection on the article, you may come up with questions or criticisms that you yourself are unable to address. If anything, coming up with difficult-to-answer questions is a sign that you’re doing things correctly. If such questions or concerns come up during your essay, we highly encourage you to add it to your submission. Good, open-ended questions will boost the quality of your submission and may be used as prompts for class discussions.
m10___ferwerda___personality_instagram.pdf

Unformatted Attachment Preview

Predicting Personality Traits with Instagram Pictures
Bruce Ferwerda
Markus Schedl
Marko Tkalcic
Department of Computational
Perception
Johannes Kepler University
Altenberger Str. 69
A-4040 Linz, Austria
Department of Computational
Perception
Johannes Kepler University
Altenberger Str. 69
A-4040 Linz, Austria
Department of Computational
Perception
Johannes Kepler University
Altenberger Str. 69
A-4040 Linz, Austria
bruce.ferwerda@jku.at
markus.schedl@jku.at
ABSTRACT
Instagram is a popular social networking application, which
allows photo-sharing and applying different photo filters to
adjust the appearance of a picture. By applying photo filters, users are able to create a style that they want to express
to their audience. In this study we tried to infer personality traits from the way users take pictures and apply filters
to them. To investigate this relationship, we conducted an
online survey where we asked participants to fill in a personality questionnaire, and grant us access to their Instagram
account through the Instagram API. Among 113 participants and 22,398 extracted Instagram pictures, we found
distinct picture features (e.g., hue, brightness, saturation)
that are related to personality traits. Our findings suggest
a relationship between personality traits and the way users
want to make their pictures look. This allow for new ways to
extract personality traits from social media trails, and new
ways to facilitate personalized systems.
Categories and Subject Descriptors
H.3.1 [Information storage and retrieval]:
Content
Analysis and Indexing; I.4.7 [Image processing and computer vision]: Feature Measurement
Keywords
Instagram, Personality, Photo filters, Picture features
1.
INTRODUCTION
Instagram is a popular mobile photo-sharing, and social
networking application with currently over 300 million active
users. 1 Instagram lets users easily connect with other social
networking platforms (e.g., Facebook, Twitter, Tumblr, and
Flickr) to share the taken pictures on, and enables users to
apply filters to their pictures. At this moment Instagram
offers 25 predefined photo filters that soften and color shift
1
https://instagram.com/press/ (accessed: 02/23/2015)
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full citation
on the first page. Copyrights for components of this work owned by others than the
author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or
republish, to post on servers or to redistribute to lists, requires prior specific permission
and/or a fee. Request permissions from permissions@acm.org.
EMPIRE ’15, September 16 – 20, 2015, Vienna, Austria
c 2015 Copyright held by the owner/author(s). Publication rights licensed to ACM.
ISBN 978-1-4503-3615-4/15/09. . . $15.00
DOI: http://dx.doi.org/10.1145/2809643.2809644
marko.tkalcic@jku.at
picture properties, for users to customize and modify their
pictures to create the desired visual style.
The ease with which a photo filter can be applied allow users to express a personal style and create a seeming
distinctiveness with the customized pictures. Through the
shared content and the way of applying filters, users are
able to reveal a lot about themselves to their social network. With that, the question arises: What do Instagram
pictures tell about the user? Or more specifically: What do
Instagram pictures say about the personality of the user?
Personality traits have shown to consist of cues to infer users’ behavior, preference, and taste (e.g., [3, 11, 13]).
Hence, knowing one’s personality can provide important cues
for systems to cater to a personalized user experience. It
can provide systems with estimations about user preferences
without the use of extensive questionnaires or observations.
There has been an increased interest in how to use personality in systems (e.g., [2, 4, 14]), and how to automatically
extract personality from online behavior trails (e.g., Facebook [1, 6, 9, 12], Twitter [5, 10]). In this work we join
the personality extraction research by specifically focusing
on the relationship between the picture features of an Instagram collection and the personality traits of the user.
Our work makes several contributions. We contribute to
personality research by showing relationships between personality traits and the visual style of users’ Instagram pictures. Additionally, we contribute to new ways to extract
personality from social media (i.e., Instagram). To the best
of our knowledge, we are the first to analyze (Instagram)
pictures in relation with personality traits.
We conducted an online survey where we asked participants to fill in the widely used, Big Five Inventory (BFI)
personality questionnaire, and grant us access to the content
of their Instagram account. We extracted 22,398 Instagram
pictures of 113 users, and analyzed them on several features
(e.g., hue, brightness, saturation). Distinct correlations were
found between personality traits and picture features.
In the remainder of the paper we will continue with related
work, materials, features, results, discussion, and conclusion.
2.
RELATED WORK
There is an increase of psychological research that investigate the relationship between personality and real-world behavior. Personality is known as an enduring factor and has
shown to be related to a person’s taste, preference, and interest (e.g., [3, 11, 13]). For example, Rawlings and Ciancarelli found relationships between personality traits and music
genre preferences [11], while Tkalcic et al. found relation-
ships between personality and classical music [13].
To categorize personality, several models have been developed. The five-factor model (FFM) is the most well known
and widely used one. It categorizes personality into five general dimensions that describe personality in terms of: openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism [8].
Based on psychological findings of personality in relation
to real-world behavior, there is an emergent interest in how
to use and implement these findings into applications (e.g., [2,
4, 14]). It can provide useful proxy measures for applications to cater to a more personalized service. For example,
Tkalcic et al. proposes to use personality traits to enhance
the nearest-neighborhood measurement for overcoming the
cold-start problem (i.e., recommending items to new users)
in recommender systems [14]. Ferwerda et al. provide a way
to use personality traits for adjusting the user interface of
music applications to fit user’s music browsing styles [4].
Personality traits have shown to consist of valuable information for personalized system, but hard to acquire (e.g.,
extensive, time-consuming questionnaires). Therefore, as
applications are increasingly interconnected (e.g., social media), research started to focus on how to extract personality
information from (online) behavior trails. Current research
on Facebook (e.g., [1, 6, 9, 12]) and Twitter (e.g., [5, 10])
have shown to consist of reliable cues to infer personality
traits from. With our work we add to the personality extraction research by showing the relationship between personality traits and picture features on Instagram.
3.
MATERIALS
To investigate the relationship between personality traits
and picture features, we asked participants to fill in the
44-item BFI personality questionnaire (5-point Likert scale;
Disagree strongly – Agree strongly [7]). The questionnaire
include questions that aggregate into the five basic personality traits of the FFM. Additionally, we asked participants to
grant us access to their Instagram account through the Instagram API in order to crawl their pictures. From hereon,
we define the picture-collection term as all the Instagram
pictures of a single user.
We recruited 126 participants through Amazon Mechanical Turk. Participation was restricted to those located in the
United States, and also to those with a very good reputation to avoid careless contributions. Several comprehensiontesting questions were used to filter out fake and careless
entries. The Mahalanobis distance was calculated to check
for outliers. This left us with 113 completed and valid responses. Age (18-64, median 30) and gender (54 male, 59
female) information indicated an adequate distribution.
4.
FEATURES
For each picture in a picture-collection that was crawled,
we extracted several features. The extracted features are
discussed below. Most of the features are color-based, some
are content-based. For color-based features we use the color
space that is most closely related to the human visual system, i.e., the Hue-Saturation-Value (HSV) color space [15].
Brightness. For each picture, we calculated the average
brightness and variance across all the pixels in the picture.
Pictures that have a high average brightness tend to be
bright, obviously. These features represent how light/dark a
picture is and how much contrast there is in the picture,
respectively. Pictures that have a high variance tend to
have both dark and light areas, whereas pictures with a low
variance tend to be equally bright across the picture. Furthermore, we divided the brightness axis into three equal
intervals and counted the share of pixels that fall into each
of these intervals (low/mid/high brightness). Pictures that
have a high value in the low brightness feature tend to be
darker, those that have a high value in the mid brightness
feature tend to have mostly neither dark nor bright areas,
while those pictures that have a high value in the high brightness feature tend to have lots of bright areas.
Saturation. We calculated the average saturation and the
variance for each picture. Pictures with low average saturation tend to be bleak, colorless, while pictures with high
saturation have more vivid colors. Pictures with a high saturation variance tend to have both bleak and vivid colors.
Here we also divided the saturation axis into three equal
intervals and calculated the share of pixels that fall into
each interval (low/mid/high saturation). pictures that have
a high value in the low saturation tend to have more bleak
colors, those with a high value in the mid saturation feature tend to have neither bleak nor vivid colors while those
pictures that have a high value in the high saturation feature tend to have vivid colors across most of the picture area.
Pleasure-Arousal-Dominance (PAD). As the filters on
Instagram intend to create a certain expression, we adopted
the PAD model of Valdez and Merhabian [16]. They created general rules of the expression of pleasure, arousal, and
dominance in a picture as a combination of brightness and
saturation levels:
1. Pleasure = .69 Brightness + .22 Saturation
2. Arousal = -.31 Brightness + .60 Saturation
3. Dominance = -.76 Brightness + .32 Saturation
Hue-related features We extracted features that represented the prevalent hues in pictures. We chose many features that represent various aspects of the hues. For each of
the basic colors (red, green, blue, yellow, orange, and violet)
we counted the share of pixels that fall into each color. As
the discrete color clustering of the hue dimension is nonlinear and subjective, we also divided the hue into 10 equal
intervals and calculated the share of pixels for each interval.
These intervals are hard to describe with subjective color
descriptions. Furthermore, we calculated the share of pixels
that fall into cold (violet, blue, green) and warm (yellow,
red, orange) colors.
Content-based features. Beside the color-centric features
we also performed picture content analysis. We counted the
number of faces and the number of people in each picture.
We used the standard Viola-Jones algorithm [17]. A manual
inspection of the Viola-Jones face detector results revealed
some false positives (e.g., a portrait within the picture) and
false negatives (e.g., some rotated and tilted faces). However, in general the users who tended to take pictures of
people (e.g., selfies) had a higher number of average number
of faces/people per picture than those users who tended to
take mostly still photographs.
5.
RESULTS
We crawled 22,398 pictures, and extracted all the features
per picture for each picture-collection. As the features in
the picture-collections show a symmetrical distribution, we
calculated mean values for each feature to create a measurement of central tendency. The mean values of the features
were used to calculate the correlation matrix (see Table 1).
Pearson’s correlation (r [-1,1]) is reported to indicate the
linear relationship between personality and picture features.
The correlation matrix shows several features related with
personality traits. We will discuss the results related to each
personality trait below. Besides significant correlations of
p<.05, we decided to report marginally significant results as well (i.e., significant levels of .1 >p>.05).
Openness to experience.
The openness factor correlates positively with the feature green, meaning that open
users tend to take pictures with a lot of green or applied a
filter to express more greenness. We observed a negative correlation with the feature brightness mean, which means that
open users tend to upload pictures that are low on brightness. This was further confirmed by the positive correlation
on brightness low, and the negative correlation on brightness
high. These correlations show that the pictures of open users
show more dark areas, and less bright areas.
Openness is also correlated with the saturation mean. This
indicate that the pictures of open users consist of more saturated, vivid colors. We also observed a positive correlation
with the feature saturation variance, which means that open
users upload pictures that have both vivid and bleak colors.
A marginally significant correlation was observed for the
warm/cold features. Pictures of open users contained less
warm colors (i.e., red, orange), but more cold colors (i.e.,
blue, green). Also, the pictures of open users tend to express
less pleasure, but more arousal and dominance. Additionally, their pictures consist of less faces and people.
Conscientiousness.
A marginally significant (positive)
correlation was found between the saturation variance feature and conscientiousness. This indicate that conscientious
users upload pictures consisting of bleak and vivid colors.
Extraversion. We mostly found marginally significant correlations with the picture features and the extraversion. Extraverts tend to upload pictures with less red and orange,
but with more green and blue tones. Additionally, their pictures tend to be darker (brightness low ), but tend to consist
of both vivid and bleak colors (saturation variance). Additionally, the emotion that the pictures of extraverts consist,
are low on pleasure, but high on dominance.
O
C
Red
-0.06
0.02
Green
0.17ˆ
0.14
Blue
-0.01
0
Yellow
0.01
0.04
Orange
-0.03
-0.07
Violet
0
-0.06
Bright.mean -0.25*
-0.1
Bright.var.
0.06
0
Bright.low
0.28**
0.09
Bright.mid
-0.09
0.06
Bright.high
-0.2ˆ
-0.12
Sat.mean
0.16ˆ
0.06
Sat.var.
0.2ˆˆ
0.16ˆ
Sat.low
-0.08
-0.02
Sat.mid
0.08
-0.09
Sat.high
0.13
0.1
Warm
-0.05ˆˆ -0.04
Cold
0.05ˆˆ
0.04
Pleasure
-0.19ˆˆ -0.08
Arousal
0.23*
0.09
Dominance
0.28**
0.11
# of faces
-0.16ˆ
0.03
# of people
-0.22ˆˆ -0.05
Note. ˆp<0.1, ˆˆp<.05, E A N -0.17ˆ -0.05 0.03 0.23ˆˆ 0.03 -0.12 0.17ˆ 0.02 -0.01 0.01 0.14 -0.07 -0.16ˆ -0.02 0.06 -0.09 -0.07 0.06 -0.19ˆ -0.07 0.22ˆ 0 -0.07 0.05 0.16ˆ -0.05 -0.16ˆ 0.04 0.15ˆ -0.06 -0.18ˆ -0.08 0.21ˆ 0.03 -0.04 0 0.19ˆˆ 0.1 -0.05 0.02 0.07 0.01 0.02 0.07 0.01 0.04 -0.01 0.01 -0.2 0 0.03 0.2 0 -0.03 -0.18ˆ -0.09 0.22ˆˆ 0.1 0 -0.08 0.17ˆ 0.05 -0.18ˆˆ 0.11 -0.11 -0.03 -0.07 -0.01 0.07 *p<.01, **p<.001. Table 1: Correlation Matrix of the picture features against the personality traits: (O)penness, (C)onscientiousness, (E)xtraversion, (A)greeableness, (N)euroticism. expression of the pictures of extraverts. They show to adjust their pictures to express more pleasure but less dominance. 6. PERSONALITY PREDICTION Given that we found significant correlations between picture features and personality traits, we explored personality prediction based on these features. We trained our predictive model with the radial basis function network classifier in Weka, with a 10-fold cross-validation. We report the rootmean-square error (RMSE) in Table 2. The RMSE of each personality trait relates to the [1,5] score scale. Personality Openness to experience Conscientiousness Extraversion Agreeableness Neuroticism RMSE 0.73 0.69 0.95 0.74 0.95 Agreeableness. A marginally significant correlation was found between agreeableness and the brightness mid feature. This means that the pictures of agreeable users do not show emphasized bright or dark areas, but are more in between. Table 2: Personality prediction with the root-meansquare error (RMSE). Neuroticism. A marginally significant correlation was found on brightness mean, brightness low, and brightness high. The positive correlation with brightness mean indicate that extravert users tend to upload pictures that are high on brightness. This is also reflected in the brightness low (negative correlation) and brightness high (positive correlation) features. Additionally, correlations were found in the emotion The RMSE values that we found are low and comparable with previous work on personality extraction from social media trails. For example, Quercia et al. [10] looked at the relationship between personality traits and Twitter usage and reported RSME scores of 0.69, 0.76, 0.88, 0.79, and 0.85, respectively for, openness to experience, conscientiousness, extraversion, agreeableness, and neuroticsm. Our results, as well as the results of prior work show that the Personality Openness to experience Conscientiousness Extraversion Agreeableness Neuroticism Picture properties Green, low brightness, high saturation, cold colors, few faces Saturated and unsaturated colors Green and blue tones, low brightness, saturated and unsaturated colors Few dark and bright areas High brightness Table 3: Picture properties in relation to personality traits. The properties apply for the pictures of users who score high in the respective personality trait. most difficult traits to predict are the extraversion and the neuroticism personality traits. 7. DISCUSSION We found Instagram picture features to be correlated with personality. Results show that the most strongly significant correlations are found in the openness to experience personality trait. Although we found weaker significant levels for the other personality traits, we were still able to find distinct correlations. See Table 3 for a summary of our findings. Based on the found correlations, we also explored the prediction of the personality traits based on the picture features. Compared with the findings of prior work (i.e., [10]), we were able to find similar results and patterns. The most successful personality traits to predict are openness to experience, conscientiousness, and agreeableness, whereas the more difficult traits are extraversion and neuroticism. 8. LIMITATIONS AND FUTURE WORK Our study contains limitations that need to be considered. Although we were able to obtain a fair amount of Instagram pictures (n=22,398), our personality measurement was limited to 113 participants. Given that we only had personality information of 113 participants to find relationships with picture features, we decided to give attention to the marginally significant results as well (i.e., significant levels of .1 >p>.05). A bigger sample size to assess personality traits should provide more conclusive results about the
marginally significant effects that …
Purchase answer to see full
attachment

How it works

1. Paste your instructions in the instructions box. You can also attach an instructions file
2. Select the writer category, deadline, education level and review the instructions 
3. Make a payment for the order to be assignment to a writer
4.  Download the paper after the writer uploads it 

Will the writer plagiarize my essay?

You will get a plagiarism-free paper and you can get an originality report upon request.

Is this service safe?

All the personal information is confidential and we have 100% safe payment methods. We also guarantee good grades