�ݺ�ߣ

Hashtags on
lnstagram
Content Description on a Mobile Image Sharing Service
Isabelle Dorsch
D e p a r t m e n t o f I n f o r m a t i o n S c i e n c e
H e i n r i c h H e i n e U n i v e r s i t y , D ü s s e l d o r f , G e r m a n y
1 2 t h S e p t e m b e r 2 0 1 8
Ta l k , I S I 5 3 0 2 K n o w l e d g e O r g a n i z a t i o n S e m i n a r

Hashtags on Instagram | Dorsch
Change of Media and Celebrities
2
“A picture is worth a thousand words”

Hashtags on Instagram | Dorsch 2
How do users index a
picture on Instagram with
at most 30 hashtags?

Indexing of Pictures
2 approaches
Features extraction
e.g. color, shape, texture
Textual description
Content-based Concept-based
(Lancaster, 2003; Rasmussen, 1997)

Folksonomy
Term for the free allocation of keywords (called tags)
(Smith, 2004, Vander Wal, 2007)
Total quantity of all assigned tags in an information service
(Peters, 2009)
Example: flickr
Folksonomy”  “folk” and “taxonomy”
https://www.flickr.com/

Hashtags
Twitter: https://twitter.com/chrismessina/status/223115412
ChrisMessina as
“creator” of the hashtag

Instagram
https://www.instagram.com
Main purpose:
Sharing photos and videos
Launched in October 2010
 With 25,000 signed-up users
 Only for iOS-Users
Since April 2012:
Also for android users
Now
 1B+ monthly active users
 500M+ daily active users
Instagram history: https://www.instagram.com/p/BaD1n2gjj1a/?hl=de&tagged=cat

Instagram
 Posted architecture picture by kajaf on Instagram

RQ1: Are there any differences in relative frequencies of hashtags in the
picture categories?
RQ2: Given a picture category, what is the distribution of hashtag
categories; and given a hashtag category, what is the distribution of
picture categories?
RQ3: Is there any association between image categories and hashtag
categories?
Research Questions

Research Model
RQ1
RQ2+3
Instagram
User Behavior
Hashtag
Categories
Picture
Categories
Hashtag
Counts
Post
Picture with
Hashtags

Methods Overview
Analysis of 14,649 hashtags for in total 1,000 Instagram pictures
+ Additional pretest with 50 Instagram pictures
Content Analysis: Development of 2 codebooks
Codebook categories: Picture categories & hashtag categories
Data collection & coding process

 Selection of 10 picture categories based on Hu et al. (2014)
 Development of a picture category codebook
Hu et al. (2014) 8 popular Instagram photo categories:
Methods
FoodActivity
Captioned
Photo
Fashion
SelfieFriends Gadget Pet

 Selection of 10 picture categories based on Hu et al. (2014)
 Development of a picture category codebook
10 picture categories used for this study:
Methods
FashionActivity Art
Captioned
Photo
SelfieFood Landscape Pet
Architecture
Friends
Gadget
Why these 10 categories?
Reason: Pretest

Methods
Examples for the used picture categories
Activity

Methods
Architecture

Methods
Art

Methods
Captioned
Photo

Methods
Fashion

Methods
Food

Methods
Friends

Methods
Landscape

Methods
Pet

Methods
Selfie

 Selection of 7 hashtag categories
 Development of a hashtag category codebook
7 hashtag categories used for this study:
Methods
Isness
Content-
relatedness
Emotiveness Fakeness
“Insta”-Tags Performativeness
Sentences

Content-relatedness
 Everything a picture directly or abstractly depicts
 Based on definitions of aboutness and ofness
(Shatford, 1986, Shatford Layne, 1994)
Hashtag examples:
If they are depicted:
#food, #person, #pet, #yellow, #colorful, #activity, #eyes, friends,
#bestfriends, #paris, #blackcats #polishgirl
Methods

Emotiveness
 Includes all hashtags which contain primary emotions or feelings:
 Love
 Happiness
 Fun
 Surprise
 Aspiration
 Besides, consideration of all possible manifestations of emotions a user
could formulate
Hashtag examples:
#love, #fun, #happy, #sad, #happygirl, #doglover
Methods
(Siebenlist, 2013)
 Sadness
 Anger
 Disgust
 Fear
 Shame

Sentences
 Includes all hashtags which are formulated as a whole sentences
 Sentences represent content
 Why not include them into the category Content-relatedness?
Methods

Excursion: Indexing and
summarizing in knowledge
representation
Methods
(Stock & Stock, 2013)

Sentences
 Includes all hashtags which are formulated as a whole sentences
 Category Content-relatedness: Only terms and phrases (Indexing)
 Category Sentences: Information condensation (Abstracts)
 According to this, a separation was made
Hashtag examples:
#WIWT  #WhatIWoreToday, #IFeelRealyBadToday
Methods

Isness (Ingwersen, 2002)
 All non-topical features of a document; for pictures:
 Technical aspects of the photograph
(e.g., camera type, length of exposure, aperture) (Stock & Stock, 2013)
 (User)name of the photographer
 Date
 Location
 Type of the picture (e.g. selfie)
 Requirement: All of them not depicted on the picture posting
 Special consideration for art pictures and people
Hashtag examples:
#is_made_with_XY, #picoftheday, #photooftheday, #foodie, #photo, #nofilter,
#foodporn, #foodgasm, #blogger, #sunday, #winter , #selfietime
Methods

Performativeness (Peters & Stock, 2007)
 Includes all hashtags for action requests or challenges on Instagram or
other platforms, e.g.:
 Contests like “participatory hashtag projects” on Instagram
 Account features
 Participation through tagging a picture with a certain hashtag
 General performative hashtags were not categorized (e.g. #petoftheday)
Hashtag examples:
#like4like, #likethis, #eatrealfood, #icebucketchallenge, #discover, #followme
Methods

Fakeness
 Deliberately incorrectly assigned hashtags
 Includes all hashtags that are not valid for the respective image or posting
description in any way
 Except
 Hashtag provides room for interpretation  no clear false/true statement possible
 Typos
Hashtag examples:
#nature, #mountains, #beautiful, #latergram,
#naturelovers, #switzerland,
#cityscape, #selfie, #instanature
Methods

“Insta”-Tags
 Contains as hashtag component “Insta,” “gram,” Instagram or a shortcut
regarding to the previously stated expressions
 Platform-specific phenomenon
Hashtag examples:
#instadaily, #instapic, #instaart, #instacat, #webstagram, #webstagramers,
#Igers, #naturegram, #latergram
Methods

Methods
Content-relatedness
#cat
Emotiveness
#happy
“Insta”-Tags
#catsofinstagram
Isness
#photooftheday
Performativeness
#excellent_cats
Sentences
#ilovemycat
Fakeness
e.g. #dog
Real world example for the used hashtag categories

Codebooks
1. Instagram picture category codebook
2. Instagram hashtag category codebook
Methods
Purpose
Category overview
STRUCTURE
Common coding rules
Full code descriptions
References

Methods
Example: Full code description
picture category codebook

Methods
Example: Full code description
hashtag category codebook

Data collection
 November 2016 – January 2017
 Pretest: 50 pictures (manually)
 Analysis: 1,000 pictures (automatically + manually)
 Picture selection according to top Instagram hashtags of their category
(e.g. Category Food  #food 198m postings; #foods 8.5m postings)
Coding process
 2 coders, „4-eyes principle“
Methods

RQ1: Frequencies of hashtags
Results

RQ2: Relative frequency of hashtag categories by picture categories
Results
Relative frequency of hashtag categories by picture categories (N=1,000 posts; 100 posts per picture category)

RQ3: Association between image and hashtag categories
 Chi-square test of independence
 A statistical association between hashtag categories and hashtag pictures
exists
 Picture categories affect hashtag categories
Results

Evaluation of Instagram users‘ tagging behavior in a wide field:
Discussion
On average 15 hashtags per picture
Content-relatedness + Isness hashtags were assigned the most
Less Emotiveness + Sentences hashtags
„Insta“-Tags  often assigned to the category Pet
Fakeness  often assigned to the category Captioned Photo
Small statistical association
RQ1
RQ2
RQ3

Limitations:
 Limited access to the data
Outlook:
 Users‘ intention?
 Analysis of Videos, Stories, etc.
 Gender-specific differences
 …
Discussion

Please feel free to ask any
questions, etc. via Twitter
@bezwitschernd
i s a b e l l e . d o r s c h @ h h u . d e
T h a n k y o u f o r
y o u r a t t e n t i o n !
The full article:
Dorsch, I. (2018). Content Description on a Mobile Image Sharing Service: Hashtags on Instagram. Journal of Information Science Theory and Practice 6(2), 46-61.

If not directly indicated under the picture:
�ݺ�ߣs 2, 4 from: Alexander Schöch
�ݺ�ߣs 11, 25, 26, 29-31, 38, 42, 44 from: https://www.iconfinder.com/iconsets/streamline-icon-set-free-pack
author website: http://www.webalys.com/minicons/
�ݺ�ߣ 38 from: https://www.iconfinder.com/iconsets/user-avatars-1
author website: https://usersinsights.com/
�ݺ�ߣ 42 from: https://www.iconfinder.com/icons/2976459/bird_fly_pet_sparrow_icon
�ݺ�ߣs 44 from: https://twitter.com
Instagram Screenshots: https://www.instagram.com (The author of the posting is indicated in the respective picture)
Further free pictures from: https://pixabay.com
Picture Credits

Hu, Y., Manikonda, L., & Kambhampati, S. (2014). What we Instagram: A first analysis of Instagram photo content and user types. In Proceedings of the 8th International
Conference on Weblogs and Social Media, ICWSM 2014 (pp. 595-598). Palo Alto, CA: AAAI Press.
Ingwersen, P. (2002). Cognitive perspectives of document representation. In H. Bruce, R. Fidel, P. Ingwersen, & P. Vakkari (Eds.), Proceedings of the 4th International
Conference on Conceptions of Library and Information Science (pp. 285-300). Greenwood Village, CO: Libraries Unlimited.
Instagram. (2010). Instagram launches. Retrieved Apr 14, 2017 from https://instagram-press.com/blog/2010/10/06/instagram-launches-2/
Instagram. (2018). Instagram statistics. Retrieved Sep 6, 2018 from https://instagram-press.com/our-story
Lancaster, F. W. (2003). Indexing and abstracting in theory and practice. Champaign, IL: University of Illinois, Graduate School of Library and Information Science.
Messina, C. (2007). how do you feel about using # (pound) for groups. As in #barcamp [msg]? @chrismessina (Twitter). Retrieved May 2, 2017 from
https://twitter.com/chrismessina/status/223115412.
Panofsky, E. (1955). Meaning in the visual arts. Garden City, NY: Doubleday Anchor Books.
Peters, I. (2009). Folksonomies: Indexing and retrieval in Web 2.0. Berlin: De Gruyter Saur.
Peters, I., & Stock, W. G. (2007). Folksonomy and information retrieval. Proceedings of the American for Information Science and Technology, 44(1), 1-28.
Rasmussen, E. M. (1997). Indexing images. Annual Review of Information Science and Technology (ARIST), 32, 169-196.
Shatford, S. (1986). Analyzing the subject of a picture: a theoretical approach. Cataloging & Classification Quarterly, 6(3), 39-62.
Shatford Layne, S. (1994). Some issues in the indexing of images. Journal of the American Society for Information Science, 45(8), 583-588.
Siebenlist, T. (2013). Emotionale suche. In D. Lewandowski (Ed.), Handbuch Internet-Suchmaschinen (pp. 299- 327). Heidelberg: Akademische Verlagsgesellschaft.
Smith, G. (2004). Folksonomy: Social classification. Retrieved May 9, 2018 from http://atomiq.org/%0Aarchives/2004/08/ folksonomy_social_classification.html
Stock, W. G., & Stock, M. (2013). Handbook of information science. Berlin: De Gruyter Saur.
Vander Wal, T. (2007). Folksonomy coinage and definition. Retrieved Apr 16, 2017 from http://www.vanderwal. net/folksonomy.html
References

�ݺ�ߣ

Hashtags on Instagram

Recommended

More Related Content

Recently uploaded (20)

Featured (20)

Hashtags on Instagram