Inferring social media users' demographics from profile pictures: A Face++ analysis on twitter users

Soon Gyo Jung, Jisun An, Haewoon Kwak, Joni Salminen, Bernard James Jansen

Research output: Contribution to journalConference article

2 Citations (Scopus)

Abstract

In this research, we evaluate the applicability of using facial recognition of social media account profile pictures to infer the demographic attributes of gender, race, and age of the account owners leveraging a commercial and well-known image service, specifically Face++. Our goal is to determine the feasibility of this approach for actual system implementation. Using a dataset of approximately 10,000 Twitter profile pictures, we use Face++ to classify this set of images for gender, race, and age. We determine that about 30% of these profile pictures contain identifiable images of people using the current state-of-the-art automated means. We then employ human evaluations to manually tag both the set of images that were determined to contain faces and the set that was determined not to contain faces, comparing the results to Face++. Of the thirty percent that Face++ identified as containing a face, about 80% are more likely than not the account holder based on our manual classification, with a variety of issues in the remaining 20%. Of the images that Face++ was unable to detect a face, we isolate a variety of likely issues preventing this detection, when a face actually appeared in the image. Overall, we find the applicability of automatic facial recognition to infer demographics for system development to be problematic, despite the reported high accuracy achieved for image test collections.

Original languageEnglish (US)
Pages (from-to)140-145
Number of pages6
JournalProceedings of the International Conference on Electronic Business (ICEB)
Volume2017-December
StatePublished - Jan 1 2017
Event17th International Conference on Electronic Business: Smart Cities, ICEB 2017 - Al Barsha, Dubai, United Arab Emirates
Duration: Dec 4 2017Dec 8 2017

Fingerprint

Demographics
Social media
Twitter
Evaluation
System implementation
System development
Owners
Test collections
Tag

All Science Journal Classification (ASJC) codes

  • Business, Management and Accounting(all)
  • Computer Science(all)

Cite this

@article{01e9d982e9584d44b464172d226a4b2d,
title = "Inferring social media users' demographics from profile pictures: A Face++ analysis on twitter users",
abstract = "In this research, we evaluate the applicability of using facial recognition of social media account profile pictures to infer the demographic attributes of gender, race, and age of the account owners leveraging a commercial and well-known image service, specifically Face++. Our goal is to determine the feasibility of this approach for actual system implementation. Using a dataset of approximately 10,000 Twitter profile pictures, we use Face++ to classify this set of images for gender, race, and age. We determine that about 30{\%} of these profile pictures contain identifiable images of people using the current state-of-the-art automated means. We then employ human evaluations to manually tag both the set of images that were determined to contain faces and the set that was determined not to contain faces, comparing the results to Face++. Of the thirty percent that Face++ identified as containing a face, about 80{\%} are more likely than not the account holder based on our manual classification, with a variety of issues in the remaining 20{\%}. Of the images that Face++ was unable to detect a face, we isolate a variety of likely issues preventing this detection, when a face actually appeared in the image. Overall, we find the applicability of automatic facial recognition to infer demographics for system development to be problematic, despite the reported high accuracy achieved for image test collections.",
author = "Jung, {Soon Gyo} and Jisun An and Haewoon Kwak and Joni Salminen and Jansen, {Bernard James}",
year = "2017",
month = "1",
day = "1",
language = "English (US)",
volume = "2017-December",
pages = "140--145",
journal = "Proceedings of the International Conference on Electronic Business (ICEB)",
issn = "1683-0040",

}

Inferring social media users' demographics from profile pictures : A Face++ analysis on twitter users. / Jung, Soon Gyo; An, Jisun; Kwak, Haewoon; Salminen, Joni; Jansen, Bernard James.

In: Proceedings of the International Conference on Electronic Business (ICEB), Vol. 2017-December, 01.01.2017, p. 140-145.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Inferring social media users' demographics from profile pictures

T2 - A Face++ analysis on twitter users

AU - Jung, Soon Gyo

AU - An, Jisun

AU - Kwak, Haewoon

AU - Salminen, Joni

AU - Jansen, Bernard James

PY - 2017/1/1

Y1 - 2017/1/1

N2 - In this research, we evaluate the applicability of using facial recognition of social media account profile pictures to infer the demographic attributes of gender, race, and age of the account owners leveraging a commercial and well-known image service, specifically Face++. Our goal is to determine the feasibility of this approach for actual system implementation. Using a dataset of approximately 10,000 Twitter profile pictures, we use Face++ to classify this set of images for gender, race, and age. We determine that about 30% of these profile pictures contain identifiable images of people using the current state-of-the-art automated means. We then employ human evaluations to manually tag both the set of images that were determined to contain faces and the set that was determined not to contain faces, comparing the results to Face++. Of the thirty percent that Face++ identified as containing a face, about 80% are more likely than not the account holder based on our manual classification, with a variety of issues in the remaining 20%. Of the images that Face++ was unable to detect a face, we isolate a variety of likely issues preventing this detection, when a face actually appeared in the image. Overall, we find the applicability of automatic facial recognition to infer demographics for system development to be problematic, despite the reported high accuracy achieved for image test collections.

AB - In this research, we evaluate the applicability of using facial recognition of social media account profile pictures to infer the demographic attributes of gender, race, and age of the account owners leveraging a commercial and well-known image service, specifically Face++. Our goal is to determine the feasibility of this approach for actual system implementation. Using a dataset of approximately 10,000 Twitter profile pictures, we use Face++ to classify this set of images for gender, race, and age. We determine that about 30% of these profile pictures contain identifiable images of people using the current state-of-the-art automated means. We then employ human evaluations to manually tag both the set of images that were determined to contain faces and the set that was determined not to contain faces, comparing the results to Face++. Of the thirty percent that Face++ identified as containing a face, about 80% are more likely than not the account holder based on our manual classification, with a variety of issues in the remaining 20%. Of the images that Face++ was unable to detect a face, we isolate a variety of likely issues preventing this detection, when a face actually appeared in the image. Overall, we find the applicability of automatic facial recognition to infer demographics for system development to be problematic, despite the reported high accuracy achieved for image test collections.

UR - http://www.scopus.com/inward/record.url?scp=85050642516&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050642516&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85050642516

VL - 2017-December

SP - 140

EP - 145

JO - Proceedings of the International Conference on Electronic Business (ICEB)

JF - Proceedings of the International Conference on Electronic Business (ICEB)

SN - 1683-0040

ER -