The impact of visual attributes on online image diffusion

Luam Totti, Felipe Costa, Sandra Avila, Eduardo Valle, Wagner Meira, Virgílio Almeida

Research output: Chapter in Book/Report/Conference proceedingConference contribution

25 Citations (Scopus)

Abstract

Little is known on how visual content affects the popularity on social networks, despite images being now ubiquitous on the Web, and currently accounting for a considerable frac- tion of all content shared. Existing art on image sharing fo- cuses mainly on non-visual attributes. In this work we take a complementary approach, and investigate resharing from a mainly visual perspective. Two sets of visual features are proposed, encoding both aesthetical properties (brightness, contrast, sharpness, etc.), and semantical content (concepts represented by the images). We collected data from a large image-sharing service (Pinterest) and evaluated the predic- tive power of different features on popularity (number of reshares). We found that visual properties have low pre- dictive power compared that of social cues. However, after factoring-out social in uence, visual features show consider- able predictive power, especially for images with higher ex- posure, with over 3:1 accuracy odds when classifying highly exposed images between very popular and unpopular.

Original languageEnglish
Title of host publicationWebSci 2014 - Proceedings of the 2014 ACM Web Science Conference
PublisherAssociation for Computing Machinery
Pages42-51
Number of pages10
ISBN (Print)9781450326223
DOIs
Publication statusPublished - 1 Jan 2014
Externally publishedYes
Event6th ACM Web Science Conference, WebSci 2014 - Bloomington, IN, United States
Duration: 23 Jun 201426 Jun 2014

Other

Other6th ACM Web Science Conference, WebSci 2014
CountryUnited States
CityBloomington, IN
Period23/6/1426/6/14

Fingerprint

Luminance

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

Totti, L., Costa, F., Avila, S., Valle, E., Meira, W., & Almeida, V. (2014). The impact of visual attributes on online image diffusion. In WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference (pp. 42-51). Association for Computing Machinery. https://doi.org/10.1145/2615569.2615700

The impact of visual attributes on online image diffusion. / Totti, Luam; Costa, Felipe; Avila, Sandra; Valle, Eduardo; Meira, Wagner; Almeida, Virgílio.

WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference. Association for Computing Machinery, 2014. p. 42-51.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Totti, L, Costa, F, Avila, S, Valle, E, Meira, W & Almeida, V 2014, The impact of visual attributes on online image diffusion. in WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference. Association for Computing Machinery, pp. 42-51, 6th ACM Web Science Conference, WebSci 2014, Bloomington, IN, United States, 23/6/14. https://doi.org/10.1145/2615569.2615700
Totti L, Costa F, Avila S, Valle E, Meira W, Almeida V. The impact of visual attributes on online image diffusion. In WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference. Association for Computing Machinery. 2014. p. 42-51 https://doi.org/10.1145/2615569.2615700
Totti, Luam ; Costa, Felipe ; Avila, Sandra ; Valle, Eduardo ; Meira, Wagner ; Almeida, Virgílio. / The impact of visual attributes on online image diffusion. WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference. Association for Computing Machinery, 2014. pp. 42-51
@inproceedings{dd559b0acb7943cd9d9755fac3ca08a8,
title = "The impact of visual attributes on online image diffusion",
abstract = "Little is known on how visual content affects the popularity on social networks, despite images being now ubiquitous on the Web, and currently accounting for a considerable frac- tion of all content shared. Existing art on image sharing fo- cuses mainly on non-visual attributes. In this work we take a complementary approach, and investigate resharing from a mainly visual perspective. Two sets of visual features are proposed, encoding both aesthetical properties (brightness, contrast, sharpness, etc.), and semantical content (concepts represented by the images). We collected data from a large image-sharing service (Pinterest) and evaluated the predic- tive power of different features on popularity (number of reshares). We found that visual properties have low pre- dictive power compared that of social cues. However, after factoring-out social in uence, visual features show consider- able predictive power, especially for images with higher ex- posure, with over 3:1 accuracy odds when classifying highly exposed images between very popular and unpopular.",
author = "Luam Totti and Felipe Costa and Sandra Avila and Eduardo Valle and Wagner Meira and Virg{\'i}lio Almeida",
year = "2014",
month = "1",
day = "1",
doi = "10.1145/2615569.2615700",
language = "English",
isbn = "9781450326223",
pages = "42--51",
booktitle = "WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference",
publisher = "Association for Computing Machinery",

}

TY - GEN

T1 - The impact of visual attributes on online image diffusion

AU - Totti, Luam

AU - Costa, Felipe

AU - Avila, Sandra

AU - Valle, Eduardo

AU - Meira, Wagner

AU - Almeida, Virgílio

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Little is known on how visual content affects the popularity on social networks, despite images being now ubiquitous on the Web, and currently accounting for a considerable frac- tion of all content shared. Existing art on image sharing fo- cuses mainly on non-visual attributes. In this work we take a complementary approach, and investigate resharing from a mainly visual perspective. Two sets of visual features are proposed, encoding both aesthetical properties (brightness, contrast, sharpness, etc.), and semantical content (concepts represented by the images). We collected data from a large image-sharing service (Pinterest) and evaluated the predic- tive power of different features on popularity (number of reshares). We found that visual properties have low pre- dictive power compared that of social cues. However, after factoring-out social in uence, visual features show consider- able predictive power, especially for images with higher ex- posure, with over 3:1 accuracy odds when classifying highly exposed images between very popular and unpopular.

AB - Little is known on how visual content affects the popularity on social networks, despite images being now ubiquitous on the Web, and currently accounting for a considerable frac- tion of all content shared. Existing art on image sharing fo- cuses mainly on non-visual attributes. In this work we take a complementary approach, and investigate resharing from a mainly visual perspective. Two sets of visual features are proposed, encoding both aesthetical properties (brightness, contrast, sharpness, etc.), and semantical content (concepts represented by the images). We collected data from a large image-sharing service (Pinterest) and evaluated the predic- tive power of different features on popularity (number of reshares). We found that visual properties have low pre- dictive power compared that of social cues. However, after factoring-out social in uence, visual features show consider- able predictive power, especially for images with higher ex- posure, with over 3:1 accuracy odds when classifying highly exposed images between very popular and unpopular.

UR - http://www.scopus.com/inward/record.url?scp=84904490651&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84904490651&partnerID=8YFLogxK

U2 - 10.1145/2615569.2615700

DO - 10.1145/2615569.2615700

M3 - Conference contribution

AN - SCOPUS:84904490651

SN - 9781450326223

SP - 42

EP - 51

BT - WebSci 2014 - Proceedings of the 2014 ACM Web Science Conference

PB - Association for Computing Machinery

ER -