TY - GEN
T1 - A glance at an overlooked part of the world wide web
AU - Trestian, Ionut
AU - Xiao, Chunjing
AU - Kuzmanovic, Aleksandar
PY - 2013
Y1 - 2013
N2 - Although according to surveys related to internet user ac- tivity it is considered one of the most popular aspects, few studies are actually concerned with internet pornography. This paper is aimed at rectifying that overlook. In particu- lar, we study user activity related to internet pornography by looking at two main behaviors: (i) watching pornogra- phy, and (ii) providing feedback on pornography items in the form of ratings and comments. By using appropriate datasets that we collect, we make contributions related to the study of both behaviors pointed out above. With regards to viewing, we observe that views are highly dependent on pornography category and video size. By studying the feedback system of pornography video websites, we observe differences in the way users rate items across websites popular in different parts of the world. Fi- nally, we employ sentiment analysis to study the comments that users leave on pornography websites and we find sur- prising similarities across the analyzed websites. Our results pave the way to understanding more about human behav- ior related to internet pornography and can impact, among others, fields such as content personalization, video content delivery, recommender systems.
AB - Although according to surveys related to internet user ac- tivity it is considered one of the most popular aspects, few studies are actually concerned with internet pornography. This paper is aimed at rectifying that overlook. In particu- lar, we study user activity related to internet pornography by looking at two main behaviors: (i) watching pornogra- phy, and (ii) providing feedback on pornography items in the form of ratings and comments. By using appropriate datasets that we collect, we make contributions related to the study of both behaviors pointed out above. With regards to viewing, we observe that views are highly dependent on pornography category and video size. By studying the feedback system of pornography video websites, we observe differences in the way users rate items across websites popular in different parts of the world. Fi- nally, we employ sentiment analysis to study the comments that users leave on pornography websites and we find sur- prising similarities across the analyzed websites. Our results pave the way to understanding more about human behav- ior related to internet pornography and can impact, among others, fields such as content personalization, video content delivery, recommender systems.
KW - Internet behavior
KW - Search behavior
KW - Sentiment analysis
UR - http://www.scopus.com/inward/record.url?scp=84893076325&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84893076325&partnerID=8YFLogxK
U2 - 10.1145/2487788.2488178
DO - 10.1145/2487788.2488178
M3 - Conference contribution
AN - SCOPUS:84893076325
SN - 9781450320382
T3 - WWW 2013 Companion - Proceedings of the 22nd International Conference on World Wide Web
SP - 1379
EP - 1386
BT - WWW 2013 Companion - Proceedings of the 22nd International Conference on World Wide Web
PB - Association for Computing Machinery
T2 - WWW 2013 Companion - Proceedings of the 22nd International Conference on World Wide Web
Y2 - 13 May 2013 through 17 May 2013
ER -