Abstract | ||
---|---|---|
Balancing personalization and privacy is one of the challenges marketers commonly face. The privacy dilemmas associated with personalized services are particularly concerning in the context of social networking websites, wherein the privacy dichotomy problem is widely observed. To prevent potential privacy violations, businesses need to employ multiple safeguards beyond the current privacy settings of users. As a possible solution, companies can utilize user social footprints to detect user privacy preferences. To take a step towards this goal, we first ran a series of experiments to examine if the privacy preference attribute is homophilous in social media. As a result, we found a set of clues that users' privacy preferences are similar to the privacy behaviour of their social contacts, signaling that privacy homophily exists in social networks. We further studied users located in different neighbourhoods with varying degrees of privacy and found a set of characteristics that are specific to public users located in private neighbourhoods. These identified features can be used in a predictive model to identify public user accounts that are intended to be private, supporting companies to make an informed decision whether or not to exploit one's publicly available data for personalization purposes.
|
Year | DOI | Venue |
---|---|---|
2018 | 10.1145/3209542.3209564 | HT |
Keywords | Field | DocType |
Social privacy, Social network analysis, Preference detection | World Wide Web,Social media,Social network,Homophily,Computer science,Social network analysis,Exploit,User privacy,Personalization | Conference |
ISBN | Citations | PageRank |
978-1-4503-5427-1 | 1 | 0.36 |
References | Authors | |
21 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Taraneh Khazaei | 1 | 16 | 4.42 |
Lu Xiao | 2 | 38 | 9.44 |
Robert E. Mercer | 3 | 254 | 46.93 |
Atif Khan | 4 | 17 | 4.16 |