Exploring personal attributes from unprotected interactions

Pritam Gundecha, Jiliang Tang, Xia Hu, Huan Liu

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

Abstract

Research, so far, has shown that many personal attributes, including religious and political affiliations, sexual orientation, relationship status, age, and gender, are predictable providing users' interaction data. To address these privacy concerns, users on a social networking site like Faceboook are usually left with profile settings to mark some of their data invisible. However, users sometimes interact with others using unprotected posts (e.g., posts from a "Faceboook page"). Although the aim of such interactions is to help users to become more social, visibilities of these interactions are beyond their profile settings and publicly accessible to everyone. The focus of this paper is to explore such unprotected interactions so that users' are well aware of these new vulnerabilities and adopt measures to mitigate them further. In particular, we ask - are users' personal attributes predictable using only the unprotected interactions? To answer this question, we design a novel problem of predictability of users' personal attributes with unprotected interactions. The extreme sparsity patterns in users' unprotected interactions pose a serious challenge for the proposed problem. Therefore, we first provide a way to mitigate the data sparsity challenge and propose a novel attribute prediction framework using only the unprotected interactions. Experimental results on Faceboook dataset demonstrates that the proposed framework can predict users' personal attributes.

Original languageEnglish (US)
Title of host publicationProceedings of the 10th International Conference on Web and Social Media, ICWSM 2016
PublisherAAAI press
Pages575-578
Number of pages4
ISBN (Electronic)9781577357582
StatePublished - 2016
Event10th International Conference on Web and Social Media, ICWSM 2016 - Cologne, Germany
Duration: May 17 2016May 20 2016

Publication series

NameProceedings of the 10th International Conference on Web and Social Media, ICWSM 2016

Other

Other10th International Conference on Web and Social Media, ICWSM 2016
CountryGermany
CityCologne
Period5/17/165/20/16

ASJC Scopus subject areas

  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Exploring personal attributes from unprotected interactions'. Together they form a unique fingerprint.

Cite this