Learning based access control in online social networks

Social networking services (SNSs) such as Facebook or Twitter have experienced an explosive growth during the few past years. Millions of users have created their profiles on these services because they experience great benefits in terms of friendship. SNSs can help people to maintain their friendships, organize their social lives, start new friendships, or meet others that share their hobbies and interests. However, all these benefits can be eclipsed by the privacy hazards that affect people in SNSs. People expose intimate information of their lives on SNSs, and this information affects the way others think about them. It is crucial that users be able to control how their information is distributed through the SNSs and decide who can access it. This paper presents a list of privacy threats that can affect SNS users, and what requirements privacy mechanisms should fulfill to prevent this threats. Then, we review current approaches and analyze to what extent they cover the requirements.

show abstract

“…Shehab et al [60] introduce a privacy policy recommender system that is based on supervised learning. Their system works in 5 steps.…”

Section: Based On Automatic Learningmentioning

confidence: 99%

Open Challenges in Relationship-Based Privacy Mechanisms for Social Network Services

Fogues

Such

Espinosa

et al. 2015

International Journal of Human-Computer Interaction

View full text Add to dashboard Cite

show abstract

“…In recent years, several models have been proposed, and some have gained serious attention. Among them are the rule based privilege computation model [23], privacy aware keyword search [37], and access control based on ontology [38], learning [39], evolution [40] and roles [41].…”

Section: Access Control As a Privacy Theorymentioning

confidence: 99%

Empowering Evolving Social Network Users with Privacy Rights

Jamil¹

2013

Preprint

View full text Add to dashboard Cite

Considerable concerns exist over privacy on social networks, and huge debates persist about how to extend the artifacts users need to effectively protect their rights to privacy. While many interesting ideas have been proposed, no single approach appears to be comprehensive enough to be the front runner. In this paper, we propose a comprehensive and novel reference conceptual model for privacy in constantly evolving social networks and establish its novelty by briefly contrasting it with contemporary research. We also present the contours of a possible query language that we can develop with desirable features in light of the reference model, and refer to a new query language, PiQL, developed on the basis of this model that aims to support user driven privacy policy authoring and enforcement. The strength of our model is that such extensions are now possible by developing appropriate linguistic constructs as part of query languages such as SQL, as demonstrated in PiQL.

show abstract

“…Recent work has proposed using machine learning to recommend social network privacy settings [5], [14]. However, the output of both proposed tools is a verbose policy, which enumerates a list of friends for each data item.…”

Section: Related Workmentioning

confidence: 99%

“…Notice that this policy is specified in terms of newlycreated lists (High School and Grad School), and that the Work Colleagues list has been repaired to include Wayne. In addition to simplifying complex rule-based policies, enList can also be used in conjunction with new machine learning tools that recommend detailed privacy policies to social network users [5], [14]. The proposed tools produce recommendations in the form of verbose permission sets.…”

Section: Introductionmentioning

confidence: 99%

enList: Automatically Simplifying Privacy Policies

Bejugam

LeFevre

2011

2011 IEEE 11th International Conference on Data Mining Workshops

View full text Add to dashboard Cite

Online social networking sites have become extremely popular. Due to the pervasiveness of these sites, it is important to provide tools that allow users to specify detailed policies controlling access to their data. However, the policies specified using existing tools are often complex, verbose, and difficult to understand. In this paper, we study the policy simplification problem. Given a complex or verbose policy, our goal is to automatically produce an equivalent policy that is easier to understand. We propose a novel framework called enList, which automatically extracts friend "lists" (semantically meaningful subgroups of a user's friends) and then simplifies an existing policy using the lists. A laboratory-based user study confirms that the resulting policies are easier for users to comprehend, remember, and maintain than the policies produced by an existing recommendation tool.

show abstract

Learning based access control in online social networks

Cited by 13 publications

References 1 publication

Open Challenges in Relationship-Based Privacy Mechanisms for Social Network Services

Open Challenges in Relationship-Based Privacy Mechanisms for Social Network Services

Empowering Evolving Social Network Users with Privacy Rights

enList: Automatically Simplifying Privacy Policies

Contact Info

Product

Resources

About