A Semantic Model for Friend Segregation in Online Social Networks - - PowerPoint PPT Presentation

a semantic model for friend segregation in online social
SMART_READER_LITE
LIVE PREVIEW

A Semantic Model for Friend Segregation in Online Social Networks - - PowerPoint PPT Presentation

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and


slide-1
SLIDE 1

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

A Semantic Model for Friend Segregation in Online Social Networks

Javed Ahmed

ICWE 2016, Lugano, Switzerland

08-06-2016

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-2
SLIDE 2

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

1

Introduction Online Social Networks Users as a Content Manager Problems with Existing Privacy Controls Lists and Circles

2

Research Problem Failure to Represent Diverse Social Relationship Context Collapse Research Questions

3

Our Approach

4

Theoretical Framework for Privacy

5

Results of User Study: Interaction Pattern

6

Results of User Study: Preferred Interaction

7

Results of User Study: Interaction Ranking

8

Ontological Model

9

Conclusion and Future Work

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-3
SLIDE 3

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Online Social Networks Users as a Content Manager Problems with Existing Privacy Controls Lists and Circles

Online Social Networks

Online social networks (OSNs) have experienced exponential growth in recent years. OSNs are the top most visited sites on the Internet.1 OSNs are the fourth most popular activity on the Internet nowadays.2 OSNs are one of the most popular fora for self representation and user interactions.

1Alexa http:/www.alexa.com/topsites 2Nielsen http://www.nielsen.com/ Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-4
SLIDE 4

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Online Social Networks Users as a Content Manager Problems with Existing Privacy Controls Lists and Circles

Users as a Content Manager

Tremendous growth of OSNs resulted in fundamental shift in status of end users Individual end users become content managers instead of just being content consumers Today, for every single piece of data shared on OSNs, the uploader must decide which of the users should be able to access the data This put huge cognitive burden on users and majority of users stick with default privacy Default privacy settings are very permissive in nature and expose data to unintended audiences

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-5
SLIDE 5

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Online Social Networks Users as a Content Manager Problems with Existing Privacy Controls Lists and Circles

Problems with Existing Privacy Controls

Multitude of privacy controls to manage access to uploaded content of OSN users Interfaces for privacy controls are too complicated to most of the normal users Current interface has limited visual feedback, and promotes a poor mental model of how the settings affect the profile visibility Privacy settings for uploaded content are often incorrect, failing to match users’ expectations

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-6
SLIDE 6

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Online Social Networks Users as a Content Manager Problems with Existing Privacy Controls Lists and Circles

Lists and Circles

Some of the social networking sites have begun providing lists and circles features This help users in organizing a large friend network into groups Grouping several hundred friends into different lists is a laborious process What criteria should be used to construct the lists and circles? Are these lists meaningful for setting privacy policies? The responsibility of maintaining the appropriateness of these lists lies solely on the user.

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-7
SLIDE 7

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Failure to Represent Diverse Social Relationship Context Collapse Research Questions

Failure to Represent Diverse Social Relationship

Online social networks fail to represent diverse social relationships of the users Most OSNs employ friendship as only type of bidirectional relationship This provide only coarse indication of nature of relationship between two users In reality social relationships are of varying tie strength, dynamic, asymmetric It is challenging task to model dynamism, asymmetry, and relationship strength in user relationships

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-8
SLIDE 8

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Failure to Represent Diverse Social Relationship Context Collapse Research Questions

Context Collapse

In reality people play diverse roles and display their personal information according to the role OSNs place employers and romantic partners on the same communication plane It makes presentation of varied versions of self difficult for OSN users Difficulty in disclosing information selectively to various life facets can lead to context collapse The collapsing of social contexts has emerged as an important problem with the rise of online social networks

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-9
SLIDE 9

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work Failure to Represent Diverse Social Relationship Context Collapse Research Questions

Research Questions

The main question for this research is how to represent diverse social relationships of the users in online social networks More specifically, we want to explore whether a users’ interaction pattern with his friends can be used as a basis for inferring relationship strength among users

1

How interaction pattern and profile similarity attributes reveals strength and context of relationship among OSNs users?

2

How to develop a semantic model for friend segregation depending on strength and context of relationship among OSNs users?

3

How to evaluate the model for friend segregation in online social networks?

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-10
SLIDE 10

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Our Approach

Theoretical framework for privacy from social perspective User study to examine interaction pattern of OSN users An ontological model to represent diverse social relationships

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-11
SLIDE 11

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Theoretical Framework for Privacy

Our theoretical framework for privacy address following three aspects: Context Segregation keeps audiences separate and compartmentalize social life of OSN user Disclosure Minimization The relationship strength plays vital role in disclosure minimization User Control enhances user control over shared resources among several users

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-12
SLIDE 12

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Results of User Study: Interaction Pattern

20 40 60 80 100 120 140 Family Friends Colleagues Classmate Acquaintances Strangers

InteractionFrequency

Never Rarely Monthly Weekly SeveralTimesAWeek Daily

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-13
SLIDE 13

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Results of User Study: Preferred Interaction

20 40 60 80 100 120 140 160 180 200 Family Friends Colleagues Classmate Acquaintances Strangers

PreferredInteraction

Not Applicable Play Games Wishing Chatting Liking Commenting Tagging Posting Messaging

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-14
SLIDE 14

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Results of User Study: Interaction Ranking

100 200 300 400 500 600 1 2 3 4 5 6 7 8 9 10

Interaction Ranking

Messaging Posting Tagging Commenting Liking Chatting Wishing Play Game

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-15
SLIDE 15

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Ontological Model

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks

slide-16
SLIDE 16

Introduction Research Problem Our Approach Theoretical Framework for Privacy Results of User Study: Interaction Pattern Results of User Study: Preferred Interaction Results of User Study: Interaction Ranking Ontological Model Conclusion and Future Work

Conclusion and Future Work

We address social aspects of privacy that are ignored by existing solutions to larger extent The model is inspired by well known social theories of Erving Goffman, Helen Nissenbaum and Granovetter It is first step towards formalizing social theories. The main challenge is evaluation of such model

Javed Ahmed A Semantic Model for Friend Segregation in Online Social Networks