The document discusses the challenges of privacy and data ownership in centralized social networks, proposing a privacy-aware and scalable content dissemination framework for distributed social networks. It introduces the pubsubhubbub protocol and the semantic hub, which allows users to manage their privacy preferences while enabling efficient content distribution. Future work aims to enhance mobile device integration and offline communication capabilities.

1. Privacy-Aware and Scalable Content
Dissemination in Distributed Social Networks
Pavan Kapanipathi 1,2, Julia Anaya 1, Amit Sheth 2, Brett
Slatkin 3, Alexandre Passant 1
1 Digital Enterprise Research Institute, Galway – Ireland
2 Kno.e.sis, Dayton, OH- USA
3 Google, San Fransisco, CA - USA
1

2. Agenda
• Background
• Motivation
• Semantic Hub – Protocol
• SMOB + Semantic Hub
• Conclusion
• Future Work
2

3. Social Networks
• Centralized Social Networks (CSN)s
• Twitter and Facebook.
• What’s missing?
• Users don’t own their data
• Consistent privacy and performance concerns
• Distributed Social Networks
• Diaspora, SMOB, StatusNet, OneSocialWeb
3

4. Distributed Social Networks
• Users own their data
• Own Social Space
• Decentralized servers
• Better Performance
• Google’s PubSubHubbub protocol
• Communication protocol used in most Distributed Social
Networks
• Improves scalability
• Near-instant notifications
4

5. • PubSubHubbub – Communication protocol
• Simple, Open, web-hook based pubsub protocol
• Extension to RSS, Atom.
Give me your
latest content for
feed X, please
Publisher Subscriber
5
I have delegated
distribution task
to a Hub

6. • PubSubHubbub – Communication protocol
• Simple, Open, web-hook based pubsub protocol
• Extension to RSS, Atom.
I want to
subscribe to
feed X
Publisher Hub Subscriber
6
Hey Did Yup, its really
u send me not a DoS
this?? attacker.

7. • PubSubHubbub – Communication protocol
• Simple, Open, web-hook based pubsub protocol
• Extension to RSS, Atom.
I have new
Here is the
content for
latest content
feed X
for feed X
Publisher Hub Subscriber
Subscriber
Subscriber
Subscriber 7
Subscriber
Give me the
Here it is latest content 7
for feed X

8. Motivation – Privacy
• Twitter – Data is either
• Private – authorized followers
• Public – Everyone (No authorization required to
follow) “Cisco just offered me a job! Now I
have to weigh the utility of a fatty
paycheck against the daily commute
to San Jose and hating the work.”
8
http://www.msnbc.msn.com/id/29796962/ns/technology_and_science-tech_and_gadgets/t/twitter-gets-you-fired-characters-or-less/

9. Motivation
• Micropost -- “good presentation about Linked Data: then, now
and next http://bit.ly/xyz”
• Decide on your audience
• Almost certain about getting the message across to the right
audience
9

10. Protocol – Extension
(Overview)
• Publishers’ Preferences
• Content also includes its preferences (recipients)
• Complexity at the Hub
• Store Social Graph of the Publisher
• Filter appropriate Subscribers to push the content
10

11. PubSubHubbub Protocol
Extension
Hey I have new Here is the
Give me new content
content for feed the new
X + my of feed X
content Follower - A
preference Y
Follower - B
Pub Semantic Hub
Follower - C
Here it Follower - D
is
Get the subscribers
Social
of Pub whose profile
Graph matches preference Y
11

12. Protocol In-Use
• SMOB
• Privacy-Aware Microblog dissemination
• Personalized Filtering
• Twitter Stream (SPIM Workshop)
12

13. SMOB + Semantic Hub
• Open and Distributed Semantic Microblogging Framework
• http://smob.me
• PuSH implementation used to broadcast content
• Privacy was a concern
• Our Privacy-Aware protocol implementation used
• http://semantichub.appspot.com
13

14. POST RDF
“Presentation at
ISWC by me on Get subscribers SMOB – B
Get my privacy settings for Sunday matching And
dbPedia:Semantic_Web #semanticweb” Access Space SMOB - D
There is a micropost
SMOB – A
RDF
Interest: Sematic Web Get micropost+ Semantic Hub
Access Space
Access Space: #semanticweb
SELECT ?user WHERE {
?user foaf:topic_interest ?topic
?topic dcterms:subject category:Semantic_Web
Push Updates
SMOB – B SMOB – C SMOB – D
Interest: Knowledge Engg Interest: Computer Networks Interest: Sematic Search
14

15. POST RDF
“Presentation at
ISWC by me on Get subscribers SMOB – B
Get my privacy settings for Sunday matching And
dbPedia:Semantic_Web #semanticweb” Access Space SMOB - D
There is a micropost
SMOB – A
RDF
Interest: Sematic Web Get micropost+ Semantic Hub
Access Space
Access Space: #semanticweb
SELECT ?user WHERE {
?user foaf:topic_interest ?topic
?topic dcterms:subject category:Semantic_Web
Push Updates
SMOB – B SMOB – C SMOB – D
Interest: Knowledge Engg Interest: Computer Networks Interest: Sematic Search
15

16. Publisher – Privacy Preferences
• Feature based preferences at Publisher
• Content generated for topic Linked Data
• Micropost containing hashtag #knoesis
• Modeled using Privacy Preference Ontology (PPO)
• Fine grained restrictions
• SPARQL Queries representing the subset of subscribers from the
Social Graph in the hub
16
Privacy Preference Ontology – Work by Owen Sacco, Alexandre Passant DERI

17. Privacy Preference Ontology (PPO)
• Publishing Data on Web as a part of Linked Open Data has
privacy issues.
• Web Access Control ontology (WAC)
• PPO -- Light weight vocabulary
• On top of Web Access Control Ontology
• Fine grained privacy preferences for RDF Data
• Example: Share email, phone number in my FOAF profile to Users
with workplace as Kno.e.sis Center.
17

18. Privacy Preferences using PPO
Example: Restrict all the documents tagged with Semantic
web to only those users who are interested in Semantic
Web
<http://example.org/privacy/3> a ppo:PrivacyPreference;
ppo : appliesToResource
<http://xmlns.com/foaf /0.1/Document>;
ppo : hasCondition [
ppo : hasProperty tag:Tag;
ppo : resourceAsObject dbpedia : Semantic Web
];
ppo: assignAccess acl :Read;
ppo : hasAccessSpace [
ppo:hasAccessQuery ”SELECT ?user WHERE { 18
?user foaf : topic interest dbpedia:Semantic Web }”
].

19. SMOB’s Simple Interface – Privacy
Preference Generation
19

20. POST RDF
“Presentation at
ISWC by me on Get subscribers SMOB – B
Get my privacy settings for Sunday matching And
dbPedia:Semantic_Web #semanticweb” Access Space SMOB - D
There is a micropost
SMOB – A
RDF
Interest: Sematic Web Get micropost+ Semantic Hub
Access Space
Access Space: #semanticweb
SELECT ?user WHERE {
?user foaf:topic_interest ?topic
?topic dcterms:subject category:Semantic_Web
Push Updates
SMOB – B SMOB – C SMOB – D
Interest: Knowledge Engg Interest: Computer Networks Interest: Sematic Search
20

21. Semantic Hub – Distributed Social
Semantic Graph
• FOAF Profiles of Publishers/Subscibers
• Semantic Social Graph at the Semantic Hub
• FOAF profiles linked using push vocabulary
• Distributed
• Every user owns a Hub
21

22. FOAF -- Exchange
Hey, I want to Great, Since it is
Publish/Subscribe your first time. I
Content of feed X need your profile
Pub/Sub Semantic Hub
OK, Let me
confirm ur
identity Profile
stored
Social 22
Here is my Graph
profile

23. Push Vocabulary
• SIOC vocabulary
• Models information from online communities
• Communication protocols are not considered
• PubSubHubbub Vocabulary (push)
• Models information of users communicating via Semantic Hub
23

24. 24

25. POST RDF
“Presentation at
ISWC by me on Get subscribers SMOB – B
Get my privacy settings for Sunday matching And
dbPedia:Semantic_Web #semanticweb” Access Space SMOB - D
There is a micropost
SMOB – A
RDF
Interest: Sematic Web Get micropost+ Semantic Hub
Access Space
Access Space: #semanticweb
SELECT ?user WHERE {
?user foaf:topic_interest ?topic
?topic dcterms:subject category:Semantic_Web
Push Updates
SMOB – B SMOB – C SMOB – D
Interest: Knowledge Engg Interest: Computer Networks Interest: Sematic Search
25

26. Semantic Hub – Content
Distribution
• RSS Extension
• Element to include the privacy preference
• privacy, accessspace
• Each updated content is received by restricted subscribers
• Privacy Preference for each content fetches the matched
subscribers
• Preferences are deleted before distribution
26

27. RSS Extension
<item>
<title>Only Friends</title>
<description>
Send this to only people I know and interested in Semantic
Web
</description>
<link>http://example.org /rss</link>
<guid>123123123123</guid>
<pubDate>March 06 2001</pubDate>
<privacy>
<accessspace>
SELECT ?user WHERE {
foaf :me foaf :knows ?user .
?user foaf : topic interest dbpedia:Semantic Web
.}
</accessspace>
… 27
</privacy>
</item>

28. Personalized Filtering of the
Twitter Stream
• Profile Generator
• Automatic generation of User Profiles
• Semantic Filter
• Annotating Twitter Stream with concepts from Linked Open Data
• Semantic Hub
• Delivering tweets to appropriate Interested Users (near real-time)
28

29. Architecture
Annotate: iPhone Get
?user foaf:interest Subscribers
The new based on
dbPedia:iPhone
iPhone has a 3.5- preference
Union
inch screen,
?user foaf:interest
released today
Category:Apple
Get Interested
Subscribers
RDF
Semantic Filter Notify Update
A
N RDF
N Store and
O
T
Query Topics Semantic Hub
A Fetch Updates
T
O RSS Store FOAF
R
Update RSS
Profile Generator
Push Updates to
Interested Users
29
Create Profile

30. Conclusion
• Extension of Google’s PuSH protocol to cope with privacy
concerns.
• Implementation of the protocol
• Use in SMOB (Microblogging)
• Push vocabulary to model activities via PuSH protocol
30

31. Future Work
• Enabling the architecture for mobile devices.
• Communicate to and from devices that are offline and still
need to be notified.
31

32. Thanks
Acknowledgements:
Fabrizio Orlandi
Owen Sacco
Contact us:
{pavan, amit}@knoesis.org
{alexandre.passant, julia.anaya}@deri.org
brettslatkin@google.com
32
This work is funded by (1) Science Foundation Ireland under grant number SFI/08/CE/I1380 (Lıon 2) and by an
IRCSET scholarship supported by Cisco Systems (2) Social Media Enhanced Organizational Sensemaking in
Emergency Response, National Science Foundation under award IIS-1111182, 09/01/2011 - 08/31/2014.

33. PubSubHubbub
• PuSH is a scalable protocol, and Google provides a public hub
that people can use to broadcast their content.
• This public hub delivers for approximately 40 million unique
active feeds
• 117 million subscriptions.
• In two years, approximately 5.5 billion unique feeds have been
delivered
• fetching 200 to 400 feeds
• delivering 400 to 600 of them per second.
• Its largest subscribers get between 20 and 120 updates per
second from the hub.
33

34. 34

35. 35