Introduction To Online Social Networks Information Technology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

At the most basic level an online social network (OSN) provides facilities for individuals to interact. End users are participating not only as passive consumers of contents provided by these websites, but also as contributors creating various contents. In Facebook, MySpace etc we feel that we are in full control of our data. Nevertheless, due to the centralized design owners of these sites have full control over our data. These social networking sites provide the concept called profile to maintain our digital identity. However, this concept does not correctly help us to map our real life identity in the digital world. In the real life, we have different relationships with people around us and things that we share with them differ depends on the relationship. We can consider them as different views of our identity that we cannot map to our digital identity. Since the audience judges our character at OSNs based on what information and interests posted, most often people tend to maintain multiple accounts on social networks to maintain these different views of digital identity. We considered this as another important privacy issue in current social networking sites. Still many companies consider use of OSN is illegitimate or inappropriate as employees may inadvertently disclose company confidential information with friends. Due to the centralized design multiple parties cannot enforce different policies is on the same profile and current solutions is companies maintain their own social networking sites and giving users the burden of maintaining multiple accounts.

Decentralized OSN, where users have full control of their data is identified as more privacy preserving solution and discussed in detailed in [1] and [2].

In order to overcome above discussed drawbacks in current OSN we introduce a decentralized architecture where users maintain their profiles on servers trusted by them. We decouple the data stored by the user and its representation. Instead of having a profile with a global view this proposed architecture allows users maintain different views of their digital identify under a single profile. Therefore, user has the opportunity to interact with data stored in multiple servers representing different aspects of their identity. For instance, some of his data may come from his office server related to the professional life and some from existing social networks such as Facebook. This approach helps organizations to maintain their own social networking sites, where they have the ownership of the data while user has the opportunity of using only one profile to maintain different aspects of his digital identity.

Introduction to Online Social Networks

Today normal web users' life without Facebook is almost unthinkable. Many of us have now integrated these online social networks such as Facebook, MySpace etc into our daily practices. More and more users use social networking sites to exchange personal information with their friends, such as their personal information, images, videos and even what they are doing now or other brief expressions, thereby they are releasing lot of personal data to social networks. Furthermore social networks provides facilities to even find new friends based on shared affinities such as groups, hobbies, interests or overlaps in friendship circles. Based on above features Boyd and Ellison defined social network sites as "web-based services that allow individuals to (1) construct a public or semi-public profile within a bounded system, (2) articulate a list of other users with whom they share a connection, and (3) view and traverse their list of connections and those made by others within the system." [3].

When we consider current popular social networking sites, features provided by them can be mainly classified into three main categories as social link establishment, digital personal spaces, and means of communication such as instant messaging (IM) or private messaging similar to webmail [4]. However, popular OSNs have now extended these capabilities by allowing third party developers to develop applications and integrate with them. For instance, Facebook users have the opportunity to use such applications ranging from simple quizzes to multiuser online games. Most of these OSNs are now becoming closer to full-fledged application development platforms. In [1] Datta et al. mention that "if a Web browser becomes an operating system for the next-generation computing devices as predicted by various technological experts, it is very likely that online social network service would be the user interface of that operating system".

Brief History, the first social network site that complied with the Boyd and Ellison's definition launched in 1997. It allowed users to create profiles, list their friends. However, most of these features existed in other sites and was the first to combine all these features [3]. attracted million of users but failed to sustain as a business and shutdown around 2000. Friendster, the next most popular social network site was launched in 2002. It was designed to help friends-of-friends meet which was not available in most of the sites during that time. With the popularity increased Friendster failed to cater rapid user growth with its design. Due to these technical issues and several other issues as discussed in [3] caused its users to lost trust with the site and leave.

From 2003-2004 several social network sites such as Cyworld, Friendster, Plaxo,, Hi5, LinkedIn, MySpace, Orkut, Facebook, and Live Spaces were emerged [5]. Even today, most of these sites are extremely popular and account for significant amount of today's web traffic. For instance according to Google's most visited sites for year 2010 [6] Facebook ranked the number one site by number of unique visitors. It scores more than 540 million unique visitors per month, reaching a sizeable chunk of 35.2 percent of the Internet population. Apart from Facebook, other sites such as Orkut and MySpace are among top 50 sites.

Current Online Social Network Architecture

Most of current popular social network sites like Facebook, MySpace and LinkedIn are based on the high-level architecture depicted in Figure 1.

Figure 1: High-level architecture of current OSN

They have a central location where all the data about users are stored. There are set of modules that operate on top of this data model and providing users social networking services such as finding other people, sharing pictures/videos, exchanging professional information, etc. This architecture paradigm is termed as centralized [7], [1]. Under this architecture, people use various client applications, mostly web browsers to interact with their profiles stored in this central repository. Application layer of these sites which is based on Web 2.0[8] allows end users participate not only as passive consumers of content provided by the websites but also as a contributor creating content collaboratively with fellow users. Although most are web based social networks, currently they support mobile interactions as well, for instance Facebook has more than 200 million active mobile users [9]. In addition, there are some mobile specific social networks such as Dodgeball.

While social networking sites currently provide wide variety of services and features these features are implemented around the concept of user profiles, which displays information of users and articulate list of friends who are also users of the same system. After joining, we are asked to fill out forms containing series questions to collect information about us and the profile is generated using the answers to these questions, which typically include descriptors such as age, location, interests, etc. Most sites also encourage users to upload a profile photo. The core application used by the members is the creation and maintenance of their online relations. The label for these relationships differs depending on the sites. Popular terms include Friends, Contacts and Fans. Mostly these sites require bi-directional confirmation for Friendship, but some do not. These one-directional ties are sometimes labeled as Fans or Followers. One interesting feature that is popular today is called Wall in Facebook and available in other sites as well, which is quite similar to guest books on private web sites. Apart from these, messaging similar to webmail and instant message are some of the common features available today.

Issues of Current Online Social Networks

As described earlier current social network has a centralized architecture, they are prone to some problems that we can categorize them as social and technical issues. Some of which have even led to the demise of many early-generation OSN sites such as SixDegrees and Friendster [3].

Technical issues arise due to centralized management of large number of user population. When we consider the Facebook the largest and most popular social network, it has more than 500 million active users and more than 30 billion pieces of content such as web links, news stories, blog posts, notes, photo albums, etc. shared each month.[9] That can lead to various performance and scalability issues most notably frequent down time and slowness and unresponsiveness. A report on slowness and poor site performance of social networking sites is given in [10]. In addition, these site operators will have to face increasing cost in managing and maintaining the infrastructures to ensure a smooth continuation and reasonable performance of the provided services.

Privacy of social network user related to his data and online identity is the main issue trying to address by this project. Additionally, integrity and availability related security risks are also present in current centralized architecture as discussed in [11] and [12]. Therefore, privacy issues present in current social networks are discussed in detail.

Privacy Issues of Online Social Networks

The main privacy issue is the privacy of users' personal data. In OSNs like Facebook, MySpace or Orkut users can adjust privacy settings to protect their content and limit access by other users. However, it was reported that more than half a million images were leaked from MySpace without any knowledge from the users [13]. Since these data is centrally administrated by service providers there is no protection against access by the owners of these sites. For example, Facebook provides an option for its users to deactivate their accounts. However, this does not guarantee that all personal information of that user is completely deleted from the site [14]. Another major threat to users' privacy is from third party applications. The famous incident is Beacon, a Facebook's online advertisement system, tracked users' online activities on third-party sites that participate in Beacon and broadcast these activities to their Facebook users [15] without users' knowledge. Various research conducted to analyze these privacy issues presented in social networks, mainly associated with Facebook are discusses in [16] and [17].

All privacy issues discussed above arises due to user's lack of control on his data and misuse by web site owners. However, there is another aspect of online privacy that is not thoroughly studied that is the users' inability to map his real life identity to online identity. In real life, the identity we have depends on the people we interact. For instance, relationships we have in professional life is different compared to relationships with normal friends or family members. Our real life relationships are more complex and now it is difficult to represent the same at digital world using a single profile and a friend list. We get people from work, family members and college friends all connected to the same profile, completely different audiences and they probably should not be seeing the same things. Because we are judged by seeing what information, interests we post on the social network. Skeels and Gridin[5] in their study in workplace use of social network among Microsoft employees, found tension arises due to mixing professional and personal contacts in social networks. Facebook's friend list tries to solve this issue but still it does not full address this issue. Hence, people try to maintain multiple accounts on the same network.

Inability to Share Information across Online Social Networks

Table 1 - Member Overlap at Social Networks [18]

Table 1 shows the percentage of users in the social network given in the row who also use other network given in the column. For example, 20% of MySpace members are also Facebook members, while 64% of Facebook members also belong to MySpace. In addition, we can see that members of smaller online social networking communities such as Plaxo, Hi5 etc also participate in one of the larger networks. Due to different relationships we have on these sites we tend to create multiple account. Since these sites do not allow users to share information with each other and act as information silos, we have to upload same information on these accounts.

Social Networking at Workplace

Social networking is identified as efficient method of communication among colleagues at workplace. Research done on identifying workplace social networking usage is discussed in [5] and [19]. In [5] Skeels and Grudin discuss a study of Microsoft employees' workplace use of Facebook and LinkedIn. It analyzes both benefits of social network at workplace as well as issues Microsoft employees face while using them at workplace and for professional purposes.

The main benefit of social networks at workplace is, it provides a better communication medium for connecting with co-workers and getting to know other employees. In addition, social networks can be easily used to share knowledge, discuss work related problems and even it can used as a medium to acquire customer feedback to aid in product development and marketing.

However, most organizations consider use of social networking site such as Facebook at workplace is inappropriate or illegitimate. According to Skeels and Grudin[5] one major concern organizations have on public social network sites is inadvertent disclosure of proprietary information by employees as they don't have any control over the information employees share on these sites and organization cannot enforce any policies over proprietary information employees share on these sites. Due to its benefits and issues of public sites most major organization now, use in house social networks such as BeeHive at IBM and IdeaStorm at Dell {CITATION}.

Skeels and Grudin[5] also identified issues employees face due to use of Facebook and LinkedIn at workplace. One major issue is the tension from mixing personal and professional persons on the same network.

The Problem

Despite various privacy infringements reported, Social networking sites such as Facebook now become an essential part of life. The main issue, we identified and try to address in this project is the privacy of users' data. As discussed previous section of this chapter due to centralized architecture, we cannot achieve full control over our data we store on these sites. Another limitation of current social networks that we try to address by this project is inability of current social networks to represent our online identity representing complex relationships we have in the real world. In addition, still we do not see any facility provided by these sites to exchange our profile information between them.

Ideally, we should move to decentralized social networking architecture. It will provide facilities for its users to store their data on server where they have full control and trust. These servers can be on the cloud or on at workplace or education institute. Based on this architecture different authorizes have the ability to enforce policies on data users store on their servers. Then there should be facilities for users to aggregate these sources and build their online profile. By making a single profile having many views that he wants to share with different connections and decouple these views from the data source user can build an online identify similar to the real world identity.


The objectives of this project is to build social networking platform with the help of existing open source solutions that provide facilities to store data on servers that user have full control and based on data coming from different sources build online identity that can easily represent different aspects of his real world character.


During the discussion of this research, following terms and abbreviations are heavily used throughout remainder of this document. In order to help reader get good understanding some of these are defined here.

DHT - Distributed Hash Table. This is similar to conventional hash table data structure. Data items are assigned a random key from a large identifier space when looking up a data item, the network address of the node responsible for that data item is returned. [20]

OSN - Online Social Network

Distributed System - A set of processes that communicate by message passing and carry out some desired functions over a period of time [20]

P2P - Peer to Peer

Scalability - The ability of a system to continue to operate correctly even when it is scaled to a larger size. In the context of this project, the system should continue to operate correctly even when the system scales up or scales down.[20]

Literature Review


A decentralized online social network is an online social network implemented on a distributed information management platform, such as a network of trusted servers or peer-to-peer systems. Many researches are currently going on in implementing distributed social networking solution. All these researches suggest decentralization of users' data and give them full control of data would preserve privacy while providing a highly scalable and high performance infrastructure. Based on these trends on how social network architecture is evolving Datta et al[1] categorize social networks as depicted in Figure 2. Along the vertical axis, the features and services provided by online social networks have been extended significantly. Early social networking sites, from a simple online tool to manage personal and professional contacts, such as in SixDegrees and Friendster, to an effective tool for sharing several kind of information and contents with a viral spread. Popular OSNs such as Facebook offer users even more services and applications, as third parties are allowed to develop and plug their applications into the site. Along the horizontal axis, we can see how current trends on social network architecture are evolving.

Figure 2 - Classification and development trend of online social network services [1]

Although we can see, some social networking approaches presented in both distributed and peer-to-peer architectures they are still on the initial research phases and not widely used. Still all the famous OSNs such as Facebook, MySpace and Orkut follow centralized approach.

As discussed in the previous chapter we can see there are many advantages in decentralized social networks compared to current centralized approach. When analyzing existing literature we can see that there are several architectural approaches in achieving decentralized social networks. Taking centralized OSNs and decentralizing them or by adding a social component to current decentralized applications that do not have a social component yet. It can also be achieved by adding both properties at the same time. [1] This project mainly focuses on the first approach.

Academic Research on P2P OSN

In this section of the report, we will look at current academic research done on peer-to-peer based decentralized social networks. Except for the Safebook other systems do not mention about any actual implementation and therefore we assumed that they are still at the research phase. Moreover, joining Safebook is based on invitation only therefore analyses of these systems are based on research papers published by them.


Figure 3 - Safebook Architecture [11]

Safebook[11] is a P2P based approach to provide a decentralized general purpose OSN with the main focus on protecting its users' privacy. In addition to data privacy, in Safebook due to its P2P overlay architecture and it provides communication privacy for its users. Connections and data sharing between users cannot be tracked at the network level.


Since data is distributed among friends, who may not be continuously online achieving high availability of user's data is considered as one of the major drawbacks of P2P solution. Vis-`a-Vis[21] proposes an distributed social networking architecture that achieve privacy as well as high availability through Virtual Individual Servers (VISs), a personal virtual machine instance hosted by a cloud computing utility such as Amazon's EC2[22].


In addition PeerSoN[4], aims at keeping the features of OSNs but overcoming two limitations: privacy issues and the requirement of Internet connectivity for all transactions. To address the privacy problem, it uses encryption and access control coupled with a peer-to-peer approach to replace the centralized authority of classical OSNs.

Distributed OSN

This section will look at another approach of decentralized social networks. In this architecture, our data is stored on servers that we trust and we have full control on them.

FOAF Based Decentralization

Figure 4 - Decentralization through FOAF [23]

Yeung et al [23] proposes decentralization as the future of online social network. They present a framework that enables users to publish their FOAF [24] profiles on dedicated trusted servers. Then users can manage their profiles through open web-based protocols such as WebDAV. Major advantage of this approach is its fully compatible with most of current social networking platforms and web based. In additions most of the standards used in this approach are supported by the W3C consortium.





Challenges for Distributed OSN

Technical Challenges

Social Challenges