The Rapid Growth of Online Social Media Networks

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.

Social networking used to connect and share information with friends.People may use

social networking services for different reasons to network with new contacts, reconnect

with the friends, maintain the relationships status, for business or project work related ,

take participate in discussions on the many topic, or just have get together meeting and

interaction with other participating users.[1].There are number of users on SocialNetwork

and Twitter. LinkedIn has positioned itself as a professional networking site profiles

include resume information and groups are created to share many questions and ideas

with other users in similar fields. Unlike traditional personal homepages people in these

societies publish not only their personal attributes, but also their relationships with

friends.It may causes the privacy violation in social networks[3].Information privacy is

needed for users. Existing techniques are used to prevent direct disclosure of sensitive

personal information.Here the focuses on social network data classification and inferring

the individuals private information. More private information are inferred by applying

collective classification algorithm. The system enhance how the online data of social

network is used for prediction some person’s private attribute that a user/person are

not interested disclose these attribute to other users(e.g. gender identification, sexual

orientation).For example in an office people connect to each other because of similar

professions. Therefore it is possible that one may be able to infer someone’s attribute

from the attributes of his/her friends. In such cases, privacy is indirectly disclosed by their

social relations rather than from the owner directly. This is called personal information

leakage from inference[10].

The rapid growth and ubiquity of online social media services has given an

impact to the way people interact with each other. Online social networking has become

one of the most popular activities on the web. Social network analysis has been a key

technique in modern sociology, geography, economics, and information science.The data

generated by social media services often referred to as the social network data. In many

situations the data needs to be published and shared with others. Social networks are

online applications allow their users for connection by different linktypes[3]. As part of

their professional network. Because of users specify details which are related to their

professional life.These sites gather extensive personal information social net- work appli-

cation providers have a rare opportunity direct use of this information could be useful to

advertisers for direct marketing. Publish data for others to analyze even though it may

create severe privacy threats or they can withhold data because of privacy concerns even

though that makes the analysis impossible.

For examples business companies are analysing the social connections in social

network data to uncover customer relationship that can benefit their services and prod-

uct sales. The analysis result of social network data is believed to potentially provide

an alternative view of real-world phenomena due to the strong connection between the

actors behind the network data and real world entities. Social-network data makes com-

merce much more profitable[7]. On the other hand the request to use the data can also

come from third party applications embedded in the social media application itself. For

instance, social sites has thousands of third party applications and the number is grow-

ing exponentially. Even though the process of data sharing in this case is implicit the

data is indeed passed over from the data owner (service provider) to different party (the

application).The data given to these applications is usually not sanitized to protect users

privacy.Desired use of data and individual privacy presents an opportunity for privacy

preserving social network data mining. That is the discovery of information and rela-

tionships from social network data without violating privacy. So using classification find

sensitive data and remove it from data set and provide highly sanitize dataset[1].

Although OSNs are quite useful in different sense, there has been some consid-

erations about privacy of users in such services. OSNs are large datastores of personal

information. This information is valuable in the sense that by statistical analysis it is

possible to extract the preference of users based on different criteria such as gender and

marital status. Such analysis can then be used for advertising and research purposes[5].

MET’s Institute of Engineering 2

Preventing techniquesused for profile data prevention inference attack on social network

Third-parties provide targeted advertisement to increase their commercial revenue

using the social platform and customize their promotions exactly based on the preferences

of visitors and increase their chances on marketing[1]. OSN providers such as Facebook

state that they will not hand private information to these third- parties. However there

has been many controversies about leakage of sensitive information to third parties where

OSN providers handed private user information along with self identifying information.

A recent investigation by the Wall Street Journal showed that personal ID of Facebook

users was being transmitted to third party advertisement and tracking companies along

with their personal interests which was against the promises made by Facebook [6]. This

is where concerns are raised about the privacy of OSN users.

The main privacy concern is that members might not be willing to expose their

profile information to everyone inside or outside a network. People need control over

their personal information and how it is being shown on the web. In OSNs users provide

their email address, photos, friends, education, career background, relationship status

and activities such as commenting. For various reasons one might be willing to hide

them from certain people. Reasons such as safety, separation of work environment and

personal life are among them. If the information is public to everyone it can cause

problems such as losing a job. Furthermore it can be collected and used for commercial

purposes without the consent of users[7],[8].

The privacy settings usually does not fully allow hiding friendship links and groups

affiliations and the connection between people and groups are publicly visible. Such links

and affiliations can lead to information leakage and expose high amount of information.

In addition many users do not protect their profiles from strangers and the network would

be a mixture of public and private profiles[13]. As a result while an individual protects

his profile using the privacy settings, it is possible that a large fraction of his friends

have an open profile which contains information about him including the friendship link,

comments and so on. Also even if there are no direct information about a person in his

friends, by statistical analysis it would be possible to infer some attributes for a user even

if he has a private profile which is the topic of this System[14],[15].

The goal of this System mainly highlights how it is possible to infer and reconstruct

private attributes of OSN users based on friendship links and personal details. Using

probability models and data mining approaches such as Naive bays learning, it is shown

that with certain possibilities it would be feasible to infer private attributes of users.

MET’s Institute of Engineering 3

Preventing techniquesused for profile data prevention inference attack on social network

To see the result on a real dataset, a well-known Facebook OSN, Profile data is col-

lected and used. Analysis of system shows that it is possible for an active OSN member

to fully protect its privacy by removal of sensitive attributes from profile data or from

dataset before releasing to third party[1],[11].

1.1 Area of Dersertation

Social networks are considered as online applications that permit the users to connect

by way of various link types. Based on the provided details, these networks let people

to list details about themselves that are appropriate to the fundamentals of the network.

Some site is a common use of social network, therefore individual users list their preferred

activities, movies and books. Conversely a professional network such as LinkedIn, users

specify details which are suited to their professional life.These sites gather extensive

personal information and thus social network application providers have a rare chance of

direct usage of this information that could be useful to advertisers for direct marketing.For

preventing inference attack proposed system is used and it improve the classification

accuracy of system by using Naive bays classification.

MET’s Institute of Engineering 4

Preventing techniquesused for profile data prevention inference attack on social network

1.2 Dissertation Plan

Month Project Activity

August Project Topic Selection

Submission of Abstract

Study of Literature Survey

September First Presentation about idea of Project

Requirement analysis (SRS Document) preparation

October Mathematical Model

Algorithm and System Analysis

Detailed Design

November Project Stage-1 Presentation


December Stating phase of implementation

Requirement gathering for implemntation

January Implementation and testing

February Implementation and testing

March Test cases designing for complete system and testing as per test cases

Changes in implementation if any

April Testing and documentation

May Testing and documentation

Table 1.1: Dissertation Plan

MET’s Institute of Engineering 5

Preventing techniquesused for profile data prevention inference attack on social network

1.3 Motivation

• Privacy to person which is concerned with the integrity of the individuals body,means

prevent the intruder entry in personal data.

• Privacy of personal behavior,This relates to different aspects of behavior such as

sexual preferences, political activities and religious thoughts both in private and

public places.

• Here the effectiveness of both local and relational classification accuracy are reduces

by using the sanitization methods and it is very helpful for preventing personal

information attack on social network.

• Privacy of personal communication in case of individuals have an interest to be able

to communicate among each other through different media without being monitored

or intercepted by other persons or organisations.

• Privacy of personal data, Individuals claim that data about themselves should not

be available to other individuals or organisations without their consent and even if

the data is processed by a third-party, the individual must be able to have consid-

erable degree of control over it data and its use.

• Here it has been proposed to design a system that explore the effect of possible

data sanitization approaches on preventing such private information leakage, while

allowing the recipient of the sanitized data to do inference on non-sensitive details.

• Desired use of data and individual privacy presents an opportunity for privacy

preserving social network, That is the discovery of information and relationships

from social network data without violating privacy .

1.4 Objectives

This system define two classification tasks. The first is that to determine whether a

person is ”conservative” or ”liberal” on the basis of user profile information .

MET’s Institute of Engineering 6

Preventing techniquesused for profile data prevention inference attack on social network

Privacy concerns of individuals in a social network can be classified into two categories:

privacy after data release, and private information leakage. Instances of privacy after

data release involve the identification of specific individuals in a data set subsequent

to its release to the general public or to paying customers for a specific usage or third

party for their advertising work. By inferring the sensitive attribute like gender, marital

status such personal information of user profile is used for different type of attack.Here

objectives of system is privacy concern as hiding the user’s personal details from outside

users means from third party, so information misusage are avoid.