User Identification Framework in Social Network Services Environment

Social Network Service is a one of the service where people may communicate with one another; and may also exchange messages even of any type of audio or video communication. Social Network Service as name suggests a type of network. Such type of web application plays a dominant role in internet technology. In such type of online community, people may share their common interest. Facebook LinkedIn, orkut and many more are the Social Network Service and it is good medium of making link with people having unique or common interest and goals. But the problem of privacy protection is a big issue in today’s world. As social networking sites allows anonymous users to share information of other stuffs. Due to which cybercrime is also increasing to a rapid extent. In this article we preprocessed the web log data of Social Network Services and assemble that data on the basis of image file format like jpg, jpeg, gif, png, bmp etc. and also propose a framework for victim’s identification.


Introduction
Now days we can see social networking services are growing day by day, First of all the term of social network is a social structure made of individuals (or organizations) called "nodes," which are tied (connected) by one or more specific types of interdependency, such as friendship, kinship, financial exchange, dislike or relationships of beliefs, knowledge or prestige [8].Social networking is the grouping of individuals into specific groups, like small rural communities or a neighbourhood subdivision, if you will [3].Generally some social services focus on the privacy issue too.Due to which one person could not see the friend link of other one.There are various types of social network are available.Personal networks-These kinds of network allow creating profiles online and linked with other person, with a focus on social linkage such as friendship.For example, Orkut, LinkedIn, Facebook, MySpace are stage for communicating with contacts [9][4] [13].These kinds' networks involve users exchanging information with people.Statuses update networks-These types of social networks are developed to allow small status update to person in order to communicate with other persons.For example, Twitter, these types of network people may exchange messages or thoughts day by day.Location networks-These networks are developed to show one's real-time location, either as public information or as an update viewable to authorized contacts.Content sharing networks-These networks are designed as platforms for sharing content, such as music, photographs and videos.For example YouTube and Flicker.Shared interest networks-Some social networks are built around a common interest or geared to a specific group of people.These networks incorporate features from other types of social networks but are slanted toward a subset of individuals, such as those with similar hobbies, educational backgrounds, political affiliations, ethnic backgrounds, religious views, sexual orientations or other defining interests.For example LinkedIn [16].But now a day's these social sites are suffering from the critical problem of cyber crime.This thing is happening due to open space functionality of these sites.There are many internet applications which are working today.If any person commits a cyber crime and interact with the social networking sites then by identifying the different activities of that particular person we can jump to its complete group even those being included in that cyber crime [11].
For this work the log data of the person is collected which gets available on the server by reading this log data we can suspect the culprit.The log file is available at The This log data later partitioned on the basis of its attributes which can be done by applying data mining techniques.By doing this if any person proves out to be criminal then its complete group who was included in that crime could be traced.For this work a special technique for data mining that is frequent pattern mining is used.When a particular user click or visit the particular link then the all activity which has been performing by the user are have shown on the Fig. 2 identified through log data and maintained by the server or database.Some social networking sites follow the unauthorized showing mechanism.And its overall content follows the mechanism or open space.Being an open source code any unauthorized user to these particular sites can click on any knowing or unknowing link [6][12] [2].Apart of it he can send any types of message too.In a broad area cyber crime is also called as computer related crime this is an illegal work committed by any means, or in relation to a computer system or network.Including such kind of illegal distribution information by means of computer system or network.Cyber crime is of many types against any person, against any business or non business organization committing crime by targeting the government [17].There are few more examples of it like financial claims, money laundering, credit card frauds, sale of illegal articles including weapon and wild life article, posting any article to your website brought from some other needs.Intellectual crime includes software piracy, copyright stuffs.Email spooling include sent the mail from one source to another source.Forgery include currency host stamp, making marks sheet through printer and scanner.
2 Literature Review kim W. et al [8] attempt to organize the status, uses and issues of social websites into comprehensive framework for discussing, understanding, using building and forecasting the future of social websites.Chu H. et al [3] provide the real case review concerning the crime sense reconstruction with respect to the previous facebook session of the victim based on the digital evidence collected and analyzed via live internal data acquisition and they also identified victim according to its log data which is maintained by the server.They have converted that log data to hexadecimal number and also found its mutual friends who also involved with the victim.If a person involved in a crime, according its session their conversation may be happened and they mined that data and identified victim and his friends.They also developed a framework of social network according to facebook sessions.Wu H. et al. [16] analyze the characteristics of users in social networking websites as well as the related contents of the website.They provide a technique of social network analysis and web mining to illustrate the networks of the blog users and according to that they find out the interest group.Ortigosa A. et al [10] Shown the data from online social networks profiles, specifically facebook profiles, can be mined in order to build classifier able to predict user personality and they also developed TP2010 applications which is general questioners to get the user personality and also applied classification tree.Jin L. et.al [7] we show that it also raises significant privacy concerns as an adversary can use it to find out some or all of the victim's friends, although, as per the privacy settings of the victim, the adversary is not authorized to see his friend list directly.We show that by using mutual friend queries, an attacker can launch privacy attacks that we refer to as mutual-friend based attacks to identify friends and distant neighbours of targeted users.We analyze these attacks and identify various attack structures that an attacker can use to build attack strategies, using which an attacker can identify a user's friends and his distant neighbours.Through simulations, we demonstrate that mutualfriend based attacks are effective.For instance, one of the simulation results show that an attacker using just one attacker node can identify more than sixty percent of a user's friends.DOI: 10.12948/issn14531305/18.2.2014.02hexadecimal number it become memory efficient and makes the computation fast comparatively.We also used it by the point of view of cryptography so that any person can hit and interact directly with your data.In hexadecimal conversion bit modification becomes quite easy.For analyzing the log data maintained by the server, or if any person is involved in criminal case, then for identifying that person, it is really necessary to find a common pattern in that.Which can only be possible by analyzing the different attributes of log data?In case of cyber crime in Social Networking Services play a vital role as a clue.It consists of many steps that could be helpful in this kind of investigation.

User Activity
When a user click or visit the particular link then the all activity which has been performing by the user are identified through log data and maintained by the server or database.

Web Log Collection
In this step, the time at which any crime had taken place on any social networking sites, the complete data of that particular duration is fetched.For example if on 15-Jan -2012 any crime get conducted then we can take the complete data of that day for investigation.In this data, different type of attributes are inbuilt like IP address, user name, timestamp, access request, status code, referrer etc.

Preprocessing of Web Log
According to the given time, for analyzing the current log, we will use the concept of data mining in which we will apply the data preprocessing technique [15] [14].Through preprocessing technique this log data could be simplified to a lot.Data can be cleaned through it or we can say noise could be removed from data.

Analyze Access Page Attribute
After splitting the log data separately, here only a column 'access page' is analyzed in which many type of information is present.DOI: 10.12948/issn14531305/18.2.2014.02

Hexadecimal Conversion
Whichever images have been identified, the page related to that image will be converted to hexadecimal number due to which it could be easily processed.In terms of cryptography it would be easier for privacy and modification of bits is also easier for computation, and we will take only those image files which exist in that session and will also count the occurrence of that page that how many times that page has occurred in that session?We displayed a table here according to the user in which it is shown that how many users hit which all image?At what time?

Result
After counting an occurrence of all images we have to analyze user who visited or viewed that images and apply data mining techniques we have identified restricted users.By applying all above steps we have proposed architecture which is shown in Figure 3.

Experimental Result
First of all we have rearranged our web log data according to its attribute as we shown in fiure4.Every user may registered in social networking services like facebook, orkut, LinkedIn etc. and that kinds of service provide a unique number or unique Id for a particular user.But through same IP address of a system any number of user can login to various social networking services, through IP address it is very complex to discriminate user exist in social networking services.

Fig. 3. Identify Images from Web Log
Converting web log in to hexadecimal number due to which it could be easily processed.In terms of cryptography it would be easier for privacy and modification of bits is also easier for computation as we shown in the  Any user which is in social network can visit same or different image when same user hit a particular image again and again all activities are resides in the server or database in terms of web log with different timings .In table 1 we showed the visited image by the user with different timing.We are showing a snippet of visited image.Suppose a crime gets conducted 15-jan-2005 then we can take all logs on that day for investigation.An investigation depends on how many users visited a particular image?At what time?Etc.For splitting our log data we have accomplished that task using java programming.Any user can interact with different kind of social networking services moreover using same IP we can interact with various social networking services but their timings are different.As we shown that Fig .showing number of pages which had visited by the user.When a crime gets conducted then we have analyzed all pages of access page on that day and get how many people involved on that day?Page column contain all activities which is done by the user.Any number of users is directly or indirectly linked together.We have divided our log data one attribute name as timing.Whenever an user viewed images then log maintain their timing.If any criminal activities get conducted, then we have included all data with respect to its timing.posting of images etc.In this research, on the basis of web log we have identified the victim, for that we have analyzed different attributes of web log.If there is any criminal activity performed in any social site then we will take the web log of that day and from that log data it can be identified the user's behaviour on the basis of page click by users.For that we have preprocess that web log.In this we have find restricted user on the basis of number of hits on an image.Which is used as to restrict unauthorized access to social sites and it will also helpful for social network providers.

1 DOI: 10 Figure 1
Figure 1 shows that the social relationships among different users.In social networking services any person can interacts another person we have shown the person using dot in the give Fig. and maintained its relationships through lines.In this network any number of users (u1, u2, u3 ...) can communicate each other.But their exists some services that do not suffer from such kind of issues like LinkedIn Facebook, orkut etc. 'Mutual friend' is that feature existing on such services which provides a way of connecting through other person.Mutual friends are the persons who are Social Networking Service (SNS) friends with both you and the person whose timeline you are viewing.Or we can say mutual friends are the common friends of a person.Suppose u1 is friend of u10 and u14 then u1 is mutual friend of u10, u14.According to Fig. 1 we are showing in the table1 of mutual friends of an individual user.

Table 2 .
An information of web log