Malicious Users Detection in Crowdsourced Social Networks
Malicious Users Detection in Crowdsourced Social Networks
ABSTRACT:
The past few years have witnessed the dramatic popularity of large-scale social networks where malicious nodes detection is one of the fundamental problems. Most existing works focus on actively detecting malicious nodes by verifying signal correlation or behavior consistency. It may not work well in large-scale social networks since the number of users is extremely large and the difference between normal users and malicious users is inconspicuous. In this paper, we propose a novel approach that leverages the power of users to perform the detection task. We design incentive mechanisms to encourage the participation of users under two scenarios: full information and partial information. In full information scenario, we design a specific incentive scheme for users according to their preferences, which can provide the desirable detection result and minimize overall cost. In partial information scenario, assuming that we only have statistical information about users, we first transform the incentive mechanism design to an optimization problem, and then design the optimal incentive scheme under different system parameters by solving the optimization problem. We perform extensive simulations to validate the analysis and demonstrate the impact of system factors on the overall cost.
PROJECT OUTPUT VIDEO:
EXISTING SYSTEM:
-
Most existing works are concerned with actively detecting malicious nodes. One approach is focusing on data received by sensors. If some data do not meet the certain criteria such as spatial correlation or frequency correlation, there may be malicious nodes. For example, sensory data in wireless sensor networks are usually location dependent. The malicious nodes can be identified if their reported data are far discrepant from that of nearby sensor nodes.
-
Stringhini et al. focused on the malicious users detection in social network using big data, studied the users’ characteristics and then built a tool to detect spammers.
-
In other existing system the authors used a classifier to identify malicious users. Firstly, the authors collected about 22 thousand users’ data set from Twitter site and then manually created a label for each user. Then authors studied the difference between normal users and malicious users and exploited two parameters (user score and tweet score) from these data to classify users in the network.
DISADVANTAGES OF EXISTING SYSTEM:
-
Different from the existing systems our work focuses on the scenario where the malicious user can not be easily detected by the system administrator or data correlation of nearly users. And we propose a novel crowdsourcing based approach to tackle the malicious users detection problem. The malicious users in social networks have a terrible impact on the network, in terms of degrading the network’s performance, reducing the network’s efficiency, increasing the cost or even disabling the whole network.
-
One fundamental issue in crowdsourcing based approach is incentive mechanism design. Since different users have different preferences for these malicious activities, many users may choose to stay silent without a proper incentive.
-
Further, malicious users may provide compensation for the victims to keep them silent. For example, a malicious user may send an advertisement to user alongside with a coupon or monetary reward. In such case, incentive provision is critical to encourage the participation of users.
PROPOSED SYSTEM:
-
In this paper, we propose an approach to detect malicious users in large-scale social networks from a radical new perspective. The system administrator is not directly participated in the detection process. Instead, it leverages the power of normal users in the social networks to accomplish such a difficult goal, i.e., crowdsourcing the detection tasks to the users.
-
When malicious users perform abnormal activities such as cyber attack or advertisement injection, the users who are the victims of these activities can report them to the system administrator. Obviously, in such a way, the detection cost for malicious cost can be significantly reduced since no additional overhead is incurred. Also, the detection accuracy can be increased.
-
To existing system issues, we investigate the incentive mechanism to encourage the user participation in the malicious user detection in a large-scale social network. Interestingly, we consider that the malicious users may provide incentives to the normal users when it performs malicious activities (cyber attack, advertisement injection, etc) towards user ui. For example, if a malicious user wants to get users’ profile information, providing some incentives can keep more users silent. Besides, users’ preferences are typically different for malicious activities. Some users are more tolerant of advertisement injection than other users. We adopt contract theory to tackle our problem i.e., we construct contractual arrangements as incentive mechanism for system administrator to encourage users to help detect the malicious user
ADVANTAGES OF PROPOSED SYSTEM:
-
We introduce a novel, efficient, and effective approach, i.e., crowdsourcing, to detect malicious users in lagerscale social networks. Based on this, in order to encourage sufficient users to perform detecting tasks, we formulate the incentive mechanism design problem.
-
We solve the incentive mechanism design problem in two scenarios: full information of users’ preferences and partial information of users’ preferences. In full information scenario, we design the optimal incentive mechanism by ordering users’ preferences. In partial information scenario, assuming that we only have statistical information about users’ preferences, we transform this problem to an optimization problem and solve it by exploring the form of its solution.
-
We perform extensive simulations to illustrate the relationship between the system’ total cost and factors, and validate our analysis.
MODULES:
- System Administrator
- Users
- Incentive Scheme
MODULE DESCRIPTIONS:
-
System Administrator
When the network size N is to a great degree extensive, it is troublesome for the system administrator to detect the malicious user without anyone else/herself. Accordingly, the system administrator needs to design an incentive mechanism that encourages all users in the network to take an interest in the detection of the malicious user. To abstain from being detected, the malicious user will likewise give incentives to a user ui when he/she builds up a link with user ui. A link between the malicious user and user ui could be a cyber attack or advertisement injection. We define the incentives as B which is a constant, in light of the fact that the malevolent user cannot distinguish the difference of users.
-
Users
Users themselves have their own preferences when the malicious user establishes a link with them. So for the so-called malicious user, they will have different response. For example, if the malicious user is trying to promote products to potential customers, their potential customers who have corresponding requirements will have a favorable impression while other users will not like it that much. We denote the preference of each user ui by pi. It is positive when ui has a favorable impression on the malicious user and negative when ui thinks it is annoying. We assume that for each ui it exactly knows its pi and it has no knowledge of other users’ preference.
-
Incentive Scheme
The system administrator has to decide an incentive scheme to encourage the report of malicious node from users. We define that ui’s incentive is ci and ci >0. Note that the incentives that the system administrator provides vary from person to person. The reason is that the system administrator can have access to some prior information about users in the system so that its incentives can differ as different users’ preferences differ. Here we assume that the system will give out its incentives only when there are more than N0 users reporting the malicious user, where N0 is a predefined threshold. It will cause dishonesty that giving incentives as soon as they report because in this way, users will report all other users including normal users to get a higher payoff. And another assumption is that the system administrator ensures that if each user does as the incentive scheme says, the system administrator can induce N0 users and each user’s payoff will be maximized. N0 should be chosen such that the probability that users in the network report others arbitrarily and finally get the incentive provided by the system is very small.
SYSTEM REQUIREMENTS:
HARDWARE REQUIREMENTS:
- System : Pentium i3 Processor
- Hard Disk : 500 GB..
- Monitor : 15’’ LED
- Input Devices : Keyboard, Mouse
- RAM : 4 GB.
SOFTWARE REQUIREMENTS:
- Operating system : Windows 10/11.
- Coding Language : C#.net.
- Frontend : Net, HTML, CSS, JavaScript.
- IDE Tool : VISUAL STUDIO.
- Database : SQL SERVER 2005.