Abstract
Existing shilling attack detection approaches focus mainly on identifying individual attackers in online recommender systems and rarely address the detection of group shilling attacks in which a group of attackers colludes to bias the output of an online recommender system by injecting fake profiles. In this project, we propose a group shilling attack detection method based on the bisecting K-means clustering algorithm. First, we extract the rating track of each item and divide the rating tracks to generate candidate groups according to a fixed time interval. Second, we propose item attention degree and user activity to calculate the suspicious degrees of candidate groups. Finally, we employ the bisecting K-means algorithm to cluster the candidate groups according to their suspicious degrees and obtain the attack groups. The results of experiments on the Netflix and Amazon data sets indicate that the proposed method outperforms the baseline methods.
Existing System
- Zhou et al. improved the individual shilling attack detection metrics and proposed a two-step method to detect group shilling attacks. While this method is effective for detecting group attacks in synthetic data sets, it is not effective in detecting group attacks with a low similarity between attackers.
- Wang et al. improved several traditional features and proposed a method for group attack detection based on these features. They first manually labeled candidate groups with high minimum support, and thereafter, they computed the group metrics and employed PCA to rank the candidate groups. Unfortunately, this method is only suitable for detecting group attacks whose shilling profiles have high similarity.
Proposed System
- We propose a method to detect group shilling attacks in online recommender systems through bisecting K-means clustering. The major contributions of this article are listed as follows.
- We propose a candidate group division method, which first mines the rating tracks of items and then divides the users in the item rating tracks (IRTs) into multiple groups according to a certain length of time. Since the attackers in an attack group must rate the target item(s) within a certain period of time, the proposed candidate group division method is more likely to divide the attackers in an attack group together, which can lay a good foundation for the group shilling attack detection.
- We propose metrics of item attention degree and user activity (UA) to analyze the candidate groups, making the judgment of attack groups more accurate. Based on the divided candidate groups, the item attention degree and the UA for each candidate group are calculated, and the suspicious degrees of these groups are obtained. Based on this, the bisecting K-means algorithm is employed to cluster the candidate groups according to their suspicious degrees, and the attack groups are obtained.
- To evaluate the performance of our method, we conduct experiments on the Netflix and Amazon data sets and compare the proposed method with four baseline methods.
Hardware Requirements
- Processor: Intel® Pentium i3 (or above)
- System Type: 64-bit Operating System
- Storage: 500 GB HDD/SSD
- Monitor: 15’’ LED (or above)
- Input Devices: Keyboard, Mouse
- Memory (RAM): 2 GB DDR (minimum)
Software Requirements
- Operating System: Windows 10 (or above)
- Programming Language: Java
- IDE / Tool: NetBeans 8.2 (or above)
- Database: MySQL or Microsoft SQL Server
Architecture Diagram
Project Modules
- Intelligent Mining and Secure Login
- Product Addition and Management
- User Feedback Collection and Review
- Product Viewing and Ratings Distribution
- One-Side Rating Shilling Attack Impact
- Simultaneous Rating Shilling Attack Impact
- Remedial Rating Techniques and Effectiveness
Components of Project Report
- Abstract
- Table of Contents
- List of Tables
- List of Figures
- Chapters
- Introduction
- Literature review
- Problem definition and requirement analysis
- Design and Implementation
- Testing and deployment
- Future enhancements
- Summary
- References
Project Report Pages : 80
Can be used in : Java
Delivery Time : Within 2 hours.
Support / Query : Call +91-7449000533
Email [email protected]
