Fake Users Research Articles

Multi-agent Attacks for Black-box Social Recommendations

The rise of online social networks has facilitated the evolution of social recommender systems, which incorporate social relations to enhance users’ decision-making process. With the great success of Graph Neural Networks (GNNs) in learning node representations, GNN-based social recommendations have been widely studied to model user-item interactions and user-user social relations simultaneously. Despite their great successes, recent studies have shown that these advanced recommender systems are highly vulnerable to adversarial attacks, in which attackers can inject well-designed fake user profiles to disrupt recommendation performances. While most existing studies mainly focus on targeted attacks to promote target items on vanilla recommender systems, untargeted attacks to degrade the overall prediction performance are less explored on social recommendations under a black-box scenario. To perform untargeted attacks on social recommender systems, attackers can construct malicious social relationships for fake users to enhance the attack performance. However, the coordination of social relations and item profiles is challenging for attacking black-box social recommendations. To address this limitation, we first conduct several preliminary studies to demonstrate the effectiveness of cross-community connections and cold-start items in degrading recommendations performance. Specifically, we propose a novel framework MultiAttack based on multi-agent reinforcement learning to coordinate the generation of cold-start item profiles and cross-community social relations for conducting untargeted attacks on black-box social recommendations. Comprehensive experiments on various real-world datasets demonstrate the effectiveness of our proposed attacking framework under the black-box setting.

Robust Estimation Method against Poisoning Attacks for Key-Value Data with Local Differential Privacy

Local differential privacy (LDP) protects user information from potential threats by randomizing data on individual devices before transmission to untrusted collectors. This method enables collectors to derive user statistics by analyzing randomized data, thereby presenting a promising avenue for privacy-preserving data collection. In the context of key–value data, in which discrete and continuous values coexist, PrivKV has been introduced as an LDP protocol to ensure secure collection. However, this framework is susceptible to poisoning attacks. To address this vulnerability, we propose an expectation maximization (EM)-based algorithm combined with a cryptographic protocol to facilitate secure random sampling. Our LDP protocol, known as emPrivKV, exhibits two key advantages: it improves the accuracy of statistical information estimation from randomized data, and enhances resilience against the manipulation of statistics, that is, poisoning attacks. These attacks involve malicious users manipulating the analysis results without detection. This study presents the empirical results of applying the emPrivKV protocol to both synthetic and open datasets, highlighting a notable improvement in the precision of statistical value estimation and robustness against poisoning attacks. As a result, emPrivKV improved the frequency and the mean gains by 17.1% and 25.9%, respectively, compared to PrivKV, with the number of fake users being 0.1 of the genuine users. Our findings contribute to the ongoing discourse on refining LDP protocols for key–value data in scenarios involving privacy-sensitive information.

Lexicon-Enhanced Long Short-Term Memory (LSTM) for Detecting of Fake News using Deep Learning

Due to its increasing popularity, low cost, and easy-to-access nature, online social media (OSM) networks have evolved as a powerful platform for people to access, consume, and share news.However, this has led to the large-scale distribution of fake news, i.e., deliberate, false, or misleading information. Spreading fake news is roughly as dangerous as spreading the virus. Fake news detection attracts many researchers' attention due to the negative impacts on the society Over the past years, many fake news detection approaches have been introduced, and most of the existing methods rely on either news content or the social context of the news dissemination process on social media platforms. In this work, we propose a lexicon-enhanced LSTM an automated model that is able to take into account both the news content and the social context for the identification of fake news. The model first uses sentiment lexicon as an extra information pre-training a word sentiment classifier and then get the sentiment embeddings of words including the words not in the lexicon. Combining the sentiment embedding and its word embedding can make word representation more accurate and to detect fake news and better predict fake user accounts and posts.We used five performance metrics to evaluate the proposed framework: accuracy, the area under the curve, precision, recall, and f1-score.The model achieves an accuracy of 99.55% compared to 93.62% against discourse structure analysis. Also, it shows an average improvement of 18.76% against other approaches, which indicates its viability against fake-classifier-based models

Fake User Detection Based on Multi-Model Joint Representation

The existing deep learning-based detection of fake information focuses on the transient detection of news itself. Compared to user category profile mining and detection, transient detection is prone to higher misjudgment rates due to the limitations of insufficient temporal information, posing new challenges to social public opinion monitoring tasks such as fake user detection. This paper proposes a multimodal aggregation portrait model (MAPM) based on multi-model joint representation for social media platforms. It constructs a deep learning-based multimodal fake user detection framework by analyzing user behavior datasets within a time retrospective window. It integrates a pre-trained Domain Large Model to represent user behavior data across multiple modalities, thereby constructing a high-generalization implicit behavior feature spectrum for users. In response to the tendency of existing fake user behavior mining to neglect time-series features, this study introduces an improved network called Sequence Interval Detection Net (SIDN) based on Sequence to Sequence (seq2seq) to characterize time interval sequence behaviors, achieving strong expressive capabilities for detecting fake behaviors within the time window. Ultimately, the amalgamation of latent behavioral features and explicit characteristics serves as the input for spectral clustering in detecting fraudulent users. The experimental results on Weibo real dataset demonstrate that the proposed model outperforms the detection utilizing explicit user features, with an improvement of 27.0% in detection accuracy.

SPAMMER DETECTION AND FAKE USER IDENTIFICATION ON SOCIAL NETWORKS

The project titled "Spammer Detection and Fake User Identification on Social Networks" aims to explore and implement a novel approach to concealing information within digital images while preserving their visual integrity. With the exponential growth of social media platforms like Instagram, the issue of spam and fake user accounts has become increasingly prevalent, posing significant challenges to the integrity and user experience of these platforms. In this project, a comprehensive approach to address the problem of spammer detection and fake user identification on Instagram is proposed. Leveraging machine learning techniques, including Decision Tree Classifier, KNN Classifier, Random Forest Classifier, and Logistic Regression Algorithm, the project aims to develop robust models capable of automatically identifying and flagging suspicious accounts. By analyzing various features such as user behaviour patterns, engagement metrics, and content characteristics, these classifiers will be trained to differentiate between genuine and fake accounts effectively. The project's ultimate goal is to contribute to the enhancement of Instagram's spam detection mechanisms, fostering a safer and more authentic social media environment for users. Key Words: Spam detection, Fake user identification, KNN Classifier, Decision Tree Classifier, Random Forest Classifier, Logistic Regression Algorithm.

Shilling Black-Box Recommender Systems by Learning to Generate Fake User Profiles.

Due to the pivotal role of recommender systems (RS) in guiding customers toward the purchase, there is a natural motivation for unscrupulous parties to spoof RS for profits. In this article, we study shilling attacks where an adversarial party injects a number of fake user profiles for improper purposes. Conventional Shilling Attack approaches lack attack transferability (i.e., attacks are not effective on some victim RS models) and/or attack invisibility (i.e., injected profiles can be easily detected). To overcome these issues, we present learning to generate fake user profiles (Leg-UP), a novel attack model based on the generative adversarial network. Leg-UP learns user behavior patterns from real users in the sampled "templates" and constructs fake user profiles. To simulate real users, the generator in Leg-UP directly outputs discrete ratings. To enhance attack transferability, the parameters of the generator are optimized by maximizing the attack performance on a surrogate RS model. To improve attack invisibility, Leg-UP adopts a discriminator to guide the generator to generate undetectable fake user profiles. Experiments on benchmarks have shown that Leg-UP exceeds state-of-the-art shilling attack methods on a wide range of victim RS models. The source code of our work is available at: https://github.com/XMUDM/ShillingAttack.

Enhanced Recommender Systems with the Removal of Fake User Profiles

The rapid expansion of information resources on the internet has made it increasingly challenging for users to get valuable and relevant information. Recommendation systems (RSs) play a vital role in addressing this issue by providing personalized suggestions based on user preferences. However, the open nature of RSs allows for the injection of fake profiles by malicious users, leading to biased ratings and manipulation of the overall system results. These fake profiles have been intended to promote or degrade specific items, which creates bias in the system and compromises the user experience. The motivation behind these fake profiles is to promote or degrade a particular item, which affects the system and makes it more biased toward certain items. This research paper aims to detect and eliminate such fake profiles in RSs, ensuring users receive genuine and reliable results. The proposed approach utilizes a voting ensemble (VE) method, combining the strengths of multiple classification algorithms including Quadratic Discriminant Analysis (QDA), K-Nearest Neighbor (KNN), and Naive Bayes classifier (NBC). By utilizing these classifiers, the system enhances its robustness against malicious activities. The proposed VE method effectively identifies the fake profiles with a minimum number of parameters, and it shows high detection accuracy on Movielens datasets. It has been demonstrated that the proposed method gives 99.4%, 99.5%, 99.4% and 99.6% accuracy using NBC, QDA, KNN, and VE, respectively and it outperforms all other competing methods in terms of accuracy, precision, recall and MCC.

ClusterPoison: Poisoning Attacks on Recommender Systems with Limited Fake Users

ClusterPoison: Poisoning Attacks on Recommender Systems with Limited Fake Users

Adversarial Attacks for Black-Box Recommender Systems via Copying Transferable Cross-Domain User Profiles

As widely used in data-driven decision-making, recommender systems have been recognized for their capabilities to provide users with personalized services in many user-oriented online services, such as E-commerce (e.g., Amazon, Taobao, etc.) and Social Media sites (e.g., Facebook and Twitter). Recent works have shown that deep neural networks-based recommender systems are highly vulnerable to adversarial attacks, where adversaries can inject carefully crafted fake user profiles (i.e., a set of items that fake users have interacted with) into a target recommender system to promote or demote a set of target items. Instead of generating users with fake profiles from scratch, in this paper, we introduce a novel strategy to obtain “fake” user profiles via copying cross-domain user profiles, where a reinforcement learning based black-box attacking framework (CopyAttack+) is developed to effectively and efficiently select cross-domain user profiles from the source domain to attack the target system. Moreover, we propose to train a local surrogate system for mimicking adversarial black-box attacks in the source domain, so as to provide transferable signals with the purpose of enhancing the attacking strategy in the target black-box recommender system. Comprehensive experiments on three real-world datasets are conducted to demonstrate the effectiveness of the proposed attacking framework.

Influence-Driven Data Poisoning for Robust Recommender Systems.

Recent studies have shown that recommender systems are vulnerable, and it is easy for attackers to inject well-designed malicious profiles into the system, resulting in biased recommendations. We cannot deprive these data's injection right and deny their existence's rationality, making it imperative to study recommendation robustness. Despite impressive emerging work, threat assessment of the bi-level poisoning problem and the imperceptibility of poisoning users remain key challenges to be solved. To this end, we propose Infmix, an efficient poisoning attack strategy. Specifically, Infmix consists of an influence-based threat estimator and a user generator, Usermix. First, the influence-based estimator can efficiently evaluate the user's harm to the recommender system without retraining, which is challenging for existing attacks. Second, Usermix, a distribution-agnostic generator, can generate unnoticeable fake data even with a few known users. Under the guidance of the threat estimator, Infmix can select the users with large attacking impacts from the quasi-real candidates generated by Usermix. Extensive experiments demonstrate Infmix's superiority by attacking six recommendation systems with four real datasets. Additionally, we propose a novel defense strategy, adversarial poisoning training (APT). It mimics the poisoning process by injecting fake users (ERM users) committed to minimizing empirical risk to build a robust system. Similar to Infmix, we also utilize the influence function to solve the bi-level optimization challenge of generating ERM users. Although the idea of "fighting fire with fire" in APT seems counterintuitive, we prove its effectiveness in improving recommendation robustness through theoretical analysis and empirical experiments.

Poisoning QoS-aware cloud API recommender system with generative adversarial network attack

With the proliferation and deepening of service-oriented architecture, more and more enterprises and organizations are exposing their computing functions and big data to the Internet in the form of cloud APIs to support service-oriented software development. This has resulted in a plethora of cloud APIs with similar functionality appearing on the Web, drowning users in a sea of cloud API choices. To solve this problem, quality of service (QoS)-aware recommender system is then widely applied to the selection of cloud APIs. Due to the dynamic and open network environment, the QoS-aware cloud API recommender systems are vulnerable to data poisoning attacks, where attackers inject poisoned data to skew the recommender system and make the recommendation direction follow the attacker's will. Given the lack of data poisoning attack methods and robustness analysis for existing QoS-aware cloud API recommender systems, in this work, we first built a general poisoning attack framework for QoS-aware cloud API recommender systems to elucidate and standardize the attack process. Then, we proposed a deep learning-based poison attack approach, which uses generative adversarial network (GAN) to learn the cloud API QoS data distribution of real users in an adversarial way, so as to generate high-quality fake user attack vectors. We conducted extensive experiments on real-world QoS datasets, and the experimental results show that our proposed GAN-based poisoning attack is effective and can better hide itself from being detected. In addition, we analyzed the data poisoning attack mechanism and the robustness of the cloud API recommender system based on four categories of twelve recommendation methods, thereby raising awareness about the security of cloud API recommendation and helping the recommender system defenders to develop more targeted defense strategies.

Evaluating and Detecting Fake Users in Social Media by Random Forest

Currently, users have been engaging in conversations, sharing information and producing web content via social media platforms. But in recent times, many users have been using these platforms to conduct identity faults, payment frauds, and many more without the knowledge of the actual user. For example: - On Instagram, according to the latest analysis, there are around 95 million fake accounts compared to the total number of users, which amount to 1 billion. Therefore, there are nearly 10% of fake accounts active at present. The obtained dataset lies approximately in thousands. Hence, we used GANs and deep learning to broaden the data to around 1 lakh. The conventional methods used for distinguishing between real and fake accounts were ineffective. Adopting machine learning-based approaches allowed us to identify fake accounts that can mislead users. The dataset is pre-processed using several Python tools, and a comparison model is created to identify a practical solution appropriate for the dataset that has been provided.

Detection of fraudulent marketing of consumer product in social networking services

In recent days, malicious users try to captivate the consumers using their fraudulent marketing URL post in social networking sites. Such malicious URL posted by fake users in Social Networking Services (SNS) is hard to identify. Therefore, there occurs a need to detect such fraudulent URLs in SNS. In order to detect such URLS, this paper proposes a SNS Fraudulent Detection (SFD) scheme. The proposed SFD scheme includes a Deterministic Finite Automata Tokenization (DFA-T) and Web Crawler (WC) based Neuro Fuzzy System (WC-NFS). DFA-T extracts the URL features and calculates a Penalty Score (PS) based on the malicious words in the extracted URL. The DFA extracted URL features with PS are fed into WC-NFS. Subsequently, the WC fetches the numeric WC-Index (WCI) value from the URLs which are added to the WC-NFS. The existing URL data set is used to identify the malicious web links and suitable machine learning techniques are used to identify the malicious URLs. From the experimental results, it is found that the proposed SFD provides 92.6 % accuracy in classifying the benign from malicious URLs when compared with the existing methods.

Face recognition based physical layer security system for next-generation wireless communication

Security system in every sector all over the world is a most demandable and crucial issue to protect the fake user especially in wireless communication system. The most common feature in current wireless communication system is only transmitted users’ information; consequently, it is difficult to identify the actual sender to attain the desired security. Face recognition is one of the most popular security systems in biometric authentication, which can detect intruders to restricted or high-security areas in wireless communication sector, and help in minimizing the face user. In this paper, we proposed a system that is the senders will transmit message information with their image to confirm the original senders. A face-recognition algorithm will be implemented at the receiver end. When a senders’ transmitted information enters to the receiver section, the image of sender will be separated from original message signal and sent to the face-recognition algorithm to be analysed and compared with an existing database of trusted user. An alarm goes off if the user is not recognized.

Practical Cross-System Shilling Attacks with Limited Access to Data

In shilling attacks, an adversarial party injects a few fake user profiles into a Recommender System (RS) so that the target item can be promoted or demoted. Although much effort has been devoted to developing shilling attack methods, we find that existing approaches are still far from practical. In this paper, we analyze the properties a practical shilling attack method should have and propose a new concept of Cross-system Attack. With the idea of Cross-system Attack, we design a Practical Cross-system Shilling Attack (PC-Attack) framework that requires little information about the victim RS model and the target RS data for conducting attacks. PC-Attack is trained to capture graph topology knowledge from public RS data in a self-supervised manner. Then, it is fine-tuned on a small portion of target data that is easy to access to construct fake profiles. Extensive experiments have demonstrated the superiority of PC-Attack over state-of-the-art baselines. Our implementation of PC-Attack is available at https://github.com/KDEGroup/PC-Attack.

MPKIX: Towards More Accountable and Secure Internet Application Services via Mobile Networked Systems

Nowadays, both Internet Application Service (IAS) providers and users face various security threats and legal issues. Due to the lack of reliable user information verification mechanisms, adversaries can abuse IASs to launch various cyberattacks, such as misinformation distributing and phishing, by using fake user accounts. IAS providers may thus inadvertently offer inappropriate content to restricted users, thereby suffering a serious risk of prosecution under local or international laws. Also, IAS users may suffer from nefarious ID theft attacks. In this paper, we proposed a novel security framework, MPKIX, designated as Mobile-assisted PKIX (Public-Key Infrastructure X.509). MPKIX secures both IAS providers and users by leveraging the broadly used PKIX services and mobile networked systems. It not only provides IAS providers with a reliable user verification mechanism while simultaneously enabling cross-IAS user privacy protection, but also largely mitigates the possibility of ID theft attacks and benefits other involved parties, such as cellular network operators and PKIX service providers. We further conduct a security analysis of MPKIX and implement an MPKIX prototype. The evaluation results based on the prototype confirm the effectiveness and efficiency of MPKIX with low overhead.

Spam Detection and Fake User Identification in Twitter: An Analysis of Machine Learning Models

Spammer detection and fake user identification are significant issues in the realm of social media, with Twitter being no exception. The detection of spammers and fake users on Twitter is essential for preserving the platform's integrity and protecting users from online scams and fraud. This project paper aims to conduct a comprehensive study of the different techniques and algorithms used for spammer detection and fake user identification on Twitter. We will evaluate the effectiveness of traditional techniques and machine learning-based methods and propose a novel approach combining both methods for spammer detection and fake user identification on Twitterpils.

Spammer Detection and Fake User Identification on Social Networks

Spammer Detection and Fake User Identification on Social Networks

Spammer Detection and Fake User Identification on Social Networks

Spammer Detection and Fake User Identification on Social Networks

Identification of Source of Misleading Information and Stop the Dissemination through Blocking the User

Introduction: At present, people are more dependent on Internet sources for any sort of information or news. So, the news/information needs to be preserved and should not be modified by any user. Providing security for the news data is a major concern. The decentralized approach of a chain of blocks is used in order to strengthen the security of the news. The existing blockchain framework that offers openness, tamper-proofing, privacy, controlling information, and monitoring is inherited in the proposed work. Precisely, the idea is to build a safe platform that can detect bogus news on social media platforms. Even if the environment is fragile, the chain of blocks-based decentralised peer-to-peer environment provides security to the published information.
 Objectives: As a result of recent innovations and advancements in the field of computer technology, social media networks have emerged as one of the most crucial aspects of contemporary human existence. Social media has developed into a well-known platform for information dissemination and news, as well as for daily reports. There are a variety of benefits associated with social media; but, on the converse, there is a great deal of misleading news and data that can mislead the reader. One of the major issues with social media is that there is a dearth of information that can be relied on as well as real world news. Because of misleading news on social media, users are misled. So, to build a trustful environment, early detection of misleading news is necessary. Innovative machine learning methodologies are useful to identify and recognize misleading news more accurately.
 Methods: Misleading news is more viral than real news. People instantly believe on the false information. So, there is a need to reduce the dissemination of misleading information on social media. In order to minimize the spread of misleading news, the source of the news needs to be traced. In overall, proposed system utilizes the chain of blocks and applies proposed machine learning methodologies in order to identify misleading news and thereafter reduce the propagation of misleading information by blocking the fake user. 
 Results: An experimental analysis reveals that the proposed classification algorithm obtains a better accuracy rate. In order to produce useful training rules and evaluate the test classifier, a number of features are extracted like TF-IDF, N-Gram features, and dependency-oriented NLP features from the data input.
 Conclusions: The proposed method analyzes every user's uploaded information, identify fraudulent users and reduce the propagation of false information by blocking the user.