Web Spam Research Articles

PRADA: Practical Black-box Adversarial Attacks against Neural Ranking Models

Neural ranking models (NRMs) have shown remarkable success in recent years, especially with pre-trained language models. However, deep neural models are notorious for their vulnerability to adversarial examples. Adversarial attacks may become a new type of web spamming technique given our increased reliance on neural information retrieval models. Therefore, it is important to study potential adversarial attacks to identify vulnerabilities of NRMs before they are deployed. In this article, we introduce the Word Substitution Ranking Attack (WSRA) task against NRMs, which aims at promoting a target document in rankings by adding adversarial perturbations to its text. We focus on the decision-based black-box attack setting, where the attackers cannot directly get access to the model information, but can only query the target model to obtain the rank positions of the partial retrieved list. This attack setting is realistic in real-world search engines. We propose a novel Pseudo Relevance-based ADversarial ranking Attack method (PRADA) that learns a surrogate model based on Pseudo Relevance Feedback (PRF) to generate gradients for finding the adversarial perturbations. Experiments on two web search benchmark datasets show that PRADA can outperform existing attack strategies and successfully fool the NRM with small indiscernible perturbations of text.

Efficient index-free SimRank similarity search in large graphs by discounting path lengths

Link-based similarity search aims to find similar nodes for a given query node in a graph, which arises in numerous applications, including web spam detection, social network analysis and web search. Among existing methods, SimRank is a well-known similarity model, which provides an effective and trustful function for similarity search. A large amount of techniques on SimRank similarity search are devoted recently, which compute the similarity scores by traversing the paths between query and candidate nodes. However, the number of paths increases exponentially as path length increases, which makes the computation expensive and cannot support fast similarity search over large graphs. In this paper, we propose an efficient index-free SimRank similarity search approach, namely DisSim, which reduces the computational cost by discounting path length. We observe that SimRank could rapidly converge at a stable state and the results change little after a few of iterations. Based on the fast convergence, the similarity between nodes is defined as the SimRank score at the second iteration. For the computation of DisSim, we divide the similarity into one-step and two-step first-meeting probabilities. The one-step first-meeting probabilities are computed by path traverses from query to candidate nodes, which reduces computational cost by skipping unnecessary nodes. And the two-step first-meeting probabilities are computed by integrating the repeated parts of the paths. For further speeding up query processing, we develop a pruning algorithm, which prunes unpromising path traverses by setting a threshold, and the accuracy loss under threshold is given through mathematical analysis. Extensive experiments on real graphs demonstrate the performance of DisSim through comparing with the state-of-the-art algorithms.

SecureEngine: Spammer classification in cyber defence for leveraging green computing in Sustainable city

In today’s era, Sustainable city is evaluated using the services provided to the society. The designing of the integral part of the society should be focused towards the benefits of people. Internet is extensively utilized by the society using Search Engines. The accuracy and time it takes for different search engines to retrieve information from a cloud computing repository around the world, varies. However, it has been discovered in the literature that webpage ranking reduces the amount of time a user spends surfing, which saves a significant amount of energy during computation and transmission across the network. The hyperlink structure of the web graph is used in most of the earlier solutions documented in the literature, which consumes a lot of energy during calculation. It may exacerbate the link leakage problem by increasing the frequency of spam pages. In light of the energy consumption of various smart gadgets, hyperlink structure alone is no longer sufficient for predicting webpage relevance. Its true importance is revealed by user surfing activity. To improve search engine accuracy and speed, it is critical to demote spam pages, lowering energy consumption. Among all the existing ranking algorithms in the literature, one of the important components is the PageRank algorithm used by Google’s ranking module. Keeping focus on these points, in this paper, various page ranking algorithms based upon supervised learning are surveyed and summarized with respect to different selected parameters and experiments performed. Using this information, a detailed taxonomy of search engine results is presented in the text. Moreover, PageRank algorithms are explored by using different supervised learning techniques applied in the existing proposals for getting and processing the results. In the nutshell, the PageRank methodology is surveyed with respect to web spam detection which is the demand of cognitive systems in smart cities.

E-mail Spam Classification Using Grasshopper Optimization Algorithm and Neural Networks

Spam has turned into a big predicament these days, due to the increase in the number of spam emails, as the recipient regularly receives piles of emails. Not only is spam wasting users’ time and bandwidth. In addition, it limits the storage space of the email box as well as the disk space. Thus, spam detection is a challenge for individuals and organizations alike. To advance spam email detection, this work proposes a new spam detection approach, using the grasshopper optimization algorithm (GOA) in training a multilayer perceptron (MLP) classifier for categorizing emails as ham and spam. Hence, MLP and GOA produce an artificial neural network (ANN) model, referred to (GOAMLP). Two corpora are applied Spam Base and UK-2011 Web spam for this approach. Finally, the finding represents evidence that the proposed spam detection approach has achieved a better level in spam detection than the status of the art.

An Improved Framework for Content- and Link-Based Web-Spam Detection: A Combined Approach

In this modern era, people utilise the web to share information and to deliver services and products. The information seekers use different search engines (SEs) such as Google, Bing, and Yahoo as tools to search for products, services, and information. However, web spamming is one of the most significant issues encountered by SEs because it dramatically affects the quality of SE results. Web spamming’s economic impact is enormous because web spammers index massive free advertising data on SEs to increase the volume of web traffic on a targeted website. Spammers trick an SE into ranking irrelevant web pages higher than relevant web pages in the search engine results pages (SERPs) using different web-spamming techniques. Consequently, these high-ranked unrelated web pages contain insufficient or inappropriate information for the user. To detect the spam web pages, several researchers from industry and academia are working. No efficient technique that is capable of catching all spam web pages on the World Wide Web (WWW) has been presented yet. This research is an attempt to propose an improved framework for content- and link-based web-spam identification. The framework uses stopwords, keywords’ frequency, part of speech (POS) ratio, spam keywords database, and copied-content algorithms for content-based web-spam detection. For link-based web-spam detection, we initially exposed the relationship network behind the link-based web spamming and then used the paid-link database, neighbour pages, spam signals, and link-farm algorithms. Finally, we combined all the content- and link-based spam identification algorithms to identify both types of spam. To conduct experiments and to obtain threshold values, WEBSPAM-UK2006 and WEBSPAM-UK2007 datasets were used. A promising F-measure of 79.6% with 81.2% precision shows the applicability and effectiveness of the proposed approach.

Artificial Intelligence and Edge Computing-Enabled Web Spam Detection for Next Generation IoT Applications

For next generation IoT applications, edge devices provides most of the computing resources close to the proximity of the end users. These devices having built-in intelligence using various AI techniques can take independent decisions in the environment where these are deployed. Motivated from these concerns, We suggest a cognitive intrusion security system to maintain the credibility of search engine results, which eliminates the advertising images from penetrating the image database of the web browser. The proposed framework provides edge intelligence for web data filteration and detects the web spam by considering three different layers, i.e., data collection services, edge computing services, and cloud services. The target is to detect the malicious images. Firstly, the features of an image such as mean, image gradient, entropy are fetched and then the retrieved data is processed in the proposed framework. Deep learning algorithms are used for the validation of the proposed system. By evaluating it on real-time collected dataset, it resulted in an accuracy of 98.77%.

Advances in spam detection for email spam, web spam, social network spam, and review spam: ML-based and nature-inspired-based techniques

Despite the great advances in spam detection, spam remains a major problem that has affected the global economy enormously. Spam attacks are popularly perpetrated through different digital platforms with a large electronic audience, such as emails, microblogging websites (e.g. Twitter), social networks (e.g. Facebook), and review sites (e.g. Amazon). Different spam detection solutions have been proposed in the literature, however, Machine Learning (ML) based solutions are one of the most effective. Nevertheless, most ML algorithms have computational complexity problem, thus some studies introduced Nature Inspired (NI) algorithms to further improve the speed and generalization performance of ML algorithms. This study presents a survey of recent ML-based and NI-based spam detection techniques to empower the research community with information that is suitable for designing effective spam filtering systems for emails, social networks, microblogging, and review websites. The recent success and prevalence of deep learning show that it can be used to solve spam detection problems. Moreover, the availability of large-scale spam datasets makes deep learning and big data solutions (such as Mahout) very suitable for spam detection. Few studies explored deep learning algorithms and big data solutions for spam detection. Besides, most of the datasets used in the literature are either small or synthetically created. Therefore, future studies can consider exploring big data solutions, big datasets, and deep learning algorithms for building efficient spam detection techniques.

Blog Backlinks Malicious Domain Name Detection via Supervised Learning

Web spam is the unwanted request on websites, low-quality backlinks, emails, and reviews which is generated by an automated program. It is the big threat for website owners; because of it, they can lose their top keywords ranking from search engines, which will result in huge financial loss to the business. Over the years, researchers have tried to identify malicious domains based on specific features. However, lighthouse plugin, Ahrefs tool, and social media platforms features are ignored. In this paper, the authors are focused on detection of the spam domain name from a mixture of legit and spam domain name dataset. The dataset is taken from Google webmaster tools. Machine learning models are applied on individual, distributed, and hybrid features, which significantly improved the performance of existing malicious domain machine learning techniques. Better accuracy is achieved for support vector machine (SVM) classifier, as compared to Naïve Bayes, C4.5, AdaBoost, LogitBoost.

A fuzzy Dempster–Shafer classifier for detecting Web spams

The Web spam identification problem can be modeled as an instance of the conventional classification problem. Web spams aim at deceiving web crawlers by advertising certain Web pages through elevation of their page rankings superficially than their actual weights. Web spams are intended to produce fraudulent results of web search queries and degenerate the client’s experience by directing users to fake Web pages. We present a fuzzy evidence-based methodology for identifying Web spams by which the spamicity of web hosts is formulated as a reasoning problem in the presence of uncertainty. However, any classification task intrinsically suffers from incomplete or vague evidence and ambiguity in the class assignment based on evidence. In this work, we combine fuzzy reasoning as the decision maker for selecting the most suitable evidence in a multi-source Dempster–Shafer (DS) based classification algorithm. The introduced approach has the benefit of providing more reliable solution to detect spams without any prior information. The evidence theory offers flexible support that takes into account the multi-dimensional nature of implementation decisions. The experimental results show that the fuzzy reasoning in combination with DS theory, reduces the conflicts among evidence leading to enhanced classification results. The aim of this paper is to describe the potential of fuzzy reasoning and the Dempster–Shafer Theory (DST) as a decision model for the web spams classification problem.

A Fuzzy-Based Approach to Enhance Cyber Defence Security for Next-Generation IoT

In the modern era, the Cognitive Internet of Things (CIoT) in conjunction with IoT evolves which provides the intelligence power of sensing and computation for next-generation IoT (Nx-IoT) networks. The data scientists have discovered a large amount of techniques for knowledge discovery from processed data in CIoT. This task is accomplished successfully and data proceeds for further processing. The major cause for the failure of IoT devices is due to the attacks, in which Web spam is more prominent. There seems a requirement of a technique which can detect the Web spam before it enters into a device. Motivated from these issues, in this article, a cognitive spammer framework (CSF) for Web spam detection is proposed. CSF detects the Web spam by fuzzy rule-based classifiers along with machine learning classifiers. Each classifier produces the quality score of the webpage. These quality scores are then ensembled to generate a single score, which predicts the spamicity of the webpage. For ensembling, the fuzzy voting approach is used in CSF. The experiments were performed using a standard data set WEBSPAM-UK 2007 with respect to accuracy and overhead generated. From the results obtained, it has been demonstrated that CSF improves the accuracy by 97.3%, which is comparatively high in comparison to the other existing approaches in the literature.

Using deep belief network to demote web spam

Many score propagation based Web Spam Demotion Algorithms (WSDAs) have been proposed in last decade. There are two major challenges those algorithms suffer from. First, the non-incremental property of score propagation based WSDAs restricted their applications in real world since Web changes rapidly and running algorithm on the entire Web graph is computation consuming. Second, the score propagation based WSDAs adopt only link structure of the web graph to demote Web spam, so that they are vulnerable to some other kind of spamming techniques, such as content spam. In this paper, we propose a preference-based learning to rank method to address the above-mentioned issues confronted by score propagation based WSDAs. Our proposal consists of two components, a preference function and an ordering algorithm. The preference function is modeled by Deep Belief Network (DBN), which can benefit from unlabeled data for better generalization. The proposed Incremental Probabilistic Ordering Algorithm (IPOA) uses the trained preference function to calculate top-ranking probabilities of Web pages, and then uses those probabilities for final ranking. Therefore, the complex object (i.e. Web page) ranking problem is reduced to real number ranking problem, which can be solved efficiently by classical sorting algorithm. We conduct experiments to compare our proposal with conventional score propagation based WSDAs as well as some popular preference based learning to rank algorithms on two public available datasets, WEBSPAM-UK2006 and WEBSPAM-UK2007. Our experimental results demonstrate the superiority of our proposed method. Specifically, compared with score propagation based WSDAs, we obtain 0.0074 absolute improvement (0.7% relative improvement) on WEBSPAM-UK2006 and 0.065 absolute improvement (7.3% relative improvement) on WEBSPAM-UK2007 in terms of spam demotion score.

Detecting Web Spam Based on Novel Features from Web Page Source Code

Search engine is critical in people’s daily life because it determines the information quality people obtain through searching. Fierce competition for the ranking in search engines is not conducive to both users and search engines. Existing research mainly studies the content and links of websites. However, none of these techniques focused on semantic analysis of link and anchor text for detection. In this paper, we propose a web spam detection method by extracting novel feature sets from the homepage source code and choosing the random forest (RF) as the classifier. The novel feature sets are extracted from the homepage’s links, hypertext markup language (HTML) structure, and semantic similarity of content. We conduct experiments on the WEBSPAM-UK2007 and UK-2011 dataset using a five-fold cross-validation method. Besides, we design three sets of experiments to evaluate the performance of the proposed method. The proposed method with novel feature sets is compared with different indicators and has better performance than other methods with a precision of 0.929 and a recall of 0.930. Experiment results show that the proposed model could effectively detect web spam.

GT2FS-SMOTE: An Intelligent Oversampling Approach Based Upon General Type-2 Fuzzy Sets to Detect Web Spam

With the growing internet, web spam is also increasing, which majorly affect the user experiences with search engines. Web spam methods target the search engine’s internal programs to push targeted web sites at the upper positions. This paper proposed an intelligent oversampling approach based upon general type-2 fuzzy sets to balance the distribution and hence enhance the classification performance for web spam detection. The proposed method is validated with the real-world benchmark dataset, WEBSPAM-UK 2007, and its performance is assessed with AUC (Area under the ROC curve), F-measure, and G-mean. It is compared with SMOTE in combination with 11 well-known base classifiers available with WEKA Tool. The computational complexity of the proposed method is the same as that of SMOTE. It is reported that when the proposed method is combined with the base classifiers, it boosts up the classifier’s performance and outperforms SMOTE in every case. Proposed combinations are also statistically analyzed using Friedman, Holm, and Wilcoxon test to know the best combination among the 11 base classifiers. It is evident from the analysis that the proposed method, in combination with random forest (GT2FS-SMOTE+RF), performed best among every other combination.

Adaptive evaluation model of web spam based on link relation

AbstractThe credibility evaluation of web pages is an important part of current network information mining. Generally, content‐based evaluation and link relationship‐based evaluation are used to detect web spam. This paper proposes an algorithm based on link relationship which adjusts the weight of the link object according to the credibility of the link object to calculate the credibility of the web page. Considering the different linking conditions, (i) low‐scoring web pages actively linking to high‐scoring web pages, resulting in the reduction of high‐scoring web pages score; (ii) in order to endorse the low‐score web page, high‐score web pages actively linking to the low‐score web pages, adjust the score transmission. Obviously, the proposed algorithm is a recursive model. To avoid the amplification of the error caused by the loop and different evaluation results caused by different initial nodes, this paper presents a novel method to adaptively adjust the weights of the links according to the link conditions, then proves the convergence of the algorithm. Finally, we validate the proposed algorithm on the public datasets, which is compared with other up‐to‐date algorithms under different experimental indexes. The results show that the proposed algorithm can effectively reduce the ranks of spam pages in all web pages and improve the detection efficiency of spam pages.

Spams classification and their diffusibility prediction on Twitter through sentiment and topic models

ABSTRACT Alike web spam has been a major issue to almost every aspect of the current World Wide Web, social spam has led to a serious threat to the utilities of online social media, particularly in information diffusion. To combat this challenge, the significance and impact of social media and its content should be analyzed critically. To address this issue, Twitter is used as a case study, where we have modeled the contents of information through topic models and coupled it with the user-oriented features to predict spam diffusion with a good accuracy. Latent Dirichlet Allocation (LDA), a widely used topic modeling technique is applied to capture the latent topics from the tweets’ documents. The major contribution of this work is twofold. Firstly, analyzing the variation of sentiment and topics in spam/non-spam information, and secondly, constructing a feature set that classifies tweets more accurately into spam and non-spam category as compared to the existing approaches. The extensive simulation reveals the variation in sentiments and topics shared by the spam and non-spam tweets.

Multi-Scale Anomaly Detection on Attributed Networks

Many social and economic systems can be represented as attributed networks encoding the relations between entities who are themselves described by different node attributes. Finding anomalies in these systems is crucial for detecting abuses such as credit card frauds, web spams or network intrusions. Intuitively, anomalous nodes are defined as nodes whose attributes differ starkly from the attributes of a certain set of nodes of reference, called the context of the anomaly. While some methods have proposed to spot anomalies locally, globally or within a community context, the problem remain challenging due to the multi-scale composition of real networks and the heterogeneity of node metadata. Here, we propose a principled way to uncover outlier nodes simultaneously with the context with respect to which they are anomalous, at all relevant scales of the network. We characterize anomalous nodes in terms of the concentration retained for each node after smoothing specific signals localized on the vertices of the graph. Besides, we introduce a graph signal processing formulation of the Markov stability framework used in community detection, in order to find the context of anomalies. The performance of our method is assessed on synthetic and real-world attributed networks and shows superior results concerning state of the art algorithms. Finally, we show the scalability of our approach in large networks employing Chebychev polynomial approximations.

An efficient deep learning-based scheme for web spam detection in IoT environment

From the last few years, Internet of Things has revolutionized the entire world. In this, various smart objects perform the tasks of sensing and computing to provide uninterrupted services to the end users in different applications such as smart transportation, e-healthcare to name a few. With the inherent capabilities of these objects to take adaptive intelligent decisions, Cognitive Internet of Things is another paradigm of Internet of Things which emerges during this era. However, while accessing data from the Internet, web spam is one of the challenges to be handled. It has been observed from the literature review that for accessing data, search engines are preferred mostly by an individual. The search engine’s effective ranking can decrease the computational cost of objects during the data access. The current solutions to this issue are aimed to discover the spam in the search engine after its occurrence. So, in this proposal, we present a cognitive spammer framework that removes spam pages when search engines calculate the web page rank score. The framework detects web spam with the support of Long Short-Term Memory network by training the link features. This training resulted with an accuracy of 95.25, as more than 1,11,000 hosts are being correctly classified. However, the content features are trained by neural network. The proposed scheme has been validated with the WEBSPAM-UK 2007 dataset. Prior to processing, the dataset is pre-processed using a new technique called ‘Split by Over-sampling and Train by Under-fitting’. The ensemble and cross validation approach has been used for optimization of results with an accuracy of 96.96%. So, the proposed scheme outperforms the existing techniques.

CNN Based Malicious Website Detection by Invalidating Multiple Web Spams

Although a variety of techniques to detect malicious websites have been proposed, it becomes more and more difficult for those methods to provide a satisfying result nowadays. Many malicious websites can still escape detection with various Web spam techniques. In this paper, we first summarize three types of Web spam techniques used by malicious websites, such as redirection spam, hidden IFrame spam, and content hiding spam. We then present a new detection method that adopts the perspective of users and takes screenshots of malicious webpages to invalidate Web spams. The proposed detection method uses a Convolutional Neural Network, which is a class of deep neural networks, as a classification algorithm. In order to verify the effectiveness of the method, two different experiments have been conducted. First, the proposed method was tested based on a constructed complex dataset. We present comparison results between the proposed method and representative machine learning-based detection algorithms. Second, the proposed method was tested to detect malicious websites in a real-world Web environment for three months. These experimental results illustrate that the proposed method has a better performance and is applicable to a practical Web environment.

A Novel Set of Contextual Features for Web Spam Detection

Web spam is one of the significant problems facing search engines. It wastes sources and time, decreases the quality of results and leads to user discontent. The two main approaches to the detection spam web pages are link and content-based analysis. In this study, we mainly focus on content-based analysis in both user-visible text and the source code of a web page to propose a set of features for web spam detection. we explore the relationship between types and frequency of HTML (HyperText Markup Language) tags used in a web page source code. We also examine the structure of the URL as the other source of information. Finally, the content of a web page visible to the user is considered semantically in order to identify relevance among the number of the existing topics in the text as well as the coherence of a text using Latent Dirichlet Allocation. Experimental results show that the proposed features increases the index of balanced accuracy from 0.33 to 0.69 and improves the web spam detection rate.

AATMS: An Anti-Attack Trust Management Scheme in VANET

Vehicular Ad-hoc Network (VANET) is a significant component of intelligent transportation system, which facilitates vehicles to share sensitive information and corporate with others. However, due to its unique characteristics, such as openness, dynamic topology and high mobility, VANET suffers from various attacks. This paper proposes an anti-attack trust management scheme in VANET called AATMS to evaluate the trustworthiness of vehicles. With the help of AATMS, vehicles in VANET can avoid malicious vehicles and cooperate with trusted vehicles. The idea of AATMS is mainly inspired by TrustRank algorithm, which is used to combat web spams. In this paper, we calculate local trust and global trust, which indicate the local and global trust relationships among vehicles. First, Bayesian inference is adopted to calculate local trust of vehicles based on historical interactions. Then we select a small set of seed vehicles according to local trust and some social factors. Once we identify the reputable seed vehicles, we use the local trust link structure of vehicles to evaluate the global trust of all vehicles. The simulation results show that AATMS can efficiently identify trustworthy and untrustworthy vehicles in VANET even under malicious attacks.