Spam Identification Research Articles

PSC-BERT: A spam identification and classification algorithm via prompt learning and spell check

The rapid growth of the Internet has led to an increase in spam activities, posing significant security threats. Spammers use various tactics, such as intentional misspellings, to evade detection systems. Current spam detection mostly relies on semantic analysis, which is inadequate given these sophisticated evasion techniques. This paper addresses the problem by constructing a Chinese Camouflage Spam dataset (CCS) and proposing a novel detection and recognition model: Prompt and Spelling Checking-based BERT (PSC-BERT). BERT, a state-of-the-art language model, has been successful in many NLP tasks but faces limitations in handling misspelled text. PSC-BERT extends BERT by integrating semantic, phonetic, and glyph information from spam texts and employs a novel hard template prompt learning method that unifies text classification and spell-check tasks. In experiments, PSC-BERT outperformed baseline models, showing a 0.81% improvement over Semorph and a 2.01% improvement over LR in binary classification. For multi-class classification, it achieved a 0.95% increase in Macro-F1 and 0.49% in Weighed-F1 over the best baseline. Two sub-datasets from Chinese Camouflage Spam dataset were created for detection and recognition, and extensive analysis confirmed the model’s efficacy, especially in few-shot learning scenarios. Additionally, a case study illustrates the attention mechanism in BERT, providing deeper insights into token in the sequences. In summary, PSC-BERT demonstrates a significant advancement in spam detection capabilities against intentional misspellings problem.

Analyzing Privacy and Security in Cloud Computing Environments

With the rapid surge of spam across the Internet and its various forms, effectively identifying and combating spam has become an urgent priority. Cloud computing offers significant advantages in terms of storage and processing capabilities, making it a viable solution for analysing vast amounts of email data. To address the dynamic nature of spam and its life cycle, an anti-spam system with feedback reassessment is proposed. This system incorporates a text filtering approach based on active learning, involving four key stages: training, filtering, feedback, and re-filtering. Compared to traditional systems, the feedback-enabled filtering system demonstrates improved keyword filtering. To further enhance the accuracy of spam detection and minimize misjudgements in legitimate emails, leveraging weighted decision-making based on email header information is recommended. Additionally, for emails with sparse content, employing title weighting in the filtering process proves to be both feasible and effective, particularly in identifying spam with minimal text content. Given the advancements of cloud-based filtering methods over traditional algorithms, leveraging cloud computing holds promise in effectively combating the escalating volume of spam. As such, this paper delves into an in-depth exploration of spam identification within cloud computing environments, focusing on text filtering systems. This study is informed by a comprehensive analysis of existing anti-spam technologies, aiming to contribute to the ongoing efforts in mitigating spam proliferation online.

Images Spam Detection on Online Social Media using CNN with Pre-Trained Model

Abstract: Nowadays attackers move to image spam techniques instead of text based. The Majority of conventional techniques solely possess the capability to identify spam confined to textual content and hyperlinks. In this research we have done “Deep Convolutional Neural Networks” (DCNNs)in conjunction with pre-trained architectures. Image classification is one of the areas that has increased in the last decade very rapidly. But due to the less computational resources it become very challenging to train a good image classification model. With the help of Transfer learning, we can overcome this type of situation and can build a good image classification model. Our proposed model utilizes deep convolutional neural networks (DCNNs) along with pre-trained architectures. By fine-tuning this pre-trained model on the specific image dataset, the model can effectively learn to detect image- based spam content. Our suggested model surpasses contemporary state-of-the-art detection models concerning both accuracy and efficiency in the realm of the image spam identification.

Email Platform: Spam Detection

Abstract: In practically every industry today, from business to education, emails are used. Ham and spam are the two subcategories of emails. Email spam, often known as junk email or unwelcome email, is a kind of email that can be used to hurt any user by sapping their time and computing resources and stealing important data. Spam emailvolume is rising quickly day by day. Today’s email and IoT service providers face huge and massive challenges with spam identification and filtration. Email filteringis one of the most important and well-known methods among all the methods createdfor identifying and preventing spam. SVM, decision trees, and other machine learning and deep learning approaches have all been applied to this problem. Together with the explosive growth in internet users, email spam has increased substantially in recent years. Individuals are using them for illegal and dishonest purposes, such as fraud, phishing, and distributing malicious links through unsolicited email that can harm our systems and attempt to access your systems. By quickly constructing phone-y/fake profiles and email accounts, spammers prey on those who are ignorant of these scams. They use a real name in their spam emails. As a result, it’s critical to identify spam emails that include fraud. This project will accomplish this by utilizing machine learning methods, and this article will examine the machine learning algorithms, put them to use on our data sets, and select the approach that can detect emailspam with the maximum degree of precision and accuracy.

Spam Detection in Social Networks Using Machine Learning

Many social media platforms have emerged as a result of the online social network's (OSN) rapid expansion. They have become important in day-to-day life, and spammers have turned their attention to them. Spam detection is done in two different ways, such as machine learning (ML) and expert-based detection. The expert-based detection technique’s accuracy depends on expert knowledge, and the manual process is a time-consuming task. Thus, ML-based spam detection is preferred in OSN. Spam identification on social networks is a difficult operation involving a variety of factors, and spam and ham have resulted in an imbalanced data distribution, which gives an advantage to spammers for corrupting our devices. Spam detection based on ML algorithms like Logistic Regression (LR), K-Nearest Neighbour (KNN), Decision Trees (DT), Random Forest (RF), Support Vector Machine (SVM), and XGB, Voting Classifier (VC), and many other algorithms are used to design the address balance and to attain high assessment accuracy. There is a non-balance issue. Text is vectorized by vectorizers and all the relative results are stored. The experimental result shows that, as compared to KN, NB, ETC, RF, SVC, LR, XGB, and DT, the proposed VC provides a higher classification accuracy rate of 97.96%. The proposed methods are effective in identifying balanced and imbalanced datasets, as evidenced by the validation results. The website was created to detect messages as spam or not.

Tweet Spam Detection Using Machine Learning and Swarm Optimization Techniques

Social media networking platforms connect people living in every corner of the world. Twitter has now become a popular microblogging service, allowing users to express themselves and keep up with current events. Twitter has attracted spammers because to its popularity and ease of use. As a result, spam identification (ID) has become one of the most pressing issues. It is vital to detect and filter spam tweets as well as their owners in order to provide a spam-free environment. In this article, a spam detection method is proposed using a swarm optimization approach on a tweet-by-tweet basis. A spam tweet detection dataset is used to train the machine learning (ML) model. Metaheuristic features are created based on the input features in the dataset. Whale swam optimization algorithm (WOA) is used to select the important features before classification. The conventional objective function of WOA is modified into stochastic gradient descent (SGD) to perform feature selection. The selected subset of features is used to train the Adaboost (AB) classifier to detect the spam in the tweets. The AB classifier produced the best results in combination with WOA and SGD. The obtained accuracy is 99.85% in testing with a minimum subset of seven features and in the least possible minimum time of 17.9 s.

Machine Learning Driven Spam Detection for IoT Devices

Smart homes are gradually incorporating Internet of Things (IoT) devices, producing massive amounts of data. Wireless methods are used to convey the vast majority of this data. Several IoT devices may be impacted by a variety of IoT risks, including cyberattacks, inconsistent network connectivity, data leakage, etc. Machine learning (ML) techniques could be quite helpful in this case to guarantee security and authorization. Data anomalies must be discovered using statistics. This study investigates the dependability of IoT equipment that interacts. When determining a spam score, the six Machine Learning models are taken into account with increased input features including xgboost, BGLM, GLM, Bagging, and Stacking. The grade demonstrates the dependability of an IoT device based on many criteria. This grade demonstrates the dependability of an IoT device based on many criteria. When compared to other recent approaches, the findings demonstrate the effectiveness of the proposed strategy. The spamicity score of the home IoT device is calculated using the Spam Score algorithm. A publicly available dataset for smart homes meteorological data from the UCI repository and synthetic data from IoT devices are used to validate the technique. The results collected show just how effective the suggested algorithm is in analyzing time series data from the Internet of Things devices for spam identification.

Retracted: Spam Identification in Cloud Computing Based on Text Filtering System

Retracted: Spam Identification in Cloud Computing Based on Text Filtering System

Using Machine Learning to Protect Social Media Accounts from Spam Emails

Spam refers to all forms of bulk, unsolicited, unwelcome digital communication. Spam can be transmitted by text messages, phone calls, social media, or email, which is how it is frequently done. The rise of social media websites like Facebook, Twitter, YouTube, etc. has given spammers increased access to people. Most spam filtering solutions employ text-based algorithms. The fundamental difficulty is classification because it is a significant one. In this study, features are extracted from emails using classifications of rules. To automatically categorize emails, numerous systems have been created, a few of them include systemic decision-making. The use of Bayesian classifiers, networked neurons, support vector machines, and sample-based approaches. A sizable and time-consuming labelled dataset is required for the supervised model training phase. SVM and Naive Bayes, two supervised learning algorithms, outperform other models in terms of spam identification. It provides comprehensive explanations of these methods in addition to several likely areas for future research in spam filtering and detection. Because most spammers employ obfuscation techniques, the rules set need ongoing updating and improvement. Refinement is occasionally necessary, and in most situations, it is automated, so end users have less bother

A Comprehensive Review of Cyber Security Vulnerabilities, Threats, Attacks, and Solutions

Internet usage has grown exponentially, with individuals and companies performing multiple daily transactions in cyberspace rather than in the real world. The coronavirus (COVID-19) pandemic has accelerated this process. As a result of the widespread usage of the digital environment, traditional crimes have also shifted to the digital space. Emerging technologies such as cloud computing, the Internet of Things (IoT), social media, wireless communication, and cryptocurrencies are raising security concerns in cyberspace. Recently, cyber criminals have started to use cyber attacks as a service to automate attacks and leverage their impact. Attackers exploit vulnerabilities that exist in hardware, software, and communication layers. Various types of cyber attacks include distributed denial of service (DDoS), phishing, man-in-the-middle, password, remote, privilege escalation, and malware. Due to new-generation attacks and evasion techniques, traditional protection systems such as firewalls, intrusion detection systems, antivirus software, access control lists, etc., are no longer effective in detecting these sophisticated attacks. Therefore, there is an urgent need to find innovative and more feasible solutions to prevent cyber attacks. The paper first extensively explains the main reasons for cyber attacks. Then, it reviews the most recent attacks, attack patterns, and detection techniques. Thirdly, the article discusses contemporary technical and nontechnical solutions for recognizing attacks in advance. Using trending technologies such as machine learning, deep learning, cloud platforms, big data, and blockchain can be a promising solution for current and future cyber attacks. These technological solutions may assist in detecting malware, intrusion detection, spam identification, DNS attack classification, fraud detection, recognizing hidden channels, and distinguishing advanced persistent threats. However, some promising solutions, especially machine learning and deep learning, are not resistant to evasion techniques, which must be considered when proposing solutions against intelligent cyber attacks.

User Consumption Behavior Recognition Based on SMOTE and Improved AdaBoost

The sudden outbreak of COVID-19 has dealt a huge blow to traditional education and training companies. Institutions use the WeChat platform to attract users, but how to identify high-quality users has always been a difficult point for enterprises. In this paper, researchers proposed a classification algorithm based on SMOTE and the improved AdaBoost, which fuses feature information weights and sample weights to effectively solve the problems of overfitting and sample imbalance. To justify the study, it was compared with other traditional machine-learning algorithms. The accuracy and recall of the model increased by 19% and 36%, respectively, and the AUC value reached 0.98, indicating that the model could effectively identify the user's purchase intention. The proposed algorithm also ensures that it works well in spam identification and fraud detection. This research is of great significance for educational institutions to identify high-quality users of the WeChat platform and increase purchase conversion rate.

Anonymous Account Detection in Social Media Using Machine Learning and Natural Language Processing

Spammers have transformed significant person to person communication destinations into a stage for the spread of a tremendous measure of inadequate and maybe hazardous substance and data. Interpersonal interaction administrations are utilized by a great many people from one side of the planet to the other. The cooperations that people have with web-based media destinations, for example, Twitter and Facebook significantly affect their everyday lives, for certain terrible repercussions now and again, also. For instance, Facebook has developed to get quite possibly the most lavishly utilized foundation ever, empowering an unsuitably immense measure of spam to be sent out of the site. Client accounts made by counterfeit clients send spontaneous tweets to different clients to advance organizations or sites, which influence genuine clients as well as motivation asset utilization to ascend too. The chance of spreading off base data to clients by means of the utilization of phony personalities has additionally expanded, possibly prompting the appropriation of unsafe things. Thus, thediscovery of spammers and the ID of phony Twitter clients have as of late emerged as an unmistakable examination subject in the space of contemporary online interpersonal organizations (OSNs). All through this article,we will take a gander at the methods that are presently being used to distinguish spammers on the web-based media stage Twitter. Besides,a scientific categorization of Twitter spam location strategies is introduced, what Separates the strategies into four classifications dependent on their capacity to distinguish I counterfeit material, (ii) spamdependent on URL,(iii) spam in hot subjects,and (iv) fake clients on the person to person communication site. Just as a scope of models like client qualities, content attributes, diagram properties and different components, the provided procedures are additionally evaluated and thought about. There are three kinds of attributes: singular attributes, underlying qualities,and transient characteristics.Eventually, we accept that the exploration we've given will be an important asset for researchers looking for the features of ongoing progressions in Facebook spam identification in a one area.

ChSO-DNFNet: Spam detection in Twitter using feature fusion and optimized Deep Neuro Fuzzy Network

The usage of social networking platforms for interaction and meeting has grown significantly. However, as social networks continue to expand, Twitter has emerged as a vital social media for real life and has become a major source of spam. To address the aforementioned problems, spam identification on social media becomes an increasingly crucial task. The features that are present in the high dimension data of social networks cannot be used effectively by the existing approaches. In order to filter the spam information in social media, a Chimp Sailfish Optimization-based Deep Neuro Fuzzy Network (ChSO-based DNFN) is proposed. The proposed method effectively performs well under high dimensional data in real platform environment using deep learning classifier. It is more robust and generates optimal result and also reduces the computational complexity problems. Additionally, the proposed approach demonstrated improved performance in terms of metrics like precision, recall, and F-measure, which were measured using a 5k continuous dataset and yielded values of 0.894, 0.903, and 0.898, respectively.

Innovation and Business Model of Life and Health Sharing Platform Based on Algorithm of Naive Bayesian Model

Including shared bicycles, shared chefs, shared ideas, shared printing, and even shared human resources, the vocabulary of sharing has quietly entered people’s lives, affecting everyone’s necessities. “Everyone can participate” and “everything can be shared” are all familiar words in the sharing economy field. The implementation of the sharing economy is carried out by relying on the sharing platform. Knowledge, as a special resource, can exert its maximum benefit only when it is widely disseminated and shared, and the process of knowledge sharing and transformation itself contains knowledge innovation. By linking the multiparticipants of the platform, sharing, cocreation, and win-win are realized. The research value of this paper lies in the use of microservices to solve the problems of high coupling, poor scalability, and difficulty in rapid iteration in traditional monolithic applications, effectively identifying and blocking spam that may appear on the platform through a relatively simple solution. Meanwhile, it has been expected that under the trend of sharing economy, with the development of blockchain and 5G technology, an Internet life and health platform can be built. The upstream and downstream in the medical and health industry can be linked to share resources between the upstream and downstream in the industry, which can create a common nakedness, thus improving the operational efficiency and profitability of the entire medical and health industry. Finally, a spam identification scheme combining the improved Aho-Corasick algorithm (AC) and the naive Bayesian model (NBM) has been proposed, and a comparative experiment was conducted between this scheme and the scheme directly using NBM. The experimental results showed that the macro F1 value of the improved scheme on the bad evaluation dataset of platform A was 3.6% higher than that of NBM alone, and the macro F1 value of the improved scheme was 1.7% higher than that of NBM alone on the B platform review dataset. The overall performance of the improved NBM algorithm was stable and better than the traditional algorithm, which verified the feasibility of the scheme.

Dynamic Classification of Sentiments from Restaurant Reviews Using Novel Fuzzy-Encoded LSTM

User reviews on social media have sparked a surge in interest in the application of sentiment analysis to provide feedback to the government, public and commercial sectors. Sentiment analysis, spam identification, sarcasm detection and news classification are just few of the uses of text mining. For many firms, classifying reviews based on user feelings is a significant and collaborative effort. In recent years, machine learning models and handcrafted features have been used to study text classification, however they have failed to produce encouraging results for short text categorization. Deep neural network based Long Short-Term Memory (LSTM) and Fuzzy logic model with incremental learning is suggested in this paper. On the basis of F1-score, accuracy, precision and recall, suggested model was tested on a large dataset of hotel reviews. This study is a categorization analysis of hotel review feelings provided by hotel customers. When word embedding is paired with LSTM, findings show that the suggested model outperforms current best-practice methods, with an accuracy 81.04%, precision 77.81%, recall 80.63% and F1-score 75.44%. The efficiency of the proposed model on any sort of review categorization job is demonstrated by these encouraging findings.

Spam Identification in Cloud Computing Based on Text Filtering System

With the rapid increase of spam on the Internet and the diversification of its forms, how to quickly and effectively identify a large number of spam on the Internet has become an urgent topic. Cloud computing has obvious advantages in storage and processing, so it can effectively calculate a large amount of mail data. Due to the uncertainty and life cycle of spam, feedback re-judgment is added to the anti-spam system, and a text filtering system based on active learning with four stages of training, filtering, feedback, and re-filtering is implemented. Compared with the original system, the filtering system with feedback can improve the filtering of keywords. In order to effectively reduce the misjudgment rate of ordinary mail and improve the accuracy of spam judgment, it is suggested to improve the use of weighted decision-making of email header information to implement effective auxiliary classification. For emails lacking content, the filtering method of title weighting is feasible and effective, which can improve the identification of spam with relatively little text content. Because the filtering method on the cloud is far more advanced than the traditional algorithm, the development of the Internet can effectively solve the infinite increase of spam. Therefore, this paper makes an in-depth study on spam identification in cloud computing based on text filtering system by summarizing and analyzing the current anti-spam technologies.

Analysis of e-Mail Spam Detection Using a Novel Machine Learning-Based Hybrid Bagging Technique.

e-mail service providers and consumers find it challenging to distinguish between spam and nonspam e-mails. The purpose of spammers is to spread false information by sending annoying messages that catch the attention of the public. Various spam identification techniques have been suggested and evaluated in the past, but the results show that the more research in this regard is required to enhance accuracy and to reduce training time and error rate. Thus, this research proposes a novel machine learning-based hybrid bagging method for e-mail spam identification by combining two machine learning methods: random forest and J48 (decision tree). The proposed framework categorizes the e-mail into ham and spam. The database is split into multiple sets and provided as input to each method in this procedure. Moreover, tokenization, stemming, and stop word removal are performed in the preprocessing stage. Further, correlation feature selection (CFS) is employed in this research to select the required features from the preprocessed data. The effectiveness of the presented method is evaluated in terms of true-negative rates, accuracy, recall, precision, false-positive rate, f-measure, and false-negative rate; the outcomes of three studies are compared. According to the results, the presented hybrid bagged model-based SMD technology achieved 98 percent accuracy.

Chinese Spam Detection Using a Hybrid BiGRU-CNN Network with Joint Textual and Phonetic Embedding

The proliferation of spam in China has a negative impact on internet users’ experiences online. Existing methods for detecting spam are primarily based on machine learning. However, it has been discovered that these methods are susceptible to adversarial textual spam that has frequently been imperceptibly modified by spammers. Spammers continually modify their strategies to circumvent spam detection systems. Text with Chinese homophonic substitution may be easily understood by users according to its context. Currently, spammers widely use homophonic substitution to break down spam identification systems on the internet. To address these issues, we propose a Bidirectional Gated Recurrent Unit (BiGRU)–Text Convolutional Neural Network (TextCNN) hybrid model with joint embedding for detecting Chinese spam. Our model effectively uses phonetic information and combines the advantages of parameter sharing from TextCNN with long-term memory from BiGRU. The experimental results on real-world datasets show that our model resists homophone noise to some extent and outperforms mainstream deep learning models. We also demonstrate the generality of joint textual and phonetic embedding, which is applicable to other deep learning networks in Chinese spam detection tasks.

Classification of Image Spam Using Convolution Neural Network

Image identification and classification is a basic issue in the fields of mainframe visualization and pattern recognition. In today’s world, a great deal of unwanted material is distributed via the Internet. The unwanted information contained inside images, i.e., image spam, endangers email-based communication systems. Unlike textural spam, image spam is difficult to be detected by many machine learning (ML) techniques. This paper intends to investigate and evaluate four deep learning (DL) methods that may be useful for image spam identification. Firstly, neural networks, especially deep neural networks, were trained on various image features. Their resilience was measured on an enhanced dataset, which was created specifically to outwit existing image spam detection methods. Next, a convolution neural network (CNN) was designed, and verified through experiments. Experimental results show that our novel approach for image spam identification outshines other current techniques in the field.

Machine Learning-Based Secure Data Acquisition for Fake Accounts Detection in Future Mobile Communication Networks

Social media websites are becoming more prevalent on the Internet. Sites, such as Twitter, Facebook, and Instagram, spend significantly more of their time on users online. People in social media share thoughts, views, and facts and create new acquaintances. Social media sites supply users with a great deal of useful information. This enormous quantity of social media information invites hackers to abuse data. These hackers establish fraudulent profiles for actual people and distribute useless material. The material on spam might include commercials and harmful URLs that disrupt natural users. This spam content is a massive problem in social networks. Spam identification is a vital procedure on social media networking platforms. In this paper, we have proposed a spam detection artificial intelligence technique for Twitter social networks. In this approach, we employed a vector support machine, a neural artificial network, and a random forest technique to build a model. The results indicate that, compared with RF and ANN algorithms, the suggested support vector machine algorithm has the greatest precision, recall, and F-measure. The findings of this paper would be useful in monitoring and tracking social media shared photos for the identification of inappropriate content and forged images and to safeguard social media from digital threats and attacks.