Spam Detection

We propose a novel bio-inspired solution for biomedical article classification. Our method draws from an existing model of T-cell cross-regulation in the vertebrate immune system (IS), which is a complex adaptive system of millions of cells interacting to distinguish between harmless and harmful intruders. Analogously, automatic biomedical article classification assumes that the interaction and co-occurrence of thousands of words in text can be used to identify conceptually-related classes of articles-at a minimum, two classes with relevant and irrelevant articles for a given concept (e.g. articles with protein-protein interaction information). Our agent-based method for document classification expands the existing analytical model of Carneiro et al. , by allowing us to deal simultaneously with many distinct T-cell features (epitomes) and their collective dynamics using agent based modeling. We already extended this model to develop a bio-inspired spam-detection system . Here we develop our agent-base model further, and test it on a dataset of publicly available full-text biomedical articles provided by the BioCreative challenge . We study several new parameter configurations leading to encouraging results comparable to state-of-the-art classifiers. These results help us understand both T-cell cross-regulation and its applicability to document classification in general. Therefore, we show that our bio-inspired algorithm is a promising novel method for biomedical article classification and for binary document classification in general.

This paper presents the use of corpus linguistics techniques on supposedly "clean" corpora and identifies potential pitfalls. Our work relates to the task of filtering sensitive content, in which data security is strategically important for the protection of government and military information, and of growing importance in combating identity fraud. A naïve keyword filtering approach produces a large proportion of false positives, and the need for more fine-grained approaches, suggests the consideration of using corpus linguistics for such content filtering. We present work undertaken on the Enron corpus, a collection of emails that has had various tasks undertaken on various portions and versions of it by other researchers. We have made some efforts to reconcile differences between the different versions by considering what impact some of these versions have on our results. Our anticipated efforts in using automatic ontology learning Gillam and Ahmad 2005], and local grammars to discover points of interest within the Enron corpus have been reoriented to the problem of discovering "confidentiality banners" with a view to removing them from the collection. Our work to date makes strong use of collocation patterns [Smada 1993] to identify the signatures of these banners, and to use the banners themselves to better understand the different versions of the Enron corpus. We further consider the use of extended collocation patterns to identify text "zones", following , and the subsequent potential for sentiment analysis ; 1

In this paper we describe an approach to information assurance in which we can prevent breach of confidentiality. Specifically, we examine aspects of the propagation of confidential information via email. Email provides one simple mechanism by which naive users can rapidly breach confidence, and effectively demonstrates the difficulty in enforcing "best practice". We propose and demonstrate an intelligent filtering system aimed at prevention of disclosure and enforcement of "best practice". We show how work in corpus linguistics and data mining can contribute significantly to information assurance and report on experiments on a relatively large email collection that demonstrates the value of such work. Our experiments further demonstrate the potential for reducing false positives. We give due consideration to the danger of missed messages that should have been prevented from propagation.

E-commerce sites, forums, and blogs have become popular platforms for people to share their views. Reviews have emerged as a crucial source of information for potential customers, influencing their purchasing decisions. Similarly for profit gain or fame, Spam reviews are deliberately written with the intention of defaming businesses or individuals. This act is known as review spamming. Spam review detection is rapidly answered by various ML techniques. Review of spamming is more challenging task in multilingual communities. Spammer behavior features and linguistic features often exhibit complex relationships that influence the nature of spam reviews. The unified representation of features is another challenging task in spam detection. Various deep learning approaches have been proposed for review spamming, including different neural networks (Convolutional Neural Network, CNN). These methods are specialized in extracting the features but lack to capture feature dependencies effectively with other features. Spam Review Detection using the Fusion Gradient Boosting Model (SRD-FGBM) is proposed with fusion of spammer behavior features and linguistic features to automatically detect and classify the spam reviews. Fusion enables the proposed model to automatically learn the interactions between the features during the training process, allowing it to capture complex relationships and make predictions based on both types of features. It apparently shows the promising result by obtaining 94.3% accuracy.

In today’s world, fake review identification and prediction is an importantarea of sentiment analysis of the E-commerce industry. The automatic fake review categorizersidentify and categorize a variety of duplicate, spam, fake and untrustworthy reviews using machine learning techniques. This paper studies various recent existing fake review detection methods using NB and RF classifiers for the Yelp and Flipkart datasets. It provides a detailed study on various fake review predictors and compares their basic and performance-based specifications. It highlights the challenges, threats, and gaps of these existing works. Further, it graphically shows the discrimination for the specifications of year-wise evolution, classifier usage, and dataset usage.

Online Social Networks (OSNs), such as Twitter, offer attractive means of social interactions and communications, but also raise privacy and security issues. The OSNs provide valuable information to marketing and competitiveness based on users posts and opinions stored inside a huge volume of data from several themes, topics, and subjects. In order to mining the topics discussed on an OSN we present a novel application of Louvain method for TopicModeling based on communities detection in graphs by modularity. The proposed approach succeeded in finding topics in five different datasets composed of textual content from Twitter and Youtube. Another important contribution achieved was about the presence of texts posted by spammers. In this case, a particular behavior observed by graph community architecture (density and degree) allows the indication of a topic strength and the classification of it as natural or artificial. The later created by the spammers on OSNs.

To the modern Search Engines (SEs), one of the biggest threats to be considered is spamdexing. Nowadays spammers are using a wide range of techniques for content generation, they are using content spam to fill the Search Engine Result Pages (SERPs) with low-quality web pages. Generally, spam web pages are insufficient, irrelevant and improper results for users. Many researchers from academia and industry are working on spamdexing to identify the spam web pages. However, so far not even a single universally efficient method is developed for identification of all spam web pages. We believe that for tackling the content spam there must be improved methods. This article is an attempt in that direction, where a framework has been proposed for spam web pages identification. The framework uses Stop words, Keywords Density, Spam Keywords Database, Part of Speech (POS) ratio, and Copied Content algorithms. For conducting the experiments and obtaining threshold values WEBSPAM-UK2006 and WEBSPAM-UK2007 datasets have been used. An excellent and promising F-measure of 77.38% illustrates the effectiveness and applicability of proposed method.

The paper reviews various approaches for detecting fake reviews using different machine learning techniques, each with distinct strengths and limitations. It examines existing literature on supervised learning methods, unsupervised techniques, graph-based models, and hybrid approaches. Among these, unsupervised models rely on pattern recognition, while supervised methods, including SVM and transformer-based models like BERT, offer high accuracy but struggle with class imbalance and computational efficiency. Unsupervised and graph-based models serve as effective alternatives when labeled data is scarce or when complex relationships between reviews and users must be analyzed. Additionally, hybrid approaches that integrate multiple techniques are gaining traction, as they enhance feature selection and model performance. In this paper, we explore different methodologies for fake review classification, analyze their advantages and drawbacks, and highlight key challenges in the field.

The exponential growth of mobile communication has intensified the threat of SMS spam, compromising user security and trust in messaging platforms. This study addresses this challenge by designing and deploying a robust spam detection system using machine learning. We analyze a publicly available SMS dataset through rigorous pre-processing, including text normalization, tokenization, and feature engineering, followed by TF-IDF vectorization. A comparative evaluation of 11 classifiers-spanning probabilistic models, ensemble methods, and linear classifiers-reveals that ensemble techniques outperform traditional algorithms. The Extra Trees Classifier and XGBoost achieve state-of-the-art results, with 97.9% accuracy and 97.5% precision, demonstrating their efficacy in distinguishing spam from legitimate messages. To bridge the gap between research and practical application, we develop an interactive Streamlit web application that enables real-time spam classification with a user-friendly interface. This work underscores the potential of ensemble learning for text classification tasks and provides a scalable framework for combating SMS spam in real-world scenarios.

In this paper, the clustering Algorithm known as Support Vector Classifier (SVC) is used. SVC offers classifiers such as logistic regression and decision trees that provide very high accuracy compared to others. The model first preprocesses the input after that extracts the required details then it classifies the details and finally detection. SVC presents a workload that is data intensive and compute. To reduce the workload the model trained offline, achieving 97.95% top-1 accuracy. And it also segmentation and preprocessing is also used to minimize input image size fed into SVM.

Users of all ages face risks on social media and messaging platforms. When encountering suspicious messages, legitimate concerns arise about a sender's malicious intent. This study examines recent advances in Natural Language Processing for detecting messagebased threats in digital communication. We conducted a systematic review following PRISMA guidelines, to address four research questions. After applying a rigorous search and screening pipeline, 30 publications were selected for analysis. Our work assessed the NLP techniques and evaluation methods employed in recent threat detection research, revealing that large language models appear in only 20% of the reviewed works. We further categorized detection input scopes and discussed ethical and privacy implications. The results show that AI ethical aspects are not systematically addressed in the reviewed scientific literature.

Online credibility is a quality pursued by users, business and brands on Internet. Having a verified page on Facebook means improvement of the social web presence, reliability and reinforcement of security against impersonations of identity, unwanted fake pages and spams. Since the Facebook's page verification request has become more complex and the requirements to receive a verified page badge are uncertain, this paper describes the use of foundations of transparency on information systems to fulfill the data on the forms of the application for verification to improve the success in receiving the verified page status.

The article explores fundamental techniques for converting text into numerical data for machine learning algorithms. It meticulously examines various methods, including word vector representation via neural networks like Word2Vec, and explains the principles behind linear models such as logistic regression and support vector machines. Convolutional neural networks (CNN) and long short-term memory (LSTM) methods are also discussed, covering their components, mechanisms, and training processes. The research extends to developing and testing software for spam detection, hate speech identification, and recognizing offensive language. Using two datasets-one for labeled text messages and another for Twitter posts-the study analyzes data to address challenges like imbalanced data. A comparative analysis among linear models, deep neural networks, and single-layer models, using pre-trained bidirectional encoder representations from transformers (BERT) network, reveals promising results. The convolutional neural network stands out with a remarkable accuracy of 0.95. The study also adapts neural network architectures for hate speech and offensive language classification.

Communicating by email has become crucial for all companies today. However, a significant amount of undesirable messages pass through mail servers. Such unwanted and unsolicited communications are responsible for a significant loss of time and money for all corporations and may be damaging for corporate reputations. We present here the email delivery solution we have implemented in our institute, to prevent users from receiving spam and viruses. Our choice was motivated by many criteria including reliability, availability and security but also flexibility and price. Indeed, this email gateway runs on a Linux server and is based on a few modular open source tools designed for security purposes. We present all the tools involved in this gateway and discuss their advantages and drawbacks. Technical explanations of system optimization and Bayesian filtering are also demonstrated in this paper. Finally, we show the rate of success we obtained on deploying this email gateway. 1 The ms-tnef is usually generated by Microsoft's mail client either Microsoft Outlook or Exchange Mail. The format encodes the attributes of messages such as fonts, colour of fonts or type face of fonts.

this research project acts as a foundational guide for students and early-career professionals interested in AI. It combines academic theory, technical practice, and ethical reflection to provide a holistic understanding of AI and machine learning.

Abstract: There are economic and socio-cultural differences that characterise the north and south of Italy. A stereotype is that university students from rural southern Italy are more disadvantaged and isolated than those from the urban north. Past research has hypothesised that differences in socio-economic status impact on student learning, which is a factor of thinking style. This study set out to explore if university students from a northern and a southern Italian university report markedly different thinking style preferences. ...

In the modern era, mobile phones have become ubiquitous, and Short Message Service (SMS) has grown to become a multi-million-dollar service due to the widespread adoption of mobile devices and the millions of people who use SMS daily. However, SMS spam has also become a pervasive problem that endangers users' privacy and security through phishing and fraud. Despite numerous spam filtering techniques, there is still a need for a more effective solution to address this problem [1]. This research addresses the pervasive issue of SMS spam, which poses threats to users' privacy and security. Despite existing spam filtering techniques, the high false-positive rate persists as a challenge. The study introduces a novel approach utilizing Natural Language Processing (NLP) and machine learning models, particularly BERT (Bidirectional Encoder Representations from Transformers), for SMS spam detection and classification. Data preprocessing techniques, such as stop word removal and tokenization, are applied, along with feature extraction using BERT. Machine learning models, including SVM, Logistic Regression, Naive Bayes, Gradient Boosting, and Random Forest, are integrated with BERT for differentiating spam from ham messages. Evaluation results

The rapid proliferation of mobile communication has kept Short Message Service (SMS) a popular channel for personal and business communications. As the usage rose, the malicious exploitation followed, leading to the increased flow of unwanted and harmful content in form of spam. Spam sent by SMS is a channel for identity theft, phishing attacks, and fraud, etc., so it is important to ensure the detection of SMS spam using improved methodologies. This research paper presents a machine learning based intelligent spam filtering framework. The SMS Spam Collection Dataset consists of 5,574 entries, each labeled as either spam message or ham message. The preprocessing included normalization, tokenization, lemmatization, and removal of tokens that were not informative. Textual data was transformed into numerical vectors using Term Frequency-Inverse Document Frequency (TF-IDF) and Bag-of-Words (BoW) approaches. Some algorithms were compared against each other, including Multinomial Naïve Bayes, Support Vector Machine, Logistic Regression, Random Forest, and K-Nearest Neighbors. In particular, the probabilistic classifier was the best performer. Ensemble strategies i.e Voting Classifier, XGBoost were used to improve the predictive stability. Statistical and structural message attributes were used for feature extraction and a dimensionality reduction method was applied to allow for better generalization. The results were much better, 98.6% accuracy, the evaluation was performed with 10-fold cross-validation. The research paper also examines possibilities for deployment on multilingual datasets and suggests enhancing the model to be robust against concept drift, adversarial inputs, and evolving linguistic patterns.

With the growth of networking the usage of mails are also enhanced. Due to rapid growth of internet, dependency of communication is mostly based on electronics mails for both commercial and business purposes. According to today's scenarios electronics mails are also plays vital role in marketing or production advertisement. Hence numerous marketing firms used emails as a tool for promoting their products and services. These types of mails are generally called spam mails. Sometimes it is quite difficult to identify important mails among such group of spam mails. Because of this identification of such types of mails are essential so that they can keep away from important mails. In this paper we are presenting various techniques that are used spam filtering.

With the growth of networking the usage of mails are also enhanced. Due to rapid growth of internet, dependency of communication is mostly based on electronics mails for both commercial and business purposes. According to today's scenarios electronics mails are also plays vital role in marketing or production advertisement. Hence numerous marketing firms used e- mails as a tool for promoting their products and services. These types of mails are generally called spam mails. Sometimes it is quite difficult to identify important mails among such group of spam mails. Because of this identification of such types of mails are essential so that they can keep away from important mails. In this paper we are presenting various techniques that are used spam filtering.

The widespread use of email as a primary communication medium has led to an increase in spam messages, which pose significant threats to privacy, productivity, and cybersecurity. Spam emails, often disguised as legitimate messages, can carry malicious links, phishing scams, and fraudulent content. This paper presents a machine learning-based approach for identifying spam emails with high accuracy. By employing natural language processing (NLP) techniques and the Naïve Bayes classifier, we preprocess a labeled dataset of email messages, extract relevant features, and train a classification model. The model's effectiveness is evaluated using performance metrics such as accuracy, precision, recall, and F1-score. The results demonstrate the reliability and practicality of machine learning in mitigating email spam, offering a scalable and adaptive solution to an ongoing digital challenge

Phishing remains one of the most prevalent and evolving cybersecurity threats, exploiting humanvulnerabilities through deceptive digital communication. This study proposes a dynamic, Windows-specific phishing detection model leveraging Random Forest machine learning techniques. By integrating Term Frequency-Inverse Document Frequency (TF-IDF) vectorization with structured email features, the model classifies phishing and legitimate emails with high accuracy. Using secondary data and publicly available datasets, the model achieved a classification accuracy of 98.31% and demonstrated balanced performance across precision, recall, and F1-score metrics. This research underscores the effectiveness of hybrid feature strategies and ensemble learning for phishing detection while outlining key limitations and future directions, including model generalization and real-world deployment readiness.

The number of email users is increasing every day worldwide. If you have to communicate officially with someone, whether in a business matter or with someone else, your electronic mail is the best option. When identifying the emails, 52-60% were spam. Spammers use such creative methods that it is difficult to determine whether the email is spam. This study focused on classifying spam emails for detection with high accuracy using a machine-learning algorithm. This research involved using multiple datasets and applying various algorithms, such as Random Forest, Logistic Regression, Naïve Bayes, and SVM, to determine how they perform. Additionally, other approaches, such as Feature Engineering, Bagging, and Boosting, have been utilized. The implementation of feature engineering techniques is aimed at enhancing the performance and accuracy of the model. The effectiveness of the framework focuses on metrics such as accuracy, precision, recall, and F1-score. These findings are expected to contribute to improved spam detection mechanisms. This research not only addresses current limitations in spam detection but also sets the stage for future advancements in the field. The findings will contribute to reducing the risk of spam filters and malicious emails and enhance cyber security.

This study suggests a machine learning-based detection system built with Python and Django to tackle the growing problem of fraudulent profiles on social networking sites. Malicious actors are progressively setting up phony identities for spamming, phishing, and disseminating false information as social media usage keeps growing. In order to accurately identify bogus profiles, the suggested system analyzes user attributes, behavioral patterns, and network properties using a variety of supervised learning algorithms, such as Random Forests, Support Vector Machines, and Decision Trees. Our approach makes advantage of Django's powerful web framework to produce an intuitive, scalable profile monitoring and analysis interface. According to experimental data, the overall detection accuracy is 92%, with 90% precision and 88% recall rates. The system greatly outperforms traditional rule-based approaches in both detection accuracy and processing efficiency, particularly when handling large datasets. The Django implementation provides real-time monitoring capabilities, reducing manual verification efforts while maintaining high detection reliability. This research contributes to enhancing online security by providing an effective tool for identifying and mitigating fake profile threats on social networking platforms.

Industrial insider risk assessment using electroencephalogram (EEG) signals has consistently attracted a lot of research attention. However, EEG signal-based risk assessment systems, which could evaluate the emotional states of humans, have shown several vulnerabilities to data poison attacks. In this paper, from the attackers' perspective, data poison attacks involving label-flipping occurring in the training stages of different machine learning models intrude on the EEG signalbased risk assessment systems using these machine learning models. This paper aims to propose two categories of label-flipping methods to attack different machine learning classifiers including Adaptive Boosting (AdaBoost), Multilayer Perceptron (MLP), Random Forest, and K-Nearest Neighbors (KNN) dedicated to the classification of 4 different human emotions using EEG signals. This aims to degrade the performance of the aforementioned machine learning models concerning the classification task. The experimental results show that the proposed data poison attacks are model-agnostically effective whereas different models have different resilience to the data poison attacks.

The spread of misinformation on the internet and social media has become a growing problem, impacting public opinion and decisionmaking. In this study, we explore how machine learning can be used to detect and classify fake news more effectively. We gathered a diverse dataset from various online sources and applied a series of preprocessing steps to prepare the data for analysis. This included techniques like tokenization, normalization, removing punctuation and stop words, and lemmatization. To improve the quality of the analysis, we developed an Enhanced Feature Engineering framework for Fake News Detection. This framework combined key features such as TF-IDF, Bag of Words, tweet length, and sentiment analysis, creating a robust dataset for training machine learning models. Among the various models we tested, the ensemble voting classifier stood out for its accuracy and reliability in distinguishing real news from fake. Its strong performance demonstrates the potential of combining multiple algorithms to tackle complex problems like misinformation. Through this research, we aim to contribute to the ongoing efforts to combat fake news, helping to create a more reliable and trustworthy digital space.

The advanced architecture of Large Language Models (LLMs) has revolutionised natural language processing, enabling the creation of text that convincingly mimics legitimate human communication, including phishing emails. As AI-generated phishing emails become increasingly sophisticated, a critical question arises: How effectively can current email systems and detection mechanisms identify these threats? This study addresses this issue by analysing 63 AI-generated phishing emails created using GPT-4o. It evaluates the effectiveness of major email services, Gmail, Outlook, and Yahoo, in filtering these malicious communications. The findings reveal that Gmail and Outlook allowed more AI-generated phishing emails to bypass their filters compared to Yahoo, highlighting vulnerabilities in existing email filtering systems. To mitigate these challenges, we applied 60 stylometric features across four machine learning models: Logistic Regression, Support Vector Machine, Random Forest, and XGBoost. Among these, XGBoost demonstrated superior performance, achieving 96% accuracy and an AUC score of 99%. Key features such as imperative verb count, clause density, and firstperson pronoun usage were instrumental to the model's success. The dataset of AI-generated phishing emails is publicly available on Kaggle to foster further research.

Nuestro objetivo con este artículo es relacionar las emociones y los estereotipos de género por medio del análisis de las narrativas que contienen los spams.  Presentaremos cuatro diferentes tipos de spam de 450 correos electrónicos, que tengan entre 4 y 52 líneas de extensión. Estos se recibieron entre finales de 2009 y mediados de 2016 y fueron escritos en inglés, español, portugués y francés. Nuestro análisis se realizará en 2 pasos. En el primer paso se describirán las estrategias lingüísticas discursivas. En el segundo paso, se abordará la expresión lingüística de las emociones. Nos basaremos en la Teoría del Posicionamiento de Rom Harré para identificar dónde se ubican los autores y los lectores en las narraciones.

In this article, we aim at relating emotions and gender stereotypes through the analysis of the narratives contained in spam e-mails. We will present four different types of spam e-mails from a corpus consisting of 450 emails, their length being between four and fifty-two lines of extension. They were received between late 2009 and mid-2013, and written in English, Spanish, Portuguese and French. Our analysis will be carried out in two steps. In the first step, the discursive-linguistic strategies will be described. In the second step, the linguistic expression of emotions will be addressed. We will rely on the Positioning Theory of Rom Harré to identify where authors and readers are positioned in the narratives.

Phishing, a social Engineering attacks are becoming more common because of the quick digitization of healthcare services. Phishing is often used to steal user data, including login credentials and credit card numbers. It occurs when an attacker, masquerading as a trusted entity, dupes a victim into opening an email, instant message, or text message and this, occurring at the healthcare services clearly puts patient data security and system integrity at danger. This study introduces a brand-new machine learning-based method for identifying phishing attempts directed at healthcare institutions. Using realworld phishing and authentic emails from a variety of healthcarerelated domains, we created an extensive dataset. By employing sophisticated feature extraction methods, such as URL analysis, email header examination, and linguistic features, we were able to identify important markers of phishing activity. Emails were categorized as either authentic or phishing using a variety of machine learning techniques, such as Random Forest, Support Vector Machines, and Neural Networks. Our tests showed that the Random Forest algorithm was the most effective, achieving the highest accuracy of 95percent with an F1 score of 0.93.

Estimating the number of triangles in a graph is one of the most fundamental problems in sublinear algorithms. In this work, we provide the first approximate triangle counting algorithm using only polylogarithmic queries. Our query oracle "Tripartite Independent Set" (TIS) takes three disjoint sets of vertices $A$, $B$ and $C$ as input, and answers whether there exists a triangle having one endpoint in each of these three sets. Our query model is inspired by the "Bipartite Independent Set" (BIS) query oracle of Beame et al. (ITCS, 2018). Their algorithm for edge estimation requires only polylogarithmic BIS queries, where a BIS query takes two disjoint sets $A$ and $B$ as input and answers whether there is an edge with endpoints in $A$ and $B$. We extend the algorithmic framework of Beame et al., with TIS replacing BIS, for triangle counting using ideas from color coding due to Alon et al. (J. ACM, 1995) and a concentration inequality for sums of random variables ...

Phishing, a prevalent social engineering attack, is frequently employed by cybercriminals to steal sensitive user data, including login credentials, credit card information, and confidential government or corporate details. Despite various methods deployed to detect phishing emails, attackers continue to evolve sophisticated techniques to bypass existing defences. Machine learning has emerged as a reliable solution due to its ability to learn and identify patterns. This reliability stems from the inherent capability of machine learning algorithms to learn complex patterns within data. In this study, Decision Tree (DT), Random Forest (RF), and Extreme Gradient Boosting (XGBoost) algorithms were employed to construct a model designed to detect email phishing attacks. The aim is to develop a machine learning-based model for detecting phishing emails. The objectives of this study were achieved by firstly implementing and evaluating three tree-bassed machine learning algorithms namely DT, RF, and XGBoost, secondly, the models were evaluated based on the accuracy, precision, recall, and F1-score of these algorithms using the Kaggle phishing email dataset, and to identify the most effective algorithm for phishing detection based on the evaluation metrics. The model was tested on Mendeley Kaggle Email Phishing Datasets. The DT classifier achieved an accuracy of 91% with a precision of 0.86. In comparison, the RF model performed better, with an accuracy of 96% and precision values of 0.94. while the XGBoost matched the performance of RF, also attaining 96% accuracy, with precision scores of 0.94. Both RF and XGBoost demonstrated superior performance over the DT in terms of accuracy and precision.

With the continuous rise in the number of mobile device users, SMS (Short Message Service) remains a prevalent communication tool accessible on both smartphones and basic phones. Consequently, SMS traffic has experienced a significant surge. This increase has also led to a rise in spam messages, as spammers seek financial or business gains through activities like marketing promotions, lottery scams, and credit card information theft. Consequently, spam classification has become a focal point of research. In this paper, we explore the effectiveness of 11 machine learning algorithms for SMS spam detection, including multinomial Naïve Bayes, K-Nearest Neighbors (KNN), and Random Forest, among others. Utilizing datasets from UCI and Bangla SMS collections, our experimental results reveal that the multinomial Naïve Bayes algorithm surpasses previous models in spam detection, achieving accuracies of 98.65% and 89.10% in the respective datasets.

Combatting email spam has remained a very daunting task. Despite the over 99% accuracy in most non-image-based spam email detection, studies on image-based spam hardly attain such a high level of accuracy as new email spamming techniques that defeat existing spam filters emerges from time to time. The number of email spams sent out daily has remained a key factor in the continued use of spam. In this paper, a simple convolutional neural network model, 123DNet was developed and trained with 28,929 images drawn from 2 public datasets and a Personally Generated dataset. The model was optimized to the least set of layers to have 1 input layer, 2 embedded Convolutional layers as a hidden layer, and 3 neural network layers. The model was tested with a total of 4,339 images of the three dataset samples and then with a separate set of 1,200 images to test performance on never-seen-before images. A Classification Performance analysis was carried out using the confusion matrix. Performance metrics including Accuracy, Precision, True Negative Accuracy, Sensitivity, Specificity, and F1 Measure were computed to ascertain the model's performance. The Model returned an F1 Score of 97% on a public dataset's test sample and 88% on Never-seen-before test samples outperforming some pre-existing models while performing significantly well on the newly generated image test samples. It is recommended that a model that performed so well with new never-seen-before spam images be integrated into spam filtering systems.

Information retrieval on the web is significant and furthermore complex activity for web mining. Because of enormously increased the number of websites on the Internet, the execution of PageRank Algorithm should be easy and faster in operation. So the system proposes the fast execution time in PageRank with the reduction of iteration step rather than the original PageRank Algorithm. The proposed system contains two main parts. The first part is to create simple website using mostly HTML, CSS and also bootstrap, front-end framework for web design in the demonstration of the proposed system. The second part is to make the line number extractor using simple web crawler to get the required information for proposed PageRank Algorithm. Then, the PageRank algorithm is executed by using the data from web crawler. Finally, the convergence value is taken to specify the iteration count during the operation of PageRank Algorithm. The main point of proposed system is faster execution time than t...

Abstract. Through their interaction with search engines, users provide implicit feedback that can be used to extract useful knowledge and improve the quality of the search process. This feedback is encoded in the form of a query log that consists of a sequence of search actions, which contain information about submitted queries, documents viewed, and documents clicked by the users. In this paper, we propose characterizing documents and queries via the information available within a query log, with the goal of detecting either query polysemy ...

Online Social Networks (OSNs) at present engages the majority of individuals, from the newborn to the elderly. Social networks provide direct connection between people and increased information sharing opportunities to provide a large audience, etc. Social networking sites, in particular Facebook and Twitter, have seen a rise in hackers and privacy violation due to its popularity among people. This has been inspired researchers to develop a new secure system that can detect these fake accounts on these Online Social Networks (OSNs). Several different experiments have been performed to detect the fake social networking accounts called fake profiles, relying on the characteristics of the user-base, graph type, content and time of accounts. So, preventing the existence of malicious accounts is significant. Machine learning-based methods have been used to identify fraudulent Facebook and Twitter users. This data will be processed by machine learning algorithms and determined fake profile accounts. This paper proposes a spammer detection model and presents a review of identifying fake profiles in social networks using machine learning algorithms.

Grouping resources into set of classes allows easy access to the resources we use in our day-to-day lives. This classification makes the search faster and easier. The process of classifying the resources manually becomes expensive. This cost effectiveness switch to automated classification of resources which depends on the content of the data. This paper deals with the semantic analysis of tags. Collaborative tagging system allows any user to annotate the web resources. The user annotations can be useful to discover the aboutness of resources and also helps to ascertain classification. The social tagging system focuses on using support vector machine (SVM) as a state-of-the-art classification algorithm. We have two large scale social tagging data sets to interpret the characteristics of social classification. This System supports user defined annotations which immensely renovate folksonomies.

Because of its ease of use and speed compared to other communication applications, email is the most commonly used communication application worldwide. However, a major drawback is its inability to detect whether mail content is either spam or ham. There is currently an increasing number of cases of stealing personal information or phishing activities via email. This project will discuss how machine learning can help in spam detection. Machine learning is an artificial intelligence application that provides the ability to automatically learn and improve data without being explicitly programmed. A binary classifier will be used to classify the text into two different categories: spam and ham. This research shows the machine learning algorithm in the Azure-based platform predicts the score more accurately compared to the machine learning algorithm in visual studio, hybrid analysis and JoeSandbox cloud.

Malicious websites are those that are created to harm visitors or exploit their information for illegal purposes. These websites are commonly utilized in attacks, such as phishing, malware distribution, and scams. Clicking on a malicious URL can result in catastrophic outcomes, such as data breaches, financial losses, and identity theft. Detecting and blocking these websites is essential for protecting individuals and organizations from online threats, preserving data security, and sustaining confidence in online platforms. Researchers presented several methods for detecting malicious websites. Due to the threat's evolution, the problem remains unsolved. This paper presents a machine-learning model for malicious website identification. To process the textual contents, the experiment uses four different text vectorization techniques such as Count, TF-IDF, Hashing and word embedding (average Word2Vec). Eight machine learning models are tested to assess performance. The outcome demonstrates that the average word2vec embedding with extreme gradient boosting and random forest obtains 94.76% and 94.70% accuracy, respectively.

Email phishing is a manipulative technique aimed at compromising information security and user privacy. To overcome the limitations of traditional detection methods, such as blacklists, this research proposes a phishing detection model that leverages natural language processing (NLP) and deep learning technologies to analyze Indonesian email headers. The primary objective is to more efficiently detect zero-day phishing attacks by focusing on the unique linguistic and cultural context of the Indonesian language. This enables the development of models capable of recognizing phishing attack patterns that differ from those in other language contexts. Four models are tested, combining Indonesian bidirectional encoder representation of transformers (IndoBERT) and FastText feature extraction techniques with convolutional neural network (CNN) and long short-term memory (LSTM) deep learning algorithms. The results indicate that the combination of FastText and CNN achieved the highest performance in accuracy, precision, and F1-score metrics, each at 98.4375%. Meanwhile, the FastText model with LSTM showed the best performance in recall, with a score of 98.9583%. The research suggests exploring deeper into email content or integrating analysis between headers and email content in future studies to further improve accuracy and effectiveness in phishing email detection.

The problem of NLP should be divided into several small parts and solved step by step. In this article, where NLP is necessary at every stage of solving the problem, all forms of text processing are considered. The step-by-step text processing is called a pipeline process in NLP. When creating any NLP model, the pipeline process is a sequence of steps that must be carried out. The planning and development of the text processing is considered as the starting point for the creation of any NLP project. This article discusses the steps involved in implementing a pipeline process and their role in solving NLP tasks. This article analyzed the most common preliminary processing steps on the NLP processing pipeline. All processing stages are pre-trained in various NLP libraries, identified as usable models. If necessary, additional, modified preprocessing steps can be developed depending on the given problem condition. One can determine how a particular initial processing stage serves a given NLP problem by many experimentations.

Fake followers are those Twitter accounts specifically created to inflate the number of followers of a target account. Fake followers are dangerous for the social platform and beyond, since they may alter concepts like popularity and influence in the Twittersphere-hence impacting on economy, politics, and society. In this paper, we contribute along different dimensions. First, we review some of the most relevant existing features and rules (proposed by Academia and Media) for anomalous Twitter accounts detection. Second, we create a baseline dataset of verified human and fake follower accounts. Such baseline dataset is publicly available to the scientific community. Then, we exploit the baseline dataset to train a set of machine-learning classifiers built over the reviewed rules and features. Our results show that most of the rules proposed by Media provide unsatisfactory performance in revealing fake followers, while features proposed in the past by Academia for spam detection provide good results. Building on the most promising features, we revise the classifiers both in terms of reduction of overfitting and cost for gathering the data needed to compute the features. The final result is a novel Class A classifier, general enough to thwart overfitting, lightweight thanks to the usage of the less costly features, and still able to correctly classify more than 95% of the accounts of the original training set. We ultimately perform an information fusion-based sensitivity analysis, to assess the global sensitivity of each of the features employed by the classifier. The findings reported in this paper, other than being supported by a thorough experimental methodology and interesting on their own, also pave the way for further investigation on the novel issue of fake Twitter followers.

Fake followers are those Twitter accounts created to inflate the number of followers of a target account. Fake followers are dangerous to the social platform and beyond, since they may alter concepts like popularity and influence in the Twittersphere-hence impacting on economy, politics, and Society. In this paper, we contribute along different dimensions. First, we review some of the most relevant existing features and rules (proposed by Academia and Media) for anomalous Twitter accounts detection. Second, we create a gold standard of verified human and fake accounts. Then, we exploit the gold standard to train a set of machine-learning classifiers built over the reviewed rules and features. Most of the rules provided by Media provide unsatisfactory performance in revealing fake followers, while features provided by Academia for spam detection result in good performance. Building on the most promising features, we optimise the classifiers both in terms of reduction of overfitting and costs for gathering the data needed to compute the features. The final result is a "Class A" classifier, that is general enough to thwart overfitting and that uses the less costly features, while being able to correctly classify more than 95% of the accounts of the training set. The findings reported in this paper, other than being supported by a thorough experimental methodology and being interesting on their own, also pave the way for further investigation.

Email is an increasingly important and ubiquitous means of communication, both facilitating contact between individuals and enabling rises in the productivity of organizations. However, the relentless rise of automatic unauthorized emails, a.k.a. spam is eroding away much of the attractiveness of email communication. Most of the attention dedicated to spam detection has focused on the content of the emails or on the addresses or domains associated with spam senders. Although, methods based on these -easily changeable -identifiers work reasonably well, they miss on the fundamental nature of spam as an opportunistic relationship, very different from the normal mutual relations between senders and recipients of legitimate email. Here we present a comprehensive graph theoretical analysis of email traffic that captures these properties quantitatively. We identify several simple metrics that serve both to distinguish between spam and legitimate email and to provide a statistical basis for models of spam traffic.

In this paper, we examine the application of various grouping techniques to help improve the efficiency and reduce the costs involved in an electronic discovery process. Specifically, we create coherent groups of email documents which characterize either a syntactic theme, a semantic theme or an email thread. All such grouped documents can be reviewed together leading to a faster and more consistent review of documents. Syntactic grouping of emails is based on near duplicate detection whereas semantic grouping is based on identifying concepts in the email content using information extraction. Email thread detection is achieved using a combination of segmentation and near duplicate detection. We present experimental results on the Enron corpus that suggest that these approaches can significantly reduce the review time and show that high precision and recall in identifying the groups can be achieved. We also describe how these techniques are integrated into the IBM eDiscovery Analyzer product offering.