Email Data Research Articles

An Optimized Approach for Detection and Classification of Spam Email’s Using Ensemble Methods

AbstractSince the advent of email services, spam emails have been a major concern because users’ security depends on the classification of emails as ham or spam. It’s a malware attack that has been used for spear phishing, whaling, clone phishing, website forgery, and other harmful activities. However, various ensemble Machine Learning (ML) algorithms used for the detection and filtering of spam emails have been less explored. In this research, we offer a ML-based optimized algorithm for detecting spam emails that have been enhanced using Hyper-parameter tuning approaches. The proposed approach uses two feature extraction modules, namely Count-Vectorizer and TFIDF-Vectorizer that provide the most effective classification results when we apply them to three different publicly available email data sets: Ling Spam, UCI SMS Spam, and the Proposed dataset. Moreover, to extend the performance of classifiers we used various ML methods such as Naive Bayes (NB), Logistic Regression (LR), Extra Tree, Stochastic Gradient Descent (SGD), XG-Boost, Support Vector Machine (SVM), Random Forest (RF), Multi-layer Perception (MLP), and parameter optimization approaches such as Manual search, Random search, Grid search, and Genetic algorithm. For all three data sets, the SGD outperformed other algorithms. All of the other ensembles (Extra Tree, RF), linear models (LR, Linear-SVC), and MLP performed admirably, with relatively high precision, recall, accuracies, and F1-score.

Leaked email data: a new source for the study of authoritarian regimes

Leaked email data: a new source for the study of authoritarian regimes

Sentiment classification for insider threat identification using metaheuristic optimized machine learning classifiers

This study examines the formidable and complex challenge of insider threats to organizational security, addressing risks such as ransomware incidents, data breaches, and extortion attempts. The research involves six experiments utilizing email, HTTP, and file content data. To combat insider threats, emerging Natural Language Processing techniques are employed in conjunction with powerful Machine Learning classifiers, specifically XGBoost and AdaBoost. The focus is on recognizing the sentiment and context of malicious actions, which are considered less prone to change compared to commonly tracked metrics like location and time of access. To enhance detection, a term frequency-inverse document frequency-based approach is introduced, providing a more robust, adaptable, and maintainable method. Moreover, the study acknowledges the significant impact of hyperparameter selection on classifier performance and employs various contemporary optimizers, including a modified version of the red fox optimization algorithm. The proposed approach undergoes testing in three simulated scenarios using a public dataset, showcasing commendable outcomes.

A Design of an SMTP Email Server

This study developed a mail server program using Socket API and Python. The program uses the Hypertext Transfer Protocol (HTTP) to receive emails from browser clients and forward them to actual email service providers via the Simple Mail Transfer Protocol (SMTP). As a web server, it handles Transmission Control Protocol (TCP) connection requests from browsers, receives HTTP commands and email data, and temporarily stores the emails in a file. Simultaneously, as an SMTP client, the program establishes a TCP connection with the actual mail server, sends SMTP commands, and transmits the previously saved emails. In addition, we also analyzed security issues and the efficiency and availability of this server, providing insights into the design of SMTP mail servers.

Correction: Analysing the email data using stylometric method and deep learning to mitigate phishing attack

Correction: Analysing the email data using stylometric method and deep learning to mitigate phishing attack

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.

Battle for Inbox and Bucks

Despite its proven pervasiveness and effectiveness, the role of email in political campaigns has been understudied. In this article, we seek to understand the use of emails by Donald Trump's and Joe Biden's campaigns during the 2020 U.S. presidential election. Employing newly collected email data (N = 4051), we compared how the two campaign teams differ in the use of appeals in their emails, and the functions of the message (e.g., attack, acclaim). We found that, while both campaigns used emails primarily for fundraising, Trump's campaign (compared to Biden’s campaign) was more likely to use all of the four appeals we defined — material, solidarity, ideology, and issue appeals. We also found that by sending emails, Biden's campaign tends to acclaim itself more while Trump's campaign attacks opponents more.

A novel and secured email classification using deep neural network with bidirectional long short-term memory

Email data has some characteristics that are different from other social media data, such as a large range of answers, formal language, notable length variations, high degrees of anomalies, and indirect relationships. The main goal in this research is to develop a robust and computationally efficient classifier that can distinguish between spam and regular email content. The benchmark Enron dataset, which is accessible to the public, was used for the tests. The six distinct Enron data sets we acquired were combined to generate the final seven Enron data sets. The dataset undergoes early preprocessing to remove superfluous sentences. The proposed model Bidirectional Long Short-Term Memory (BiLSTM) apply spam labels and to examine email documents for spam. On seven Enron datasets, DNN-BiLSTM performs better than other classifiers in the performance comparison in terms of accuracy. DNN-BiLSTM and convolutional neural networks demonstrated that they can classify spam with 96.39 % and 98.69 % accuracy, respectively, in comparison to other machine learning classifiers. The risks associated with cloud data management and potential security flaws are also covered in the paper. This research presents hybrid encryption as a means of protecting cloud data while preserving privacy by using the hybrid AES-Rabit encryption algorithm which is based on symmetric session key exchange.

Analysing the email data using stylometric method and deep learning to mitigate phishing attack

Analysing the email data using stylometric method and deep learning to mitigate phishing attack

Primary Care Clinic Approaches to Facilitating Patient Health Behavior Change in Alabama.

Background Non-communicable chronic diseases (NCCDs), such as cardiovascular disease, diabetes, and cancer, are the leading cause of death and disability and the leading driver of healthcare costs in the U.S. It is estimated that 80% of chronic diseases and premature deaths are attributable to modifiable lifestyle factors related to smoking and alcohol intake, poor eating patterns, and physical inactivity. Inadequate sleep also plays a significant role. Among other directives, primary care providers (PCPs) have the opportunity to contribute to preventing and treating NCCD in their patients. Comprehensive, evidence-based behavioral counseling interventions are recommended to PCPs as a first-line approach to improving outcomes. However, presumably due to a lack of PCP time, training or resources, most patients report not receiving such services. Currently, the extent to which PCPs in Alabama offer or refer patients to health behavior change (HBC) services is unknown. Objectives This study aims to assess the following: (1) Alabama PCPs' current approaches in facilitating patient HBC in the domains of eating patterns, physical activity, sleep, and stress and (2) the likelihood of the Alabama PCPsreferring patients to virtual HBC programs, once developed by an osteopathic medical school in the state. Methods Data were collected from clinic personnel who were knowledgeable regarding the clinic's approach to facilitating patient HBC via scripted telephone interviews and online surveys sent via email. The clinic list utilized for the study was derived from a list of VCOM-Auburn clinical preceptors. Primary care and specialty clinics were included. Data were analyzed descriptively to determine the number of clinics that (1) provide, recommend, or refer programs, services, or resources to patients to facilitate HBC related to eating patterns, physical activity, sleep, and stress management and (2) are likely to refer patients to free virtual HBC programs, once developed by an osteopathic medical school in the state. Results Of the 198 clinics that were contacted, 75 were excluded, 46 were "no response," 53 agreed to participate, and 50 completed the survey. Of the 50 clinics that completed the survey, 33 indicated offering resources or referrals for diet, 29 stated they offered resources or referral services for physical activity, 33 indicated offering resources or referrals for sleep, and 28 indicated offering or recommending resources for stress management to patients. Most of the clinics (29/50) felt that their patients would benefit most from a program that facilitates improvement in eating patterns, and 41/50 clinics said that they are either "somewhat" or "extremely" likely to refer patients to a free VCOM-Auburn HBC program, once available. Conclusions Findings indicate that a significant percentage of PCP clinics are not offering HBC resources to patients and that most PCP clinics would consider referring patients to free VCOM-Auburn HBC programs, once available. Phone data were significantly different from email data. The primary limitations were a low response rate and potential response bias.

Towards privacy-aware exploration of archived personal emails

This paper examines how privacy measures, such as anonymisation and aggregation processes for email collections, can affect the perceived usefulness of email visualisations for research, especially in the humanities and social sciences. The work is intended to inform archivists and data managers who are faced with the challenge of accessioning and reviewing increasingly sizeable and complex personal digital collections. The research in this paper provides a focused user study to investigate the usefulness of data visualisation as a mediator between privacy-aware management of data and maximisation of research value of data. The research is carried out with researchers and archivists with vested interest in using, making sense of, and/or archiving the data to derive meaningful results. Participants tend to perceive email visualisations as useful, with an average rating of 4.281 (out of 7) for all the visualisations in the study, with above average ratings for mountain graphs and word trees. The study shows that while participants voice a strong desire for information identifying individuals in email data, they perceive visualisations as almost equally useful for their research and/or work when aggregation is employed in addition to anonymisation.

Negotiating joint commitment in collaborative work project: Focus on text-based requests and news deliveries in atypical work

The shift to service and gig economy and increasing polymediality have created communicative contexts where the workers have to construct varying social relations in different kinds of digital and text-based interaction environments. This article examines how transprofessional collaboration is managed in such contexts in the field of applied arts. Based on email and mobile messaging data, we study a project where an artist creates an installation artwork for the use of an organisation. By applying the methods of conversation analysis, we investigate text-based requests and news delivery sequences related to applying for external project funding. Our results show how the participants negotiate the aspects of knowledge, power and emotion within these sequences, and by doing so, maintain mutual professional relationships and display various levels of commitment to the continuation of the project. The article illuminates the facets of transprofessional collaboration in digital professional communication and in atypical work.

When East Meets West: A Comparative Analysis of Credibility and Persuasive Appeals in Cross-Cultural Email Negotiations between Genders

The bulk of studies focusing on cross-cultural variations in persuasive appeals examines letters or newspaper articles (Al-Ali 2006; Alhudhaif, 2005; Bhatia, 2014; Connor et al., 1995; Ismail, 2010). It is rare for a study to examine persuasive appeals within intercultural settings via email communication for academic proposal purposes. Typical findings of these cross-cultural studies suggest that non-native English speakers do not use rational appeals as frequently as native speakers and focus more on effective appeals. However, these studies did not base their judgements on interactive communications and the cross-cultural comparison of newspapers or political texts is hardly comparable. To this end, this paper provides a more fine-grain analysis by linking between the new impoliteness framework and persuasive appeals (impeded within genre analysis) to explore gender and cultural differences. the differences in email negotiations persuasive appeals where relevant by using comparable email data written by both cultural groups (i.e. Australians and Saudis). The results indicate that both focused on rational appeals, using far fewer credible and affective appeals. However, there were significant differences in the way both cultures utilised these appeals. In the affective appeals, Saudi men made more compliments and used more emotional and (im) polite requests than Saudi females. With the credibility appeals, Saudi women made more self-promotional moves than men, portraying the self as competent and distinguished.

A novel and secured email classification and emotion detection using hybrid deep neural network

Compared to other social media data, email data differs from it in various topic-specific ways, including extensive replies, formal language, significant length disparities, high levels of anomalies, and indirect linkages. In this paper, the creation of a potent and computationally effective classifier to categorize spam and ham email documents is proposed. To assess and validate spam texts, this paper employs a variety of data mining-based classification approaches. On the benchmark Enron dataset, which is open to the public, tests were run. The final 7 Enron datasets were created by combining the six different types of Enron datasets that we had acquired. We preprocess the dataset at an early stage to exclude any useless phrases. This method falls under several categories, including Logistic Regression (LR), Convolutional Neural Networks (CNN), Random Forests (RF), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), and suggested Deep Neural Networks (DNN). Using Bidirectional Long Short-Term Memory (BiLSTM), email documents may be screened for spam and labeled as such. In performance comparisons, DNN-BiLSTM outperforms other classifiers in terms of accuracy on all seven Enron datasets. In comparison to other machine learning classifiers, the findings demonstrate that DNN-BiLSTM and Convolutional Neural Networks can categorize spam with 96.39 % and 98.69 % accuracy, respectively. The report also covers the dangers of managing cloud data and the security problems that might occur. To safeguard data in the cloud while maintaining privacy, hybrid encryption is examined in this white paper. In the AES-Rabit hybrid encryption system, the symmetric session key exchange-based Rabit technique is combined with the benefits of the AES algorithm for faster data encryption.

Transformer models for mining intents and predicting activities from emails in knowledge-intensive processes

Process mining is an interdisciplinary field that combines Artificial Intelligence and Business Process Management to extract insights from historical event data. Knowledge-intensive processes, which predominantly involve knowledge work, are often inadequately monitored by process-aware information systems. Consequently, the event data necessary for applying process mining techniques are frequently unavailable. Emails are widely used in knowledge-intensive processes for scheduling meetings, sharing documents, and reporting on the completion of outstanding tasks, which makes them suitable candidates for replacing event logs as the primary data source. In this work, we focus on the task of extracting the set of next recommended activities from incoming emails. Yet, we face two major challenges. Firstly, emails do not express process information explicitly but rather contain subtext that implies what the next best actions would be. Secondly, email data lacks domain-specific labels that would enable the use of machine learning. We overcome these limitations by utilizing an email taxonomy to represent user intents, thus bridging the gap between textual information and process semantics, as well as leveraging pre-trained transformer models applied in zero-shot and few-shot settings that require little to no labeled email data. An evaluation of our method on real-world unlabeled email communications demonstrates its effectiveness in recognizing intents and extracting activities.

Modeling and monitoring multilayer attributed weighted directed networks via a generative model

As data with network structures are widely seen in diverse applications, the modeling and monitoring of network data have drawn considerable attention in recent years. When individuals in a network have multiple types of interactions, a multilayer network model should be considered to better characterize its behavior. Most existing network models have concentrated on characterizing the topological structure among individuals, and important attributes of individuals are largely disregarded in existing works. In this article, first, we propose a unified static Network Generative Model (static-NGM), which incorporates individual attributes in network topology modeling. The proposed model can be utilized for a general multilayer network with weighted and directed edges. A variational expectation maximization algorithm is developed to estimate model parameters. Second, to characterize the time-dependent property of a network sequence and perform network monitoring, we extend the static-NGM model to a sequential version, namely, the sequential-NGM model, with the Markov assumption. Last, a sequential-NGM chart is developed to detect shifts and identify root causes of shifts in a network sequence. Extensive simulation experiments show that considering attributes improves the parameter estimation accuracy and that the proposed monitoring method also outperforms the three competitive approaches, static-NGM chart, score test-based chart (ST chart) and Bayes factor-based chart (BF chart), in both shift detection and root cause diagnosis. We also perform a case study with Enron E-mail data; the results further validate the proposed method.

Spam Spoiler for Public and Private E-Mail Services

Abstract: In recent years, cyber security incidents have occurred frequently. In most of these incidents, attackers have used different types of spam email as a knock-on to successfully invade government systems, well-known companies, and websites of politicians and social organizations in many countries. The detection of spam mail from big email data has been paid public attention. However, the camouflage technology of spam mail is becoming more and more complex, and the existing detection methods are unable to confront the increasingly complex deception methods and the growing number of emails. In this project, we proposed to design a novel efficient approach named Spam Spoiler for big email data classification into four different classes: Normal, Fraudulent, Harassment, and Suspicious E-mails by using LSTM-based GRU. The new method includes two important stages, the sample expansion stage and the testing stage under sufficient samples. This project the LSTM-based GRU efficiently captures meaningful information from E-mails that can be used for forensic analysis as evidence. Experimental results revealed that Spam Spoiler performed better than existing ML algorithms and achieved a classification accuracy of 98% using the novel technique of LSTM with recurrent gradient units. As different types of topics are discussed in E-mail content analysis. Spam Spoiler effectively outperforms existing methods while keeping the classification process robust and reliable.

Spam Email Filtering Leveraging Improved Text Convolutional Neural Network

This paper outlines the development of anti-spam technology in the last two decades and introduces a novel approach using a convolution neural network (CNN) to tackle the problem of spam filtering. The study uses the TREC06c dataset, containing Chinese spam email data, the dataset is split between a training set and a test set. The paper also introduces the concept of word embeddings, which converts each word in the text into a real-valued vector, better reflecting the semantic relationships between words. The TEXT-CNN algorithm is then discussed, which applies convolutional neural networks to text data and is generated by modifying the TEXT-CNN model to improve its performance in the spam filter. The conclusion of this article is that TEXT-CNN model demonstrates great results in the task of identifying spam emails, and the classification efficiency can be further improved by introducing an attention mechanism and batch processing mechanism by improving the model. The article also provides some ideas for further improvement.

SPAM EMAIL DETECTION USING DIFFERENT CLASSIFIERS

Due to the widespread use of social media, the amount of unwanted emails has increased, necessitating the implementation of a reliable system to filter them out. Spam emailsare now the most common issue on the internet, despite email beingone of the fastest and most cost-effective forms of communication. Over the past few years, there has been a significant increase in spam emails due to the growing number of email subscribers. This study employs four classifiers (Random Forest, XG Boost, Naïve Bayesian) to classify email data, with varying data and feature sizes. The final classification result is '1' if the email is spam and '0' if it is not. The study was conducted using Python and implemented in a Jupyter notebook. This technique can also be utilized in monitoring social media and brand activity. The paper presents a machine learning approach for identifying spam emails by detecting spam content within the message. With machine learning, computers can learn how to perform a task without being explicitly programmed. This method uses data to generate a program that performs a task, such as classification. Unlike knowledge engineering, machine learning techniques require pre-classified data to create a training dataset that is used to fit the learning algorithm in the machine learning studio. The objective of this study is to train, test, and compare various classifiers. The rest of the paper is organized as follows: Section 2 defines the researchers' contribution in this field. Section 3 describes the experimentation framework, dataset, procedures, andlibraries. Section 4 summarizes the findings, and Section 5 concludes the paper's potential for future research.

Streamlining IT Help Desk and Incident Management: Harnessing the Power of the ITIL Framework for Enhanced Efficiency in IT Services

This study examines the potential of leveraging incident management and IT Helpdesk principles from ITIL V3 to enhance the daily operations of an IT division. The research involved collecting email communication data between users and the IT division, as well as conducting interviews with individuals associated with the IT division and various company employees. The ITIL Maturity Level (ITML) method was utilized to analyze the findings, revealing that IT Helpdesk achieved a score of 3.5 (Quality Control), while incident management received a score of 2 (Process Capability). However, both areas fell short of meeting the requirements for level 3 (Products) according to the ITIL Self-Assessment table, resulting in a classification of "Not Comply." Nevertheless, incident management demonstrated the ability to standardize workflows and sustain service performance, which can be continually monitored and reviewed to ensure stable quality of IT services.