In today’s digitalized era, Online Social Networking platforms are growing to be a vital aspect of each individual’s daily life. The availability of the vast amount of information and their open nature attracts the interest of cybercriminals to create malicious bots. Malicious bots in these platforms are automated or semi-automated entities used in nefarious ways while simulating human behavior. Moreover, such bots pose serious cyber threats and security concerns to society and public opinion. They are used to exploit vulnerabilities for illicit benefits such as spamming, fake profiles, spreading inappropriate/false content, click farming, hashtag hijacking, and much more. Cybercriminals and researchers are always engaged in an arms race as new and updated bots are created to thwart ever-evolving detection technologies. This literature review attempts to compile and compare the most recent advancements in Machine Learning-based techniques for the detection and classification of bots on five primary social media platforms namely Facebook, Instagram, LinkedIn, Twitter, and Weibo. We bring forth a concise overview of all the supervised, semi-supervised, and unsupervised methods, along with the details of the datasets provided by the researchers. Additionally, we provide a thorough breakdown of the extracted feature categories. Furthermore, this study also showcases a brief rundown of the challenges and opportunities encountered in this field, along with prospective research directions and promising angles to explore.

Detection of minor leaks in oil or gas pipelines is a critical and persistent problem in the oil and gas industry. Many organisations have long relied on fixed hardware or manual assessments to monitor leaks. With rapid industrialisation and technological advancements, innovative engineering technologies that are cost-effective, faster, and easier to implement are essential. Herein, machine learning-based anomaly detection models are proposed to solve the problem of oil and gas pipeline leakage. Five machine learning algorithms, namely, random forest, support vector machine, k-nearest neighbour, gradient boosting, and decision tree, were used to develop detection models for pipeline leaks. The support vector machine algorithm, with an accuracy of 97.4%, overperformed the other algorithms in detecting pipeline leakage and thus proved its efficiency as an accurate model for detecting leakage in oil and gas pipelines.

The significant growth in the use of the Internet and the rapid development of network technologies are associated with an increased risk of network attacks. Network attacks refer to all types of unauthorized access to a network including any attempts to damage and disrupt the network, often leading to serious consequences. Network attack detection is an active area of research in the community of cybersecurity. In the literature, there are various descriptions of network attack detection systems involving various intelligent-based techniques including machine learning (ML) and deep learning (DL) models. However, although such techniques have proved useful within specific domains, no technique has proved useful in mitigating all kinds of network attacks. This is because some intelligent-based approaches lack essential capabilities that render them reliable systems that are able to confront different types of network attacks. This was the main motivation behind this research, which evaluates contemporary intelligent-based research directions to address the gap that still exists in the field. The main components of any intelligent-based system are the training datasets, the algorithms, and the evaluation metrics; these were the main benchmark criteria used to assess the intelligent-based systems included in this research article. This research provides a rich source of references for scholars seeking to determine their scope of research in this field. Furthermore, although the paper does present a set of suggestions about future inductive directions, it leaves the reader free to derive additional insights about how to develop intelligent-based systems to counter current and future network attacks.

These days, we are witnessing unprecedented challenges to network security. This indeed confirms that network security has become increasingly important. Firewall logs are important sources of evidence, but they are still difficult to analyze. Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) have emerged as effective in developing robust security measures due to the fact that they have the capability to deal with complex cyberattacks in a timely manner. This work aims to tackle the difficulty of analyzing firewall logs using ML and DL by building multiclass ML and DL models that can analyze firewall logs and classify the actions to be taken in response to received sessions as “Allow”, “Drop”, “Deny”, or “Reset-both”. Two sets of empirical evaluations were conducted in order to assess the performance of the produced models. Different features set were used in each set of the empirical evaluation. Further, two extra features, namely, application and category, were proposed to enhance the performance of the proposed models. Several ML and DL algorithms were used for the evaluation purposes, namely, K-Nearest Neighbor (KNN), Naïve Bayas (NB), J48, Random Forest (RF) and Artificial Neural Network (ANN). One interesting reading in the experimental results is that the RF produced the highest accuracy of 99.11% and 99.64% in the first and the second experiments respectively. Yet, all other algorithms have also produced high accuracy rates which confirm that the proposed features played a significant role in improving the firewall classification rate.

The Internet of Things (IoT) has become widely adopted in businesses, organizations, and daily lives. They are usually characterized by transferring and processing sensitive data. Attackers have exploited this prospect of IoT devices to compromise user data’s integrity and confidentiality. Considering the dynamic nature of the attacks, artificial intelligence (AI)-based techniques incorporating machine learning (ML) are promising techniques for identifying such attacks. However, the dataset being utilized features engineering techniques, and the kind of classifiers play significant roles in how accurate AI-based predictions are. Therefore, for the IoT environment, there is a need to contribute more to this context by evaluating different AI-based techniques on datasets that effectively capture the environment’s properties. In this paper, we evaluated various ML models with the consideration of both binary and multiclass classification models validated on a new dedicated IoT dataset. Moreover, we investigated the impact of different features engineering techniques including correlation analysis and information gain. The experimental work conducted on bagging, k-nearest neighbor (KNN), J48, random forest (RF), logistic regression (LR), and multi-layer perceptron (MLP) models revealed that RF achieved the highest performance across all experiment sets, with a receiver operating characteristic (ROC) of 99.9%.

Tokenization is an important early step in natural language processing (NLP) tasks. The idea is to split the input sentence into smaller units, called tokens, for further processing. Words are the most commonly used tokens in text classification tasks but other tokenization ideas are also popular such as subword and character tokens. The increasing availability of training corpora has posed challenges for the word tokenization technique, primarily due to vocabulary size expansion. This has underscored the importance of exploring alternative tokenization, especially for morphologically rich languages like Arabic. In this study, we assess the efficacy of character trigrams for Arabic sentiment analysis and text classification, particularly their robustness against misspelled words. We compare character trigrams with word and WordPiece tokenization across five datasets, encompassing tweets sentiment analysis, reviews sentiment analysis, and news classification. These datasets, which range from small to large in vocabulary size, facilitate a comparative examination of vocabulary and Out-Of-Vocabulary (OOV) sizes for word and character trigram embeddings. The word and character trigram embeddings are integrated into a deep learning (DL) model featuring a dense layer connected to a Bidirectional Gated Recurrent Unit (BiGRU) for classification purposes. Meanwhile, WordPiece embeddings serve to fine-tune the AraBert model. Our findings reveal that the word embedding approach, when applied to extensive corpora, leads to a significant increase in vocabulary and OOV rates, whereas character trigram embedding maintains manageable sizes for both. On test sets devoid of spelling mistakes, the AraBert model surpasses the other models in performance. However, both word and character trigram embedding models exhibit similar levels of performance. In contrast, datasets with misspellings reveal a performance degradation for the WordPiece and word embedding models, with decline rates ranging between 2%-14% and 5%-19%, respectively. Meanwhile, the character trigram models exhibit stable performances, with a drop rate of 0%-8%. Notably, the performance of WordPiece and word embedding models suffers due to their inability to recognize misspelled words.