The rise of cyberbullying in digital communication platforms has triggered widespread concern, not just for its reach but for the lasting psychological harm caused. Identifying such harmful behavior online is difficult in general, but when the target language is Arabic, the task becomes more complicated. The issue is not just that Arabic is written in multiple dialects, each with its own informal vocabulary, spelling variations, and structure. What complicates matters further is that meaning often shifts based on region, tone, and the social context, making abusive content harder to catch using conventional tools. This study aims to improve Arabic cyberbullying detection mechanisms by introducing a feature-selection strategy. Its main contribution involves utilizing a Genetic Algorithm (GA)-based feature selector to pinpoint harmful language patterns in a corpus of 46k Arabic Instagram comments. The GA effectively reduced the feature space by approximately half, preserving essential semantic structures while removing noise and redundancy. Four classifiers were evaluated, and GA-driven selection improved F1-scores by (3.45-14.96%) and reduced classification time by a factor of 2.32-12. These findings suggest that genetic-feature optimization enhances model precision while significantly improving runtime and reducing complexity, thereby enabling scalable, context-sensitive cyberbullying detection for Arabic and morphologically rich languages.

In human cognitive psychology, the greatest planners adapt their strategies to their present situation and their appraisal of the best experiences of others. Based on this concept, we proposed an update to the Cat Swarm Optimisation (CSO) model, which we believe will enhance performance and promote convergence. The inspiration method focuses on using a cat's self-perception to guide its search direction based on its current position. The top-performing cat should focus solely on its own position without influence from others, simulating human confidence. Conversely, the remaining other cats are guided by their past successes and influenced by the global best position, emulating human reliance on personal experience. So, in this study, we revised the original CSO and suggested Self-Regulating Cat Swarm Optimisation (SR-CSO) as an enhancement over the CSO algorithm. The performance of SR-CSO is examined by combining the SRCSO with the Random Forest “RF” classifier for feature selection on large data. The findings reveal that it improves classification results for the six benchmark datasets used in the experiment. The experiment results were compared to the original CSO and Improved cat swarm optimization (ICSO), results show that the suggested SRCSO algorithm achieves much quicker convergence and higher accuracy than the basic CSO and ICSO.

Botnets are a major security threat in the Internet of Things (IoT), posing significant risks to user privacy, network availability, and the integrity of IoT devices. With the increasing availability of large datasets that contain hundreds or even thousands of variables, selecting the right set of features can be a challenging task. Feature selection is a critical step in developing effective machine learning-based botnet detection systems, as it enables the selection of a subset of features that are most relevant for detection. This paper provides a comprehensive review of filtering based feature selection techniques for botnet detection in IoT. It examines a range of filtering based techniques and evaluates their effectiveness in addressing the challenges and limitations of botnet detection in IoT. It aims to identify the gaps in the literature and areas for future research, and discuss the broader implications of findings for the field of IoT botnet detection. This review provides valuable insights and guidance for researchers and practitioners working on botnet detection in IoT, and highlights the importance of effective feature selection in developing robust and reliable detection systems.

The proliferation of internet of things (IoT) devices has led to unprecedented connectivity and convenience. However, this increased interconnectivity has also introduced significant security challenges, particularly concerning the detection and mitigation of botnet attacks. Detecting botnet activities in IoT environments is challenging due to the diverse nature of IoT devices and the large-scale data generated. Artificial intelligence and machine learning based approaches showed great potential in IoT botnet detection. However, as these approaches continue to advance and become more complex, new questions are opened about how decisions are made using such technologies. Integrating an explainability layer into these models can increase trustworthy and transparency. This paper proposes the utilization of explainable artificial intelligence (XAI) techniques for improving the interpretability and transparency of the botnet detection process. It analyzes the impact of incorporating XAI in the botnet detection process, including enhanced model interpretability, trustworthiness, and potential for early detection of emerging botnet attack patterns. Three different XAI based techniques are presented i.e. rule extraction and distillation, local interpretable model-agnostic explanations (LIME), Shapley additive explanations (SHAP). The experimental results demonstrate the effectiveness of the proposed approach, providing valuable insights into the inner workings of the detection model and facilitating the development of robust defense mechanisms against IoT botnet attacks. The findings of this study contribute to the growing body of research on XAI in cybersecurity and offer practical guidance for securing IoT ecosystems against botnet threats.

To train a machine to “sense” a users’ feelings through writings (sentiment analysis) has become a crucial process in several domains: marketing, research, surveys and more. Nevertheless in times of crisis like COVID. Typo is one of the underestimated challenges processing user-generated text (comments, tweets, ..etc), it affects both learning and evaluation processes. Word tokenization outcome changes drastically even with a single character change, hence as expected, experiments have shown significant accuracy decreases due to typo. Adding a spelling correction as preprocessing layer, building one for every language, is a very time and resources expensive solution, a huge challenge against large data and real-time processing. Alternatively, a CNN model consuming the same text, once tokenized on characters level and once on words level while inducing typo, showed that as the typo percentage approaches 10% of the text, the results with characters tokens surpasses words tokens. Finally, on %30 typo of the text, the model consuming characters tokenization outperformed itself with the word level by a significant %22.3 in accuracy and %24.9 in F1-Score, using the same exact model. This approach in solving the inevitable typo challenge in NLP proved to be of significant practicality, saving huge resources versus using a spelling-correction model beforehand. It also removes a blocker challenge in front of real-time processing of user-generated text while preserving acceptable accuracy results.

Social media platforms are among the most widely used means of communication. However, some individuals exploit these platforms for nefarious purposes, with "cyberbullying" being particularly prevalent. Cyberbullying, which involves using electronic means to harass or harm others, is especially common among young people. Consequently, this study aims to propose a model for detecting cyberbullying using a deep learning algorithm. Three datasets from Twitter, Instagram, and Facebook were utilized to predict instances of bullying using the Long Short-Term Memory (LSTM) method. The results obtained revealed the development of an effective model for detecting cyberbullying, addressing challenges faced by previous cyberbullying detection techniques. The model achieved accuracies of approximately 96.64%, 94.49%, and 91.26% for the Twitter, Instagram, and Facebook datasets, respectively.

Ever since the invention of software, change has been a destabilizing factor. Although many new software changes are being applied, the terminologies used to describe them are often inconsistent. This restricts practitioners to designing and evaluating their changes. This article aims to develop a conceptual framework of software change based on six main dimensions regarding the source, essence, and consequences of software change. To evaluate the proposed framework, benchmarking is applied against selected 11 previous studies.

Internet of Things is shaping the quality of living standard. With the rapid growth and expansion of adopting IoT-based approaches, their security represents a growing challenge for both manufacturers and consumers. There is a recent rising trend towards employing artificial intelligence approaches to enhance the security of IoT infrastructure. This survey paper focuses on reviewing recent developments in applying artificial intelligence to intrusion detection in the IoT domain. Selected articles are classified according to the applied AI algorithm. This study provides an in-depth survey highlighting the recent advances in artificial intelligence for improving the security of IoT. It summarizes and organizes the recent related research, then presents a comprehensive discussion on research challenges, open issues, and needed future research.

Software-defined networks (SDNs) have been growing rapidly due to their ability to provide an efficient network management approach compared to traditional methods. However, one of the major challenges facing SDNs is the threat of Distributed Denial of Service (DDoS) attacks, which can severely impact network availability. Detecting and mitigating such attacks is challenging, given the constantly evolving range of attack techniques. In this paper, a novel hybrid approach is proposed that combines statistical methods with machine-learning capabilities to address the detection and mitigation of DDoS attacks in SDN environments. The statistical phase of the approach utilizes an entropy-based detection mechanism, while the machine-learning phase employs a clustering mechanism to analyze the impact of active users on the entropy of the system. The k-means algorithm is used for clustering. The proposed approach was experimentally evaluated using three modern datasets, namely, CIC-IDS2017, CSE-CIC-2018, and CICIDS2019. The results demonstrate the effectiveness of the system in detecting and blocking sudden and rapid attacks, highlighting the potential of the proposed approach to significantly enhance security against DDoS attacks in SDN environments.

Elections are the most effective democratic mechanism for encouraging people to choose their representatives, which significantly impacts a nation's future and the lives of its residents. The election can be organized as traditional paper-based or electronic voting (e-voting). Traditional elections have challenges with privacy, security, transparency, and efficiency. Furthermore, it takes too long to count the votes, and in e-voting, the validity and integrity of the votes cast cannot be trusted by a sizable number of people. Therefore, security and privacy concerns persist. This paper presents e-voting based on blockchain, which provides a decentralized ledger that is distributed, secure, and immutable. Additionally, the proposed system adds layers of encryption to the votes by using AES as a symmetric key and RSA as an asymmetric key to prevent administrators or users who have access to system data from hacking voter privacy or influencing the transparency of the election. The result of the model is the protection of voter privacy and increased transparency in the election process. In several experiments, the transaction timing performance is evaluated and compared with other research results, confirming that the model performs better.

In the digital data landscape, businesses are increasingly reliant on advanced tools for extracting actionable insights, with sentiment analysis being at the forefront. This study breaks new ground by exploring sentiment analysis in audio data through the lens of Neutrosophy. Our approach hinges on Harris Hawk Optimization (HHO) augmented with Neutrosophic Sets for efficient feature selection. Utilizing the Libri TTS train clean 100 dataset, our experiments demonstrate the superiority of our model in reducing feature numbers while enhancing sentiment analysis accuracy. Remarkably, the Neutrosophic Sets significantly boost HHO's performance, achieving a best fitness of 0.96, and enabling the selection of a minimal number of features (39 out of 300) with a rapid convergence rate. Classification accuracy, using SVM, KNN, and DT classifiers, shows a notable improvement over other fitness functions, particularly with Neutrosophic Sets, where accuracy increases to 0.96. Further, our model outperforms baseline algorithms in sentiment classification, with an 18\% error rate reduction compared to the nearest competitor. Clustering analysis of the LibriSpeech dataset using Single-Valued Neutrosophic Sets reveals insightful patterns, underlining the model's potential to transform sentiment analysis by integrating both audio and textual data. This research not only advances sentiment analysis techniques but also pioneers the application of Neutrosophy in audio data analysis, offering novel insights and a robust framework for businesses and researchers in the digital era.

Image compression is a crucial task in image processing and in the process of sending and receiving files. There is a need for effective techniques for image compression as the raw images require large amounts of disk space to defect during transportation and storage operations. The most important objective of image compression is to decrease the redundancy of the image which helps in increasing the storage capacity and then efficient transmission. This study introduces a system for lossless image compression that is built to work on fingerprint image compression. It uses lossless compression to take care of the first image during processing. However, there is a serious problem which is the low ratio of compression. In order to make the ratio higher, there are five lossless compression techniques used in this study which are Elias Gamma Coding (EGC), Huffman Coding (HC), Arithmetic Coding (AC), Run-Length Encoding (RLE) and Lempel Ziv Welch (LZW). With these techniques, there are three types of transforms are used; they are Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT), and Discrete Shearlet Transform (DST). The results conclude that discrete shearlet transform with the Lempel-Ziv Welch coding technique outperforms the other lossless compression techniques and its Compression Ratio (CR) is 3.678023.

Cryptography plays a vital role in protecting information that has increased as a result of digitalization.Personal sensitive information, including electrocardiogram (ECG) signals, is widely transferred around the world, and as a result, protecting the data from unauthorized access by attackers is critical.One of the algorithms that is frequently used is the advanced encryption standard (AES) algorithm, due to its remarkable reliability and its usage in a wide range of applications.However, key exchange is still necessary to execute computations on the encrypted data, whereas time is considered the essence of the efficiency of the encryption algorithm.This study proposes and investigates the use of the proposed improved reduced round AES algorithm in conjunction with the new proposed fully homomorphic encryption (FHE) in order to preserve data privacy and eliminate key exchange restrictions.This research was employed to process, encrypt, and decrypt digital ECG signals.The suggested algorithm improves the security level and encryption and decryption times by using fewer rounds of encryption.In order to achieve this goal, it was suggested that the evaluation process be added to the AES algorithm as an extra level of security.With the evaluation process, it is possible to execute a number of computation operations homomorphically on the encrypted data without decrypting it.Additionally, the number of rounds in the AES encryption was decreased from 10 to 5. Results indicate that the proposed algorithm would take 5.39475 × 10 32 years to break, making it more efficient than the traditional AES and others.Also, the proposed algorithm has a classification sensitivity of 95.83% while simultaneously displaying an accuracy of 100%.Additionally, it may be utilized for real-time internet of things (IoT) applications.

many organizations and research groups have adopted methods, models, and standards to improve their research. However, despite these efforts, they can still find it difficult to put clear planning framework for the research that can be used for both researchers and supervisors. This paper aims to provide a theoretical framework for the researchers who are about to write their research proposal. This paper proposes a systematic framework for planning scientific research. The proposed framework extends is based on the multi-method approach for research developed by Nunamaker et al. The Framework is developed and applied based on a systematic review of existing literature and is complemented with the authors’ practical experience. Finally, the framework validation is presented through five real cases studies.

Managing change is a challenging task in today’s complex software engineering. Understanding the diversity of changes and their relationship to current technologies is critical for dealing with volatile business systems. This article aims to identify and assess the state of the art toward understanding software change for the sake of providing a deeper understanding of its causes, mechanisms, and effects.

Recently, Internet of Things (IoT) infrastructures are developing various applications in sustainable smart cities and societies. However, there are numerous challenges in smart cities, such as security, privacy, trust, verifiability, communication latency, scalability, and centralization preventing faster adaptations of IoT. Machine Learning (ML) is an important analytic tool that provides a scalable and accurate analysis of data in real time. However, there are several obstacles to designing and developing a usable large data analysis tool utilizing ML, such as centralized architecture, security, and privacy, resource limits, and a lack of sufficient training data. Blockchain, as opposed to that, promotes a decentralized architecture as new technology. It encourages the secure sharing of data, and resources among the various nodes of the IoT network, removing centralized control and overcoming ML's current difficulties. As a result, this study provides a smart city intrusion detection system. This system consists of three modules: a trust module based on designing an address-based blockchain reputation system, a two-level privacy module based on blockchain-based enhanced Proof of Work technique, and an intrusion detection module. We provide a blockchain-IPFS integrated Edge-Fog-Cloud infrastructure, named Cloud-Block, Fog-Block, and Edge-Block, to utilize the system proposed for smart cities, related to the inherited strengths and shortcomings of Edge-Fog-Cloud architecture. The sBoT-IoT and TON-IoT datasets are utilized to evaluate the system. Finally, a comparison of our implementation results shows that our system outperforms other state-of-the-art systems.

Some teenagers actively participate in cyberbullying, which is a pattern of online harassment of others. Many teenagers are unaware of the risks posed by cyberbullying, which can include depression, self-harm, and suicide. Because of the serious harm it can cause to a person's mental health, cyberbullying is an important problem that needs to be addressed. This research aimed to develop a technique to identify the severity of bullying using a deep learning algorithm and fuzzy logic. In this task, Twitter data (47,733 comments) from Kaggle were processed and analyzed to flag cyberbullying comments. The comments embedded by Keras were fed into a long short-term memory network, composed of four layers, for classification. After that, fuzzy logic was applied to determine the severity of the comments. Experimental results suggest that the proposed framework provides a suitable solution to detect bulling with values of 93.67%, 93.64%, 93.62% achieved for the accuracy, F1-score, and recall, respectively.

Abstract Internet of Things has many applications requiring the use of wireless communications networks. It utilizes data collection from sensor nodes connected to Wireless Sensor Networks. As such the wireless sensor networks is considered an important key for data transmission between sensor node and the gateways which are connected to the internet. Of the main concerns is the lifetime of the network which is affected by the battery power of the sensor nodes. It is noteworthy that transmission energy that dominates overall energy consumption is proportional to the distance between the transmitter and receiver. Thus, there is a need to send data from source node to a destination node in the most efficient way when it comes to battery level. Although there are many algorithms that tried to perform energy efficient routing, we will propose an intelligent algorithm to further improve this routing problem. In this paper we proposed an AI algorithm to enhance the lifetime of wireless sensor networks. We show that our algorithm improves the lifetime of the network by up to 75%, depending on the traffic rate, over existing algorithms.

The Internet of Things (IoT) is receiving increasing attention from academia and industry. However, improving the security of the IoT environment is critical for fostering trust in it and contributing to its growth in the manufacturing market. This study comparatively analyzes current methods for detecting intruders and malicious activities in IoT networks by introducing tree-based machine learning algorithms. It presents a research gap analysis of the current literature. Furthermore, an empirical evaluation study is presented to explore the potential of tree-based approaches to detect intruders in IoT networks. It compares the performance of bagging and boosting techniques in botnet detection by conducting an extensive experimental benchmarking.