The global growth of the e-commerce industry has heightened concerns among consumers, businesses, , and financial institutions regarding fraud while using credit and debit cards and safeguarding personal information. It is necessary to secure information disseminated over insecure channels to prevent unauthorized access. Cryptography and steganography are widely used for this purpose. However, completely relying on the combined usage of both may not be enough in today’s world, resulting in weak security. By combining visual cryptography, more levels of security can be added, thus boosting the security of secret information. This study suggests a data security framework that employs multiple levels of security for information. At the first level of security, cryptography is used to encrypt the secret information, and at the second level of security steganography is used to conceal the encrypted text. After that, visual cryptography is applied, which generates the shares of the image obtained. Finally, image steganography is used to hide the generated shares in different color images. When used together, these data security methods greatly improve the secrecy, trustworthiness, and effectiveness of secret messages. The precision of text is determined through Mean Square Analysis (MSE) and correlation coefficient, which involves a comparison of sent and received text. MATLAB environment is used for the implementation.

Video summarization, a critical task in managing the ever-growing volume of digital video content, has witnessed significant advancements with the integration of Natural Language Processing (NLP) techniques. In this paper, a unique technique is provided to NLP-based video summarising that makes use of textual metadata to improve the relevance and coherence of the summaries that are produced. With the proliferation of internet videos across platforms like YouTube, Instagram, etc., there is a growing need for effective summarization methods to condense diverse content. Its primary objective is to create brief and precise video summaries of YouTube content. The proposed method initially condenses YouTube video transcripts, forming the basis for generating the summarized video. Additionally, an android application is developed to facilitate user interaction. This application enables users to input a YouTube video link. Upon successful processing, the summarized video output is generated and showcased on the application. As the volume of online video content continues to surge, efficient summarization techniques become increasingly vital for users to quickly grasp essential information and navigate through the vast array of available material. This paper's methodology not only addresses this growing demand but also offers a user-friendly interface for easy access and utilization of the summarization tool. The NLP-driven video summarization system achieved high-quality summaries, demonstrating improved coherence, relevance, and informativeness compared to traditional methods.

The proliferation of social media platforms has revolutionized communication, but it has also given rise to social media bots that spread misinformation, manipulate public opinion, and compromise online discourse integrity. This study addresses the critical issue of detecting social media bots on Twitter, where traditional detection methods often fall short due to the evolving nature of these bots and the vast amount of data involved. To overcome these challenges, this research proposes a hybrid ensemble model that combines feature engineering and natural language processing techniques. The ensemble model is a meta-model that predicts the nature of a Twitter account based on the outputs of two base models. The first model uses a combination of engineered profile and content-based features, while the second model employs automatically extracted natural language processing features from posted tweets. By integrating these distinct features, the hybrid model captures a broader spectrum of bot behaviors and characteristics, leading to more accurate and robust detection. This combined approach allows the meta-model to identify bots that might evade detection when only one type of feature is used, offering a more holistic understanding of bot behavior across both structural and content dimensions. Extensive experiments demonstrate significant improvements in detection performance, achieving an impressive F1-score of 90.22% on the challenging Twibot-20 dataset, outperforming state-of-the-art models.

The Internet of Things (IoT) is transforming across the globe with its emerging applications in diverse aspects of life, namely healthcare, automated remote monitoring, smart wearables, sensing, etc. The IoT environment enriches the experience of its users by providing a platform to connect a large number of smart devices, such as smartphones, tablets, watches, etc., as well as share information worldwide. The increased popularity of IoT and smart devices has resulted in a menace as most users’ data is stored on these devices, making them a potential target for network attacks. Thus, it becomes extremely imperative to address malware threats in IoT devices. To combat this problem, the paper presents a detailed investigation to analyze the behavior of IoT malware using network forensics of six IoT botnets. We performed modeling on 55 IoT botnet samples from Twitter Honeypot. We performed botnet analysis in two dimensions: Activities and Networks. We examined botnet activities in terms of vulnerable ports, popular geolocations, protocols, and attack vectors. In terms of its topological features, severity, and packet length. To detect the botnet category, we applied six machine learning classifiers. Neural networks attained the best precision.

Fake products have become a major concern for businesses and consumers alike, resulting in significant economic losses and potential risks to consumers. Traditional methods of product authentication often fall short in the face of sophisticated counterfeit operations. Blockchain technology, with its decentralized and immutable nature, offers a promising solution to tackle this issue. This article provides a comprehensive overview of the use of blockchain for fake product identification. It begins by discussing the prevalence and impact of counterfeit products on various industries, highlighting the urgent need for an effective anti-counterfeiting system. The paper uses the fundamentals of blockchain technology, explaining its core concepts and features that make it suitable for this application. Next, the paper explores the key components of a blockchain-based system for fake product identification, including product registration, supply chain tracking, and consumer verification. It examines how blockchain can provide a tamper-proof and transparent ledger for recording and verifying product information at each stage of the supply chain. Additionally, the paper analyzes various consensus mechanisms and smart contract functionalities that enhance the security and efficiency of the system. The article also addresses the challenges and limitations associated with implementing blockchain-based anti-counterfeiting solutions, such as scalability, interoperability, and user adoption. It discusses potential strategies and emerging technologies that can mitigate these challenges and drive widespread adoption of blockchain for product authentication.

As the number of subscribers for Long Term Evolution (LTE) technology and associated resources increase, it has become mandatory for the LTE mobile operators to evaluate how efficiently the existing LTE cell network will perform in real-time scenarios. Controlling the transmitting services in the LTE cell network is an important factor as it assists in increasing the entire network execution. The best way to attain this is to attain the integrity between the users or achieve the efficient values of spectral efficiency (SE) and energy efficiency (EE). Spectral efficiency and energy efficiency are important parameters for a cellular network to exhibit its network performance level. In this paper, the real-time information records of approximately 50000 BTS sites, which were comprised of Key Performance Indicators (KPIs) such as Radio Resource Control (RRC), E-UTRAN Radio Access Bearer (ERAB), Packet drop, etc. have been considered. From the data set, throughput has been predicted using the Random Forest Algorithm. Further Spectral and Energy efficiency has been predicted. Spectral efficiency (SE) helps in controlling the transmitting power of the cell network and Energy Efficiency helps in enhancing the network quality of service (QoS). The maximum value of the spectral and energy efficiency at the uplink throughput and downlink throughput had been taken from the Nokia Drive Test sheet. If the predicted values lie beyond the limits of the network, it will indicate that the network performance has deteriorated and has scope of improvement. The resulting variables were chosen on the basis of the real field assessment by the Nokia Network Pvt. Ltd in India.