In real-world applications, particularly in platforms reliant on user-generated content like review websites and e-commerce platforms, evaluating the quality of webpage content is essential to ensure that users access reliable, relevant, and up-to-date information. Traditional classification methods face difficulties in determining the quality of a webpage, especially when dealing with new, unseen reviews that exhibit diverse patterns and contexts. To address these challenges, a novel Residual Convolutional Neural Networks and Drop Connect Long Short-Term Memory with Deep Deterministic Policy Gradient (RCNN-DCLSTM-DDPG) reinforcement learning is proposed. This model utilizes a Residual Convolutional Neural Networks (RCNN) for feature extraction, the Dilated CNN (D-CNN) captures long-range dependencies within review text by expanding the receptive field without increasing computational complexity. Additionally, the Residual Network (ResNet) architecture incorporates threshold-weighted mapping, enabling the network to focus on the most relevant features while eliminating unnecessary layers, which enhances classification accuracy. The Drop connect regularization is applied to the LSTM, randomly removing connections during training to reduce overfitting, enhancing the model's robustness. In the Deep Deterministic Policy Gradient (DDPG) framework, the actor network utilizes these feature representations from the RCNN-DCLSTM to predict the quality classifications of reviews, such as very high quality, high quality, moderate quality, low quality, or very low quality. The critic network incorporates the output from the RCNN-DCLSTM classifier to evaluate the accuracy of the actor's decisions, providing feedback in the form of a reward signal based on the accuracy of the predictions. This enables the system to adjust its parameters based on the feedback, improving the quality classification process. The proposed RCNN-DCLSTM-DDPG is tested on four datasets and compared to previous methods. Experimental results for dataset-4 show that the proposed technique achieves an accuracy of 99.3

One of the prevailing areas of contemporary research involves the differentiation and identification of diverse objects within a given scene through automated systems. The field of study under consideration presents a multitude of obstacles, including but not limited to issues such as diminished lighting conditions, occlusion, and camouflage. The captured image exhibits variations in illumination, resulting in uneven brightness, reduced contrast, and the presence of noise. The fundamental basis of computer vision algorithms lies in the process of extracting features from datasets and subsequently discerning these features through neural networks. The task of extracting distinct feature key points from images captured under low lighting conditions is exceedingly challenging. To address this issue, the present study seeks to employ deep learning models to implement image enhancement techniques specifically designed for low-light conditions. The primary emphasis lies in obtaining key feature points that are differentiable, thereby enabling the utilization of this annotated data for specific tasks such as object detection. The task of identifying occluded and camouflaged objects has been successfully accomplished, yielding an impressive accuracy rate of 93% in total. The mean average precision has been achieved as 85% which is reasonably high compared to many earlier works.

Horizontal scalability is very crucial in cloud applications. For microservices applications that are installed in AWS cloud environment, auto-scaling feature will automatically scale applications based on the configurations. Amazon ECS can calculate service performance as per CPU and memory resources consumed at a point in time and it provides data to CloudWatch metrics, such as ECSServiceAverageCPUUtilization and ECSServiceAverageMemoryUtilization. This paper discusses how auto-scaling can be achieved using the metrics and applying the scaling policies proportionally. Details of how the metrics collection strategies can be used for Application Auto-Scaling to scale services installed in the AWS cloud environment are discussed. It is observed that auto-scaling not only supports scaling up the instances in peak hours but also scales down the instances when there is minimal or no load on the application. Auto-scaling keeps monitoring the instance metadata. It helps in identifying the health status of the instances. It is observed that in peak load situations, if there is a demand for instant user requests or there is a surprise increase in user transactions increase in user requests, auto-scaling automatically adds more resources to handle the situation, which makes the system fault tolerant.

Due to the exponential growth of cloud data and network services, computational resources and cloud data security have one of the most attractive research issues in the real-time cloud environment. Numerous types of cloud services are combined into various domain applications, including, e-health, defence, clinical databases, and so on, for data storage and resource computing. However, a traditional challenge to consider is that the Data Owner (DO) no longer has the ability to modify the access control policy or key policy, resulting in data sharing inflexibility, inefficiency, and key abusers. To address these concerns, we propose an enhanced d level cut-off point-Quantum Secret Sharing (EdLCp-QSS) scheme based on an efficient Monitoring Key Ciphertext-ABE (MKCABE) –access control with Blockchain and Key Controller (EQMBK) mechanism for security. The proposed EQMBK mechanism provides security for monitoring the data from getting accessed by un-authorized users as well as offers updating security when a new user joins the cloud environment. The proposed EQMBK system ensures that data is not accessible by unauthorised users and that security is updated when a new user joins the cloud environment. Finally, experimental simulation reveals that the suggested technique is very efficient in preserving the privacy of the user’s data against unauthorised parties in terms of computation time encryption time, system initialization time, key generation time, and decryption time. Furthermore, when compared to state-of-the-art techniques, the proposed methods are more resilient and provide greater performance.

The Agile technique, which has a few guiding standards, is used by nearly all IT businesses. With dexterous, organizations can use an iterative approach to provide value to their clients more quickly and with fewer problems or escalations in both the environment of extend administration and computer program development. Customer-focused, self-organizing, and cross-functional teams that collaborate to identify needs and make progress arrangements are key skills in program improvement. Spry organizations place a strong emphasis on “collaboration,” which is essential to the development of their code. The problem that arises from this is “Following of Numerous Adaptations of Code Composed by Numerous Engineers at Distinctive Times.” Code changes must be planned for in order for the IT sector to successfully provide a final customer-centric item. IT works in this manner. The problem with centralized VCS is that it is centralized in nature and always depends upon developers to commit their changes on a central server. In the initial stages, it looks easy to set up and configure the one goes setup for centralized systems. In this work, various practical problems, from DevOps perspective, on using a centralized version control system are analyzed, and further approaches for migrating from a centralized version control system to a decentralized version control system are explored.

Species dispersal from one territorial zone to another is a complex process. The reasons for species dispersal are determined by both natural and human factors. The purpose of this study is to develop a cost surface for a hypothetical landscape that accounts for various species dispersion features. With tigers (Panthera tigris tigris) as the focal species, a computational model for a landscape has been proposed to predict the dispersion patterns of the species’ individuals from one habitat patch to another. Knowing how tigers disperse is very crucial because it improves the likelihood of successful conservation. The likelihood is raised because it strengthens conservation efforts in the targeted regions identified by the proposed model and encourages landscape continuity for tiger dispersal. Initially, four major factors influencing tiger dispersal are explored. Following that, grids are overlaid over the tiger-carrying landscape map. Further, game theory assigns a score to each grid in the landscape matrix based on the landscape features in the focal landscape. Specific predefined ratings are also utilized for scenarios that are very complex and may change depending on variables, such as the interaction of the dispersing tiger with co-predators. The two scores mentioned above are combined to create a cost matrix that is shown across a landscape complex to estimate the impact of each landscape component on tiger dispersal. This approach helps wildlife managers develop conservation plans by recognizing important characteristics in the landscape. The results of the model described in this work might be beneficial for a wide range of wildlife management activities, such as corridor management, smart patrols, and so on. A cost surface over any focal landscape may serve as a basis for policy and purpose design based on current landscape conditions.

For enterprises of all sizes and types in the IT sector, data security is essential. Data security is the process of giving databases and websites secure data privacy protections and preventing unauthorized data access. By encrypting digital data, software/hardware, and hard drives, unauthorized individuals or hackers can only access unreadable material, which is why encryption is a crucial data protection strategy. Data is produced by computing operations, which are then used to encrypt and send the data into the network. For low bandwidth channels, an encryption strategy has been put out in this work. The proposed method uses the properties of homomorphic transform to make it more unique and suitable for text encryption. Conversion of plain text to cipher text is done by using simple union and symmetric difference operators. Which makes it more reliable in terms of execution time.

In recent times, human intervention has been significantly reduced with the help of Internet-of-Things devices. A better monitoring control and faster response is received with the help of this technology. Human machine interaction is increasing continuously and Internet-of-Things is a step ahead by connecting the smart objects with the internet and therefore making overall working easier. In this work, a cost effective, energy efficient and flexible design of home automation is proposed which connects and controls various home appliances using NodeMCU, Google assistance and Blynk app. It provides a helping hand for the old aged and for differently abled persons. Human intervention is significantly reduced and in case of emergencies it has detecting capabilities which make it more reliable. User friendliness is provided through Google Assistant using voice commands. Blynk application helps to connect through an app and by using IFTTT, user can create a customised command for each job for Google Assistant. Proposed design also take care of the gas, temperature and humidity of the house using DHT11; the temperature and humidity sensor and MQ02; which is a gas sensor. Values of these sensors are displayed in the application which is connected to the relay. NodeMCU has been used for connecting Blynk app. Blynk server has been configured with NodeMCU using authentication key and further Blynk server is connected to the webhooks server which acts as a broker. For connecting Google Assistant, IFTTT has been used using IF-THEN conditions.

Nowadays, internet has numerous of web contents but it is difficult to find the web page quality. For predicting the quality of web page, a technique is necessary. Therefore, for determining the quality of the web page, a novel Residual Convolutional Neural network and Drop Connect Long Short Term Memory (RCNN-DCLSTM) technique is proposed. It consists of two stages: pre-processing and classification. In the preprocessing stage, tokenization, identification of slang, stop word removal, and lemmatization processes improve the level of accuracy during classification. In the classification stage, the proposed deep learning classifier based on RCNN-DCLSTM is used to classify the quality of web page as very high quality, high quality, moderate quality, low quality, and very low quality based on reviews. Here, the Drop connect regulation system on hidden-to-hidden weight metrics with LSTM is used to avoid the fitting problem. The proposed RCNN-DCLSTM accuracy is tested on four data sets and compared with previous methods. Based on the estimation result, the proposed RCNN-DCLSTM gives the accuracy of 0.91, precision of 0.909, recall of 0.908, and F-1 measure of 0.91. Hence, it is proved that the proposed RCNN-DCLSTM technique accurately finds the quality of web page.

In today’s era, traffic congestion is the widest spread problem observed all over the world, arising as consequence of exponential rise in vehicle count at the traffic intersections. This growth has largely affected the people as they are experiencing enhanced delay in travelling time and increased fuel consumption which led to wastage of billions of dollars. The current road infrastructure design and traffic signal controlling using a cycle of fixed time phase of green/red/yellow lights are not adequate to tackle the rising demands of traffic in an optimum way. These traditional traffic signal systems cannot handle the dynamics of road traffic at the intersections and hence results in exceeding delays. Also, the volume of traffic at any intersection at different times of the day is uncertain and hence it is hard to get an exact mathematical model for this problem. Many researchers have proposed some solution to this problem and their work is reviewed extensively in this paper. Due to its ability to deal with uncertainty, fuzzy logic is considered as the most appropriate technique to solve this problem and is highly recommended method for implementing automated traffic controllers. Due to its inherent advantages, most of the research in the field of traffic engineering is carried out using fuzzy logic techniques. Hence, this paper presents a systematic review of various techniques that are used for an effective management of traffic, especially focusing on different fuzzy based traffic controllers and their performance comparison to identify the best input output parameter.

Dynamic traffic control is a challenging task that involves meeting rising traffic demands and cutting down on intersection delays. The existing yellow/red/green light fixed transition periods used by traffic controllers make it impossible for them to adapt to changing real-time traffic conditions at intersections. Furthermore, it would be impractical to hire traffic officers for every intersection throughout the day due to a lack of personnel, and even if sufficient personnel are available, it would be a very expensive set up. A fuzzy based traffic model was designed and simulated in real time conditions using the developed traffic simulator algorithm to control the traffic jamming at road intersections. The developed fuzzy model was based on three fuzzy inputs and its performance was measured for 13 cases of varying road width. The developed model outperformed the traditional fixed-time delay model in all the cases and the level of improvement was further increased when the congestion was high. Narrower roads were more congested and the improvement with fuzzy systems as compared to its fixed time delay counterparts was as high as 26%. This research findings clearly support the use of fuzzy logic for handling the most challenging problem of traffic congestion in densely populated regions.

One of the critical issues in detecting depression is using facial expressions with image data classification. In This research paper, we proposed Fusion Fuzzy Logic((FFL) with deep learning for identifying depressed people based on their facial expressions. Our proposed model the based on an advanced fuzzy algorithm with deep learning for unordered fuzzy rule(FR) initiation to offer appropriate and suitable opinions based on depressed people's facial expressions(FE), to allow Depression Recognition(DR) from image files and recorded video files. The primary goal of this research work was to use the fusion method to turn these facial expressions (FE) into the detection of depressed states. To elevate the performance of the Fusion Fuzzy Logic((FFL) (fuzzy logic and CNN)), delivering them entreated them several times to imitate specific facial expressions. Our proposed FFL with the CNN model produces exact and dependable results with a 94.3% overall accuracy comparable to human recognition.

Smart Meters and Smart Grids are the present and future of power, water, and gas distribution systems around the world. The Supervisory Control and Data Acquisition (SCADA) system, which is an automated remote command and control system, form the backbone of these smart grids. This makes smart meters and SCADA systems highly lucrative target for cyber-attacks for any organisation/country. For the communication in a smart grid to be secure, it must provide confidentiality, integrity, and availability of the data from end-to-end. Various defence mechanisms for a few types of cyber-attacks on a SCADA system are proposed. For the defenders, it becomes extremely important to plug the vulnerabilities and make the Smart Grids more robust to counter such attacks. Study of vulnerabilities of SCADA systems has been the focus. Work has been carried out with the help of experiments and thereby brought out a few vulnerabilities which are critical to the security of SCADA systems. Using an experimental smart grid various cyber-attacks have been demonstrated successfully which includes Man-in-the-Middle and Denial-of-Service attacks. Possible counter-measures are proposed for protection against such types of attacks.

From the perspective of the Industry 4.0 paradigm, the machine learning (ML) discipline has had a significant influence on the manufacturing sector. The industry 4.0 concept promotes intelligent sensors, gadgets, and equipment to create technology infrastructure sectors that collect information constantly. By analyzing the obtained data, machine learning approaches allow actionable insight to boost industrial productivity without dramatically altering the necessary resources. Furthermore, the capacity of machine learning applications to provide actionable analytics has facilitated the detection of complex manufacturing trends and paved the path for an integrated intelligent process in various activities in the supply chain, including smart and constant inspection, preventative maintenance, quality enhancement, process optimization, supply chain advancement, and workflow scheduling. This paper aims to present recent advances in the field of quality inspection in Industry 4.0 and develop a framework for quality inspection that can be fully utilized in the Industry 4.0 context using adaptive bilateral filtering and Feature Correlated Auto encoder (FCA) machine learning technique. The suggested approach makes full use of information from all sources along the manufacturing chain. Therefore, it complies with quality management standards within the context of Industry 4.0. The suggested model makes use of corrective measures based on data patterns discovered through predictive analysis. Result analysis was shown on some pre-trained deep learning models such as ResNet18, Vgg19, Alexnet, Squeezenet, auto encoder, and FCA and observed that the proposed FCA(Feature Correlated Auto encoder) achieved a better result.

Data security is critical for businesses of all sizes and types in the IT industry. Data security refers to the provision of secure data privacy protections to databases and websites, as well as the prevention of unwanted data access. Encryption is a critical data security approach in which unauthorized people or hackers gain access to unreadable material by encrypting digital data, software/hardware, and hard drives. Computing processes generate data, which is then encrypted and sent into the network using the key. In this work, an encryption approach has been proposed for low bandwidth channels. The proposed approach is compared with various other approaches like DES, 3DES, AES, blowfish, and RSA in terms of execution time and file size. It is observed that the proposed approach has a faster execution time than existing algorithms for data sizes less than 1 MB. However, once the data size exceeds 1 MB, the proposed approach takes longer to execute than previous algorithms. This implementation focuses on shortening the encrypted message in this proposed technique, increasing the complexity of the attacker's decryption. The key is generated by the message and sent together with the encrypted message. The key is the same length as the cipher text, and the cipher text is then reduced to the original message. Finally, the encrypted text and key are the same size as the original message.

Due to the massive amount of information accessible on the internet, it has become a challenging task for users to discover the desired information. Automatic document summarization has become an emerging technology to address these issues. This allows the users to get the relevant information in a shortened version. However, the summary should have high content coverage and low redundancy to generate a good quality summary. Therefore, an enriched Dragonfly-Fuzzy Logic (FL) Single Document summarization is presented in this paper. Initially, the web document is preprocessed in which some functions such as segmentation, stop word removal, URL removal, stemming, etc. are performed. After preprocessing, the important features such as sentence location, proper nouns, numeric data, cue phrases, etc. are extracted from the web document. Here, the significance of the extracted features is decided by providing weights to each of the features using the Enriched Dragonfly Optimization Algorithm (EDOA). Once the weights are allotted for the features, the importance of the sentence is determined by using the FL system to form a summary. Finally, the sentence similarity in the generated summary is calculated, and then the similar sentences are eliminated from the summary to avoid redundancy issues. The performance of the proposed Dragonfly-FL summarization is tested in the CNN/Daily Mail dataset, and finally, the results are compared with the existing techniques such as MAMHOA, ExDoS, Karci summarization, and regression-based technique, DSN, Semantic approach, and BERTSUMEXT in terms of ROUGE-1, ROUGE-2, and ROUGE-L measures. The observation demonstrates that the proposed technique performs better than the existing techniques with precision, recall, and an F-score of 0.11, 0.05, and 0.01 respectively.

In every country, there are a plethora of laws whose very foundation stands on the age of the concerned. Similarly, successful gender recognition is essential and critical for many applications in the commercial domains, like human–computer interactions: such as computer-aided physiological or psychological analysis of a human. In this work, a face and gesture detection and verification system are proposed that classifies them on the basis of gender while providing the most probable age range of the concerned face and also detecting the gesture of the hands using convolutional neural network architecture. The principal idea behind the system is to compare the image with the reference images stored as templates in the database and to determine the age and gender.

Most organizations today rely heavily on their data warehouse to make enterprise level decisions. Data Warehouse pulls data from various heterogeneous sources and thus, when setting up a data warehouse, there are three ways to process data: ELT (Extract, Load and Transform), ETL (Extract, Transform and Load) and reverse ETL. It can be challenging to select the best approach when deciding how to implement a data warehouse because it has to do with costs, procedures, performance, and ongoing company improvement. In this paper, we'll be discussing the three approaches and their use cases.

With the availability of various economical sensors and the implementation of Internet-of-Things (IoT), agriculture industry is moving towards more precise, data centric and smarter than ever. Almost every industrial domain has been resigned, including smart agriculture, with the rapid emergence of IoT based technologies with the aid of economical sensor technology. The use of agricultural robots in agriculture field is increasing to cover the necessity ever increasing population with static piece of land. In this work, a multi-utility agricultural IoT based robot has been proposed for performing various agricultural activities. It provides control of the agricultural activities remotely through VLAN and also through cloud network (be it Server based or Serverless), i.e. a hybrid control model. The proposed method requires minimal human labor to perform various farming activities. The end user can control the machine in the field and all agricultural activities can be performed remotely from anywhere with the help of Android application. Three modes of operation are supported, manual, semi-automatic and automated. The proposed system is to be controlled and trained for few days manually so that it gains accuracy in the path being followed to perform the task. Proposed system has solar powered system as the primary source of energy and Lithium-ion polymer battery for battery backup support. Anti-theft mechanism is provided with the help of GSM module.