Advances in Science and Technology Vol. 124

Abstract: This task Online Grocery Shopping is one of most recent usefulness upgrade instruments utilized broadly by all associations any place there is a need of booking of objection by means of administrator and investigation of protests which are made or are forthcoming. Our site C2C.com is an internet based grievance the board framework where the issues of the clients can be enrolled on the web and settled by the various degrees of architects. Likewise adaptability is given to the clients can without much of a stretch determination their issues by speaking with engineer over web. C2C.com is a site that goes about as an extension among client and friends wherein client straightforwardly register their grumbling to organization through web. Absence of paper developments gives objection the board activities a speed which was never imagined in manual mode. Site permits client to enroll protest and consequently timetables and prompts administrators to source grumbling to concerned departments. CRM abbreviation represents Customer Relationship Management and it is characterized as the idea which permits we to sort out the deals and contact the board subtleties which we want to maintain our business on an everyday basis. The following prime exercises are done in the CRM Manager. Make a rundown of business contacts. Advance the different items and sell the really. Record the items that we sell. Keep a welcoming relationship with the clients. Oversee and keep up with the data set. Essentially the CRM suite proposed fulfills the need of any association to have a superior relationship with the client through legitimate client criticism cycle separated from keeping up with the different client information and overseeing them.

Abstract: Cyber security is a major worry for anyone with an internet-connected gadget in today's ever-changing environment. Cyber security has become a nightmare due to numerous issues such as intrusion detection, virus categorization, spam analysis, and phishing prevention. Our paper proposes a feature image generation and augmentation method that is integrated with a static analysis of harmful code using convolutional neural networks to address these difficulties (CNN). With the use of this approach, we are able to not only reduce the risk of letting the malware executing on our host system, also have a better availability of features due to the image augmentation that is applied to the feature images. When compared to previous methods, this CNN technique uses less resources and gives a more accurate output.

Abstract: Due to rapid growth of the internet most of the people started using internet through mobile and web apps to satisfy their needs. Such as online shopping and banking. Under OWSAP top 10 vulnerabilities, sensitive data exposure is one of the common threats that is identified in recent years and phishing is found to be a key source. Sensitive data exposure is majorly occurring in the internet using various phishing techniques and phishing is found to be a key sources of data stealing. Attackers, not only targeted the financial sectors and e-commerce industries, also in the field of defense and security . To detect the phishing attacks in webpages, many software was used. Some of the method of detection the phishing is, by using the URL of the webpage and by using contents of the webpage. Still, there is no robust and accurate software solution to detect the phishing attacks. The purpose of the research is to use both URL and contents of the webpage to identify the phishing. The proposed work is to build an automated and hybrid model using Random Forest (RF) algorithm in Machine learning with the Convolutional Neural network algorithm (CNN) in Deep Learning is applied to detect and classify the phishing in URL and web page contents in an automated manner .

Abstract: Medical imaging is very important in medical diagnosis. X-rays, ultrasound images, CT scans, brain pictures, and patient mri images are examples of images that contain sensitive information. However, poor communication channels and loopholes in hospital and medical centre storage systems risk accessing these images by unauthorised individuals who utilise them for nefarious purposes other than diagnostics. Image encryption is a common strategy for enhancing the integrity of communication and storage channels for protecting medical images from unauthorized access. This task proposes a biometrics-based method of secret sharing. Instead of sharing secrets between participants as in encryption, the biometric properties of participants create a single biometric configuration. If the biometric vaults are verified for the required number of actual during the authentication process, participants must provide a valid encryption key from the configuration is disclosed.

Abstract: Cloud Computing (CC) is a platform where resources and services are huge such as platforms, infrastructure, software and much more. Cloud computing builds its entire environment on the framework based on the user's requirement. Although many interventions are implemented for the problems that are identified in cloud security systems, intrusion and security issues on various services are rising day by day. This research focuses on cloud security systems where trusted access can be guaranteed for various resources and services using deep learning techniques. Deep Learning techniques can detect the anomaly variation based on selected features to find the intruder in the service provider's environment. A Novel Recurrent Neural Network (NRNN) - Auto Encoder (AE) model with a dataset is used to identify the abnormal and behavioral variation in the network. The proposed algorithm NRNN-AE is basically identifying the uncertainty of different types of malicious theft where the auto-encoder predicts the attacks against the unexpected network security challenges along with a genetic algorithm for optimization. Attacks based on the service are identified on each hidden layer based on classification that is processed in the cloud system. The results are obtained from the comparison of NSL-KDD dataset and KDD Cup 99 dataset for monitoring the behavioral and frequent changes in patterns. The system can improve the detection rate and achieve accuracy of 96% compared to the existing RC-NN model. Also the detection rate is reduced to 0.0008 which has a precision value in both positive and negative rate as a gradual increase in performance.

Abstract: With the growth of Internet of Things (IoT), which connects billions of small, smart devices to the Internet, cyber security has become more difficult to manage. These devices are vulnerable to cyberattacks because they lack defensive measures and hardware security support. In addition, IoT gateways provide the most fundamental security mechanisms like firewall, antivirus and access control mechanism for identifying such attacks. In IoT setting, it is critical to maintain security, and protecting the network is even more critical in an IoT network. Because it works directly at local gateways, the Network Intrusion Detection System (NIDS) is one of the most significant solutions for securing IoT devices in a network. This research includes various IoT threats as well as different intrusion detection systems (IDS) methodologies for providing security in an IoT environment, with the goal of evaluating the pros and drawbacks of each methodology in order to discover future IDS implementation paths.

Abstract: Docker container technology is a new virtualization technique that is extremely efficient throughout the development and deployment phases. Although Docker container technology is more convenient than traditional virtualization technology (virtual machines); it suffers from weak security due to inexperienced Docker image auditing techniques. To protect the host computer or local Docker containers from malicious Docker containers, it is required to detect potential hazards in Docker images and identify risks when Docker container instances are running on the host computer. This paper proposes a tool to give the cumulative report of the three major open-source vulnerability scanners like Trivy, Clair, and Grype.

Abstract: Replication, in general, is defined as repeating a study's technique and assessing whether the previous finding re-occurs. Research can become replicable when a person can copy the same content and arrive at the same conclusion as the original study. In this paper, an approach control mechanism for article observation replicas is been proposed. The usage of encrypted pictures or encoded attribute plots has been shown to be successful in preventing unwanted approach to models. The approach's efficiency has only been verified in image organization models and semantic analysis models but not in article recognition models. For the first time, encoded feature plots have proved to be successful in the control of article observation replicas. A safe and efficient technique based on completely homomorphic encryption is used and its usefulness for a variety of real data is been demonstrated. The suggested technique is the first to directly replicate an algorithm on ciphertext, which is one of the best performers on the plaintext feature selection problem. Furthermore, the suggested protocol is extensible to the scenario of more than three data owners.

Abstract: Digital images play a very important role in different areas in the modern technological scenario. Changing and manipulating the content of the digital image is a very easy task by using powerful image editing tools. In today's technology environment, digital photographs serve a critical function in a variety of fields. Using advanced image editing tools, changing and rearranging the content of a digital image is a simple operation. It is now possible to add, edit, or remove essential aspects from an image despite leaving any perceptible alterations. In addition to determining if the picture is authentic or forged, the metadata of the image may be examined, however, metadata can be altered. In this example, the authors use Error Level Analysis on each picture and matching parameters for error rate analysis to detect images of modifications using Deep Learning on a dataset of a false image and real photos. This experiment shows that by running through 100 epochs, we obtain the best training accuracy of 99.17 % and 95.11 % of accuracy validating.

Abstract: A Cyber-attack is a deliberate intent to take illegal access to one’s computer and data. The ascent of the web has turned into the groundwork of the vast majority's day-to-day schedules, and online administration has raised security worries. The rising measure of information, dividing among the cloud and the clients, additionally makes an attack surface. The attack surface has likewise extended with the ascent of organizations and the rising number of individuals utilizing them. The capacity of existing discovery plans to approve the goal and the earlier acknowledgment of assaults is falling apart. In the event that no effective assurance mechanism is carried out, the web will turn out to be substantially more helpless, expanding the gamble of information spillage or hacking. The focus here is to put forward a model (IDS) that detects network intrusions or anomaly detection by classifying all the network traffic packets as non-attack (harmless) or attack (vindictive) classes and also classifying the type of malicious classes using Support Vector Machine algorithm. The machine learning algorithm Support Vector Machine works for classification as well as regression problems. Decision boundaries are usually used in Support Vector Classification (SVC). We have used two different datasets of cybersecurity, namely KDDCUP 1999 and UNSW_NB15. The proposed model has been evaluated using performance metrics, namely accuracy, precision, recall (Detection rate), and F-measure. The test results exhibit that our framework has better identification execution for various cyberattacks. This model achieves an accuracy of 99.8 percent with the KDDCUP 1999 dataset and 98.2 percent with the UNSW_NB15 dataset, and remarkable detection rates of attacks.