Search

In recent decades, computer networks have played a key role in modern life and also have escalated the number of new attacks on internet traffics to avoid malicious activities. An Intrusion Detection System (IDS) is imperative for researching firewalls, anti-viruses and intrusion (bad connection). Many researchers are striving to overcome the challenges of IDS and focus on getting better accuracy to predict automatically normal data connection and abnormal data. To resolve the above problems, many researchers are focused on traditional machine learning and deep learning algorithms to detect automatically internal and external connections of network protocol. This paper adopts various Machine Learning (ML) techniques such as Bayes Network, Random Forest, Decision Table and Nearest Neighbour. The data set KDDcup-1999, which is the most reliable data set, contains a wide range of network environments. A framework for to catch attacks is also proposed with a detection rate of more than 98%. It suggested the potential application of this framework in practice to detect intrusion and contribute to the cybersecurity field.

This paper aims to evaluate the transition from conventional SAP to a business solution called S/4HANA, and understand the attached risks and challenges. The study starts with an introduction of the SAP ERP system and S/4HANA. Then, the implementation methodologies for the transition are analysed. The research object is an organisation transitioned to S/4HANA recently. Lastly, the business offerings of S4HANA would be discussed that make life easier on implementation. This research employs a mixed method. Statistical analysis is employed to analyse the data collected from questionnaires and assess the awareness and experience during the transition to S/4HANA. In addition, a qualitative approach based on interviews is used to identify the risks and challenges of the transition. The research finds that although a transition is not as smooth as thought by many organisations as just a migration, it is a need for organisations in this highly completive digital economy.

The rise of mobile applications has helped to provide information in a broader network of products remotely. They simplify the identification of products by using their barcode or even an image of the item. This paper, therefore, aims to create an e-commerce assistant Android application that incorporates machine learning, more precisely, image classification, to assist potentially disadvantaged people on a tight budget in looking to save money. This is achieved by collecting an image dataset of essential products, training a machine learning model, and applying it to the developed application. Although the proposed solution appears to be useful and consistent with all the defined goals, the available datasets are currently insufficient, taking into account both their size and quality of individual images, which negatively affect the machine learning model and limit the potential of the solution being developed. We concluded that our final product succeeds in serving the basic functionality of the app's requirements. In the future, we will reach a wider network of users and investigate their needs, then develop these functions into the application.

This paper aims to introduce, compare and contrast various commonly used business process management (BPM) tools (suites). In the first section, this research reviews the adoption of BPM tools in the industry with some concrete examples and descriptions. The continuous development trend and the benefits of integrating BPM tools into the businesses are presented. The project will then review the academic literature on BPM that would cover BPM's definition, the development of BPM theories, and the associated tools. Three BPM tools, namely Bizagi BPM Suite, ProcessMaker and Flokzu, analyse specifically will be analysed in this paper, in detail with a real business process as an example. A method adjusted from Vuksic et al. 's (2007) guidelines for selecting discrete event simulation (DES) tools will be used to evaluate and compare the BPM tools. This paper found that among the tools evaluated, Bizagi and Flokzu have a more intuitive user interface and process maker is the least user-friendly tool. Bizagi is recommended for large and middle-sized organisations and Flokzu for smaller organisations.

Digital transformation is associated with many benefits and the commonly recognised usage is machine condition monitoring. Internet of things (IoT) technologies such as radio frequency identification and wireless sensor networks (WSN) are commonly applied in different situations, but the current study primarily focuses on WSN to establish the importance of IoT in manufacturing. This study establishes the impact of IoT in manufacturing based on the impact of WSN on machine health monitoring. Both primary and secondary sources was used whereby data was obtained from the literature data in databases and existing studies. The results indicate that interconnected devices enable communication and sending alerts regarding defects and machine malfunctions. IoT helps to rectify the machine to ensure effective performance, reduce production costs, and improve safety and mass customisation. Consequently, the main benefits of IoT include faster and more informed decisions, reduced downtime, higher product quality, predictive maintenance and greater energy efficiency.

This paper aims to explore the patterns of online interaction of users of the pretty good privacy (PGP) algorithm to identify the most important and influential users in the social network. While PGP is widely used in protecting email privacy, there are some encryption defects that can raise users' concerns about data privacy and security. It is therefore essential to identify the most influential and active users who are trusted widely, getting numerous keys in the PGP web of trust. However, it is not always known whether the user actually gained trust from others or it is one who illegally forged the keys. In order to identify the most important users in the PGP network, social network analysis (SNA) is used to analyse their online interaction conditions. Along with the most traditional centrality analysis, a less frequent used K-means clustering analysis is also conducted to obtain more precise and accurate results. The SNA results show that: 1) PGP algorithm users' online interaction patterns are rather different, which include both frequent versus isolated; 2) people with higher centrality use the PGP algorithm more frequently and may become the target peeks to seek; 3) in the analysed network, all important nodes are in the same cluster when applying K-means model to divide the community.

This study presents the influence of COVID-19 and the pandemic on individuals diagnosed with hepatocellular carcinoma and intrahepatic cholangiocarcinoma, the two most common types of primary liver cancer. The study compares the effects before and after the pandemic on these patients. Additionally, it endeavors to predict the likelihood of survival for liver cancer patients. Our research will employ various methodologies to investigate this. Exploratory data analysis techniques are utilized, including univariate analysis, correlation analysis, bivariate analysis, chi-square testing, and T-sample testing. For predictive analytics, machine learning algorithms such as Logistic Regression, Decision Trees, Classification And Regression Tree (CART), Artificial Neural Networks (ANN), K-Nearest Neighbors (KNN), and Support Vector Machines (SVMs) will be applied. For our outputs, Logistic Regression and SVMs emerged as top-performing algorithms, boasting a remarkable accuracy rate of 93%. The study reveals that COVID-19 affected all age groups similarly. However, a gender-based difference was observed, indicating that males faced a higher risk of both cancer and mortality. Furthermore, the study found that variables such as year, month, bleeding, cirrhosis, and previously known cirrhosis did not significantly influence patient survival.

Machine learning technology, particularly neural networks, provides useful tools for diagnosing diseases. This study focuses on how convolutional neural networks can be implemented to diagnose COVID-19 through the processing of x-ray images. This study demonstrates how the convolutional neural networks DenseNet201, ResNet152, VGG16, and InceptionV3 can aid healthcare providers in the diagnosis of COVID-19. The models returned accuracies of 98.73%, 97.23%, 91.25% and 98.38% respectively. The results from these experiments are compared to previous studies by evaluating F1-score, accuracy, precision and recall. Additionally, the important problems of hyperparameter tuning and data imbalance are explored and addressed. Areas for future research in this area are also suggested.

Distance education has become a widely used means in both campus and training institutions. Yet, a huge gap still exists between the market shares of online courses and traditional classroom courses. This study seeks to investigate the factors that influence students' decisions of not accepting online courses over traditional classroom courses. The researchers conducted interviews and questionnaires to survey 318 Xi'an Jiaotong-Liverpool University (XJTLU) students and nine interviewees. The main findings are as follows: 1) in general, students are more likely to select traditional courses over online courses; 2) external factors such as price, characteristics of instructors and courses, and internal factors like gender and age play an essential part in decision making; 3) incorporation between digital techniques and online courses is too insufficient to maximise the effectiveness of online courses. The researcher believed the biggest obstacle is that the training agency lacks careful consideration from the perspective of students.

The purpose of this study is to investigate the factors that influence college student satisfaction with e-learning. Authors also attempt to explore the relationship between usage intention and factors affecting e-learner satisfaction. Both qualitative and quantitative methods were adopted for this study. 224 undergraduates from Suzhou participated in the data collection process. Correlation analysis and multiple regression analysis were conducted to examine relationships between various antecedents and the usage intention of online courses. It is found that external factors such as class interaction and the quality of teachers impose a significant impact on student satisfaction. Internal factors such as the level of self-discipline and the degree of concentration also influence satisfaction. Both external factors and internal factors had positive effects on the usage intention of online courses. The findings may assist faculty in designing quality online courses to meet students' needs better and improve student satisfaction.

Scientists are developing hand gesture recognition systems to improve authentic, efficient, and effortless human–computer interactions without additional gadgets, particularly for the speech-impaired community, which relies on hand gestures as their only mode of communication. Unfortunately, the speech-impaired community has been underrepresented in the majority of human–computer interaction research, such as natural language processing and other automation fields, which makes it more difficult for them to interact with systems and people through these advanced systems. This system’s algorithm is in two phases. The first step is the Region of Interest Segmentation, based on the color space segmentation technique, with a pre-set color range that will remove pixels (hand) of the region of interest from the background (pixels not in the desired area of interest). The system’s second phase is inputting the segmented images into a Convolutional Neural Network (CNN) model for image categorization. For image training, we utilized the Python Keras package. The system proved the need for image segmentation in hand gesture recognition. The performance of the optimal model is 58 percent which is about 10 percent higher than the accuracy obtained without image segmentation.

Breakthroughs in healthcare analytics can help both the doctor and the patient. Analytics in healthcare can help spot and diagnose diseases early on. Therefore, they can also be used to improve healthcare quality and patient outcomes. Machine learning models can be used to find patterns in data and generate predictions based on these patterns. They are employed in healthcare applications for disease diagnosis, prognosis, and treatment. With the development of new algorithms and other technological innovations, these models have become more effective than ever at delivering patient treatment. The primary objective of this research is to apply different machine learning algorithms to predict the diagnosis of diabetes. Furthermore, these models are compared to determine the most effective model in this regard by evaluating their accuracy of prediction, alongside other performance metrics such as precision, recall and F1 score. Of the models investigated, Random Forest significantly outperformed the others, achieving an accuracy of 82.26%.

This paper applies social network analysis in two experiments. In the first experiment, social network analysis is conducted on student friendship networks to find relational patterns. Then, three community detection methods are used to divide the student network. The RSiena package is used to illustrate the coevolution of friendship networks with smoking and drinking behavior. In this experiment, it was determined that in the closed network, same-sex reciprocated relationships are preferred. The second experiment analyzes a weighted trust network that involves users trading with Bitcoin on the BTC-Alpha platform. Since the dealers of Bitcoin are anonymous, there is an urgent need to record every dealer’s credit history to prevent fraud and other security problems. The second experiment aims to improve security problems within the Bitcoin trust network by applying social network analysis.

In the present information age, a wide and significant variety of social media platforms have been developed and become an important part of modern life. Massive amounts of user-generated data sourced from various social networking platforms also provide new insights for businesses and governments. However, it has become difficult to extract useful information from the vast amount of information effectively. Sentiment analysis provides an automated method of analyzing sentiment, emotion and opinion in written language to address this issue. In the existing literature, a large number of scholars have worked on improving the performance of various sentiment classifiers or applying them to various domains using data from social networking platforms. This paper explores the challenges that scholars have encountered and other potential problems in studying sentiment analysis in social media. It gives insights into the goals of the sentiment analysis task, the implementation process, and the ways in which it is utilized in various application domains. It also provides a comparison of different studies and highlights several challenges related to the datasets, text languages, analysis methods and evaluation metrics. The paper contributes to the research on sentiment analysis and can help practitioners select a suitable methodology for their applications.