IFIP Open Access Digital Library
Not a member yet
22614 research outputs found
Sort by
Remote and Centralized Billing System Using Cloud Computing
No full textPart 2: Data AnalyticsInternational audienceThe process of establishing, maintaining, and utilizing a billing system is highly intricate and expensive. Small-scale suppliers will therefore be unable to afford to install a complete billing system in their stores. In this study, we propose a cloud-based billing system as a solution to this issue. ReactJS may be used for front-end development, Spring Boot can be used for business application development, and MySQL can be used for back-end development and data storage while creating this cloud-based billing system. This technology stack ensures that the billing system operates securely and effectively. Any cloud computing platform, including AWS EC2, Microsoft Azure, Google Cloud Platform, or Open stack, can host this billing system. Without the need for any specialized gear, this cloud-based billing system will be available from any device, anywhere in the globe. All you need to use to access this system is an internet-connected web browser. It makes this billing program affordable for a large group of users. This allowed the store vendors to concentrate on growing their businesses instead of worrying about inventory management, billing, and invoice production
Deep Learning Ensemble for Diabetes Prediction: Integrating LSTM, DCNN, and SMOTE for Enhanced Risk Assessment
No full textPart 2: Data AnalyticsInternational audienceDiabetes is a prevalent and chronic medical condition that poses a significant public health challenge worldwide. Accurate early prediction of diabetes risk can facilitate timely interventions and improve patient outcomes. In this study, we propose a comprehensive approach to diabetes risk assessment through the integration of deep learning techniques and data augmentation strategies. Our research leverages the power of deep learning ensembles by combining Long Short-Term Memory (LSTM) networks, Deep Convolutional Neural Networks (DCNN), and Synthetic Minority Over-sampling Technique (SMOTE) for addressing class imbalance. We begin by collecting and preprocessing a dataset comprising patient information, medical histories, and, where applicable, image data. The LSTM model is tailored to capture temporal dependencies in the data, while the DCNN extracts meaningful features. By integrating SMOTE, we tackle class imbalance challenges commonly encountered in medical diagnosis tasks. Our ensemble model seamlessly integrates the predictions from these diverse models, allowing for a holistic assessment of diabetes risk. Extensive evaluation and hyperparameter tuning are performed to optimize performance and ensure the model’s robustness. Cross-validation techniques are employed to gauge the model’s generalization capabilities. The proposed approach aims to not only provide accurate predictions but also to offer insights into the contributions of individual models within the ensemble. Ultimately, our research seeks to enhance diabetes risk assessment, offering a valuable tool for healthcare practitioners in making informed decisions and potentially improving the early detection and management of diabetes. This present study represents a significant step towards more reliable and comprehensive diabetes risk assessment, leveraging the potential of deep learning and ensemble techniques to advance predictive accuracy and broaden our understanding of this critical healthcare challenge. An accuracy score of 74% suggests that the model performs reasonably well in making correct predictions overall, although it doesn’t achieve a very high level of accuracy. On the other hand, an ROC AUC value of 0.81 indicates that the model is effective at distinguishing between positive and negative instances, showing its ability to differentiate between the two classes
AI-Driven Interviewer: Enhancing Interview Experience Through Conversational AI
No full textPart 2: Data AnalyticsInternational audienceIn the realm of job interviews, real-time challenges exist for both interviewers and job seekers. Interviewers often grapple with the time-consuming nature of sifting through numerous resumes, manually screening candidates, and conducting. Repetitive initial interview stages. Additionally, they encounter challenges related to ensuring standardisation and objectivity in evaluating candidates, thus potentially leading to bias and subjectivity in the process. Conversely, job seekers commonly face issues such as lack of feedback, limited preparation opportunities, and difficulties in showcasing their skills effectively within the confines of a traditional interview format. The AI-Driven Virtual Interviewer project resolves these challenges by integrating advanced technologies. It streamlines the initial screening process for interviewers, saving time and resources by automating the extraction of relevant information from resumes and generating standardised interview questions. Moreover, the project mitigates bias by offering an objective evaluation process, ensuring every candidate is assessed against the same criteria. For job seekers, the project provides a more realistic and immersive interview experience, thereby addressing the lack of adequate preparation opportunities. It allows candidates to practise in a lifelike environment, receive immediate feedback, and consequently improve their interview skills and confidence
Harnessing Machine Learning-Based Classification Techniques to Optimize Crop and Fertilizer Recommendations
No full textPart 2: Data AnalyticsInternational audienceWith machine learning, agriculture which is essential to the world’s food security is experiencing a revolutionary change. Using methods such as XG Boost, Support Vector Machine (SVM), and Decision Tree, this research investigates three important agricultural domains: crop prediction, fertilizer management, and crop disease diagnosis. Accurate crop forecast that maximizes resource allocation with an average accuracy of 94%. Fertilizer management increases crop quality by 15%, lowers expenses by 20%, and increases output by 25% when it is based on soil parameters, historical data, and meteorological considerations. With XG Boost, a Crop Recommendation System that takes into account the specific climatic circumstances of the area provides customized guidance on crop selection and fertilizer application, guaranteeing a 99% average recommendation accuracy. Machine learning can revolutionize Indian agriculture, promote informed decision-making, and increase agricultural productivity while maintaining traditional methods. To fully realize this promise, more study on improving accessibility is essential
Human Abnormal Activity Detection Using CNN and LSTM
No full textPart 2: Data AnalyticsInternational audienceThe detection of human abnormal activities plays a crucial role in numerous applications, including security surveillance and healthcare monitoring. In this study, a novel approach is proposed for human abnormal activity detection using 3DCNN and LSTM. The proposed method leverages the spatial and temporal relationship within the video sequences to accurately identify abnormal activities. Firstly, the video frames are extracted and transformed into volumetric representations to capture spatial information. Then, a CNN model is employed to learn informative spatiotemporal features from the video volumes. The learned features are subsequently fed into an LSTM network to model temporal dependencies. This enables the system to effectively capture the dynamic nature of human activities. To facilitate timely response, an email and beep alert mechanism is designed to notify the responsible parties in case of any detected abnormal activity. This ensures that immediate actions can be taken to mitigate potential threats or provide timely assistance. The proposed approach is validated on benchmark datasets and achieves promising results, outperforming existing methods. The combination of CNN and LSTM proves to be effective in accurately capturing human activities and detecting abnormalities. The integration of email and beep alerts further enhances the practicality of the system, enabling real-time monitoring and incident response. Overall, this research contributes to the development of intelligent surveillance systems for various applications where abnormal activity detection is critical
Crop Irrigation Advisory System Using Federated Logistic Regression
No full textPart 2: Data AnalyticsInternational audienceFederated learning (FL) is a collaborative learning algorithm that builds ML models using knowledge from distributed clients at various locations. Federated learning is valuable when data is collected from heterogeneous environments, and a unified ML model needs to be built with minimal sharing of data. In this study, a federated irrigation advisory system is developed for an agricultural application that helps the farmer to water their field based on agricultural field parameters. The application of machine learning techniques in the agricultural domain is still in its infancy. Agricultural farms employ different sensors and have natural variations in soil type and environmental conditions. A federated model framework is suitable for the development of an irrigation advisory system as heterogeneous data is gathered using a variety of sensors spread across several farms. FL-based Logistic Regression is designed to predict whether to water the field or not based on the field parameters: temperature, humidity, soil moisture content, number of days since planting the crop, and crop type. Federated learning algorithms are designed with a client-server architecture. Flower, an open-source federated framework, which used to build and test the irrigation advisory application. Two clients and a server are used to build the FL-based Logistic Regression model by exchange of model parameters. Weights and bias of the logistic regression model are aggregated by the server from multiple clients. These aggregated model parameters are sent to clients to build the regression model with its own data. Experiments were conducted to understand the performance of FL in learning the model parameters for prediction. The model’s performance is evaluated using metrics to understand the converge patterns and parameters that affect FL
Factor Analog Reasoning Model and Its Solution Research
No full textPart 2: Causal ReasoningInternational audienceAnalogical reasoning is one of the most common forms of thinking that people use existing knowledge to reason, and it is a key phenomenon of human intelligence. However, when analogical reasoning is faced with simple reasoning tasks, the result is not accurate. Computationally complex problems arise when faced with complex inference tasks in big data. In order to solve the problem of analogical reasoning, based on the [U, I] image matching principle in factor space theory, this paper proposes a factor analogical reasoning model, gives the steps of factor analogical reasoning algorithm, analyzes an example in UCI data set, and compares and analyzes the factor analogical reasoning algorithm proposed in this paper with the factor analysis algorithm. The results show that the factor analogy reasoning algorithm proposed in this paper can realize the effective reasoning of the analogy reasoning problem, and the algorithm has the advantages of accurate calculation results and short calculation time. The conclusion of analogical reasoning based on factor space expands the theory and application of factor space
Music Recommendation System Based on Facial Expression using CNN
No full textPart 2: Applications of AI/ML in Image ProcessingInternational audienceThis music recommendation system focuses on the creation of an innovative music recommendation system designed to suggest songs tailored to users’ facial emotions and preferences derived from their online behavior. Extensive user details are scrutinized and summarized to comprehend individual musical tastes. Real-time facial emotion detection, coupled with social web content summarization, aids in the identification of singer names. The system recommends songs based on detected emotions and identified singer names, offering suggestions for popular songs and playlists. An insightful observation reveals users’ inclination toward happy songs during moments of joy, prompting a unique approach to address a recognized issue. Rather than exacerbating low moods with sad songs, the system recommends motivational or happy songs, showcasing its commitment to uplifting users. User studies affirm the precision and utility of the project, underscoring the overarching goal to provide a music recommendation system that not only caters to users’ musical preferences but also enhances their emotional well-being. The integration of continuous facial emotion detection and website summarization into the recommendation process adds a novel and dynamic facet to the system's capabilities
Improved Evaluator for Subjective Answers Using Natural Language Processing
No full textPart 1: Applications of AI/ML in Natural Language ProcessingInternational audienceExamination is a preliminary process for evaluating our knowledge. While objective examinations consist of multiple-choice questions, which are easy to evaluate using simple methods, whereas subjective examinations, which often require written responses, having a significant challenge in uniform evaluation. This type of answer evaluation is a time-consuming process to undertake. This can lead to inconsistencies and biases in grading due to the stress of examiners having to carry out the same task repeatedly. With the technological advancements in the field of Natural Language Processing, which offers a range of techniques that can be employed to evaluate text responses, we have proposed a novel solution for creating an evaluator that increases the accuracy as near to human evaluators compared to other such existing systems for evaluating subjective answers. This model works by taking inputs like the user response, expected sentences and expected keywords and works with the help of numerical vectors, natural language processing and deep learning approaches and other mathematical calculations to calculate the aggregate score for an answer
COOL: Classification of Online Offensive Language Using Machine Learning and Deep Learning
No full textPart 1: Applications of AI/ML in Natural Language ProcessingInternational audienceIn the dynamic realm of online communication, the surge in offensive language and hate speech has emerged as a critical concern. The proliferation of digital platforms has led to a distressing uptick in such behaviors, challenging the establishment of a secure and inclusive online environment. Identifying and categorizing such conduct is imperative from both an ethical standpoint and for the prevention of further harm. The training and validation dataset encompass a diverse array of online platforms, ensuring inclusivity across various communication styles and linguistic nuances. To capture the nuanced characteristics of abusive language, a range of feature extraction techniques were included, consisting of traditional NLP methods and state-of-the-art deep learning architectures. In tandem with these approaches, comprehensive experiments employing a variety of classifiers, such as logistic regression, SVM, stochastic gradient descent, decision trees, and ensemble models were conducted. In summary, this research contributes significantly to the ongoing battle against online toxicity and the promotion of more constructive online conversations. The RNN algorithm’s 99.47% accuracy rate in detecting hate speech has significant societal and platform-level ramifications. It represents a strong barrier against hate speech spreading online, creating a safer atmosphere