Browse Items (14421 total)
Sort by:
-
A Pipeline for Speech-to-Text Summarization and Question Identification for Enhanced Chatbot Interactions
The rapid advancements in natural language processing provide strong support for the new potential application of integrating Google Speech Recognition API, BART, and BERT to create a full pipeline for speech recognition, text summarization and question answering without breaking human interaction. The research aims to develop such a holistic pipeline involves integrating the Google Speech Recognition API to perform speech-to-text, BART for text summarization, and finally BERT for question answering based on both the summary and original transcript. The system was tested under various criteria such as testing accuracy, real-time processing performance, and stress tests for scalability where the findings include an average of 60% text compression with BART, an 88% accuracy in BERT-based question answering, and scores indicating high user satisfaction (4.3/5). Real-time processing latency can be able to cater to interaction within 2-3 seconds and the capacity of the system has proven without performance loss during simultaneous users. The research done can practically find applications in areas like intelligent virtual assistants, customer service automation and e-learning applications that improve accessibility and user experience. 2025 IEEE. -
A post covid machine learning approach in teaching and learning methodology to alleviate drawbacks of the e-whiteboards
Deep learning has paved the way for critical and revolutionary applications in almost every field of life in general. Ranging from engineering to healthcare, machine learning and deep learning has left its mark as the state-of-the-art technology application which holds the epitome of a reasonable high benchmarked solution. Incorporating neural network architectures into applications has become a common part of any software development process. In this paper, we perform a comparative analysis on the different transfer learning approaches in the domain of hand-written digit recognition. We use two performance measures, loss and accuracy. We later visualize the different results for the training and validation datasets and reach to a unison conclusion. This paper aims to target the drawbacks of the electronic whiteboard with simultaneous focus on the suitable model selection procedure for the digit recognition problem. 2021 Tamkang University. All Rights Reserved. -
A Potential Review on Self-healing Material Bacterial Concrete Methods and Its Benefits
Building plays an important role for survival of human being in a safe place to live and store basic requirements. The building can be constructed for any purpose and the architecture of each building (official, residential) differs according to the plan. Beyond the plan for a building, it is also significant in designing plans for the construction of bridges, dams, canals, etc. In all the construction, the key goal is the strength of a building which completely depends on the materials that are chosen for each work. Hence, it is essential to prefer high quality materials for the construction of a building and the major materials are such as cement, concrete, steel, bricks, and sand. Among these materials, the concrete is often used for construction which enables to harden the building by combining cement, sand, and water. The concrete looks like a paste that reinforce to prolong life of the building. In this paper, we discuss a review on the use of bacteria in concrete that has the ability of self-healing cracks in the building. The procedural process behind the activation and reaction of bacteria into concrete is studied with the benefits of this process. This bacterial concrete usage assures to enhance the property of durability and but still it is yet to be introduced in the industries. Hereby, we review the recent research works undergone in concrete using bacteria. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
A POWERFUL ITERATIVE APPROACH for QUINTIC COMPLEX GINZBURG-LANDAU EQUATION within the FRAME of FRACTIONAL OPERATOR
The study of nonlinear phenomena associated with physical phenomena is a hot topic in the present era. The fundamental aim of this paper is to find the iterative solution for generalized quintic complex Ginzburg-Landau (GCGL) equation using fractional natural decomposition method (FNDM) within the frame of fractional calculus. We consider the projected equations by incorporating the Caputo fractional operator and investigate two examples for different initial values to present the efficiency and applicability of the FNDM. We presented the nature of the obtained results defined in three distinct cases and illustrated with the help of surfaces and contour plots for the particular value with respect to fractional order. Moreover, to present the accuracy and capture the nature of the obtained results, we present plots with different fractional order, and these plots show the essence of incorporating the fractional concept into the system exemplifying nonlinear complex phenomena. The present investigation confirms the efficiency and applicability of the considered method and fractional operators while analyzing phenomena in science and technology. 2021 The Author(s). -
A pragmatic study on heuristic algorithms for prediction and analysis of crime using social media data
Advancement in technology and Social media has grown to become one amongst the foremost powerful communication channels in human history and this is where individuals are sharing their perspectives, thoughts, suppositions, and feelings. Law enforcement units are having hard time fighting crime with evergrowing population, regional issues and political con-sequences. The adoption of social media data for crime analysis is increasing day by day. Crime analysis can help use the resources wisely. A crime prediction alerts the department at the right time to focus their staff with better equipment in suspected areas. Crime analysis prevents threats to life and money loss in terms of damage. In recent days, the collection of crime data from different heterogeneous sources becomes a primary step for the crime analysis and prediction. In this paper Overview of Heuristic Based Crime Prediction and Analysis algorithms identified by different authors. Also, various sources of social media used for analysis and prediction are also reviewed in detail. This information can be considered for one of the prominent asset for crime investigation through social media data procedure and also, we had identified the different algorithms and research gaps of that algorithms with related to crime analysis and prediction. 2019, Institute of Advanced Scientific Research, Inc. All rights reserved. -
A Pragmatic Study on Movie Recommender Systems Using Hybrid Collaborative Filtering
The Movie Recommendation System (MRS) is part of a comprehensive class of recommendation systems, which categorizes information to predict user preferences. The sum of movies is increasing tremendously day by day, and a reliable recommender system should be developed to increase the user satisfaction. Most of the approaches are made to prevent cold-start, first-rater drawbacks, and gray sheep user problems, nevertheless, in order to recommend the related items, various methods are available in the literature. Firstly, content-based method has some drawbacks like data of similar user could not be achieved, and what category of these items the user likes or dislikes are also not known. Secondly, this paper discusses about collaborative filtering to find both user and item attributes that have been considered. Since there exist some issues pictured with collaborative filtering, so this paper further aims into hybrid collaborative filtering and deep learning with KNN algorithm of ratings of top K-nearest neighbors. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
A Pre-trained YOLO-v5 model and an Image Subtraction Approach for Printed Circuit Board Defect Detection
Almost every electronic product used regularly contains printed circuit boards, which in addition to being used for business purposes are also used for security applications. Manual visual inspection of anomalies and faults in circuit boards during manufacture and usage is extremely challenging. Due to a shortage of training data and the uncertainty of new abnormalities, identifying undiscovered flaws continues to be complicated. The YOLO-v5 technique on a customized PCB dataset is used in the study to incorporate computer vision to detect six potential PCB defects. The algorithm is designed to be feasible, deliver precise findings, and operate at a considerable pace to be effective. A technique of image subtraction is also implemented to detect flaws in printed circuit boards. The structural similarity index, a perception-based method, gauges how similar non-defective and defective PCB images are to one another. 2023 IEEE. -
A Precise Computational Method for Hippocampus Segmentation from MRI of Brain to Assist Physicians in the Diagnosis of Alzheimer's Disease
Hippocampus segmentation on magnetic resonance imaging is more significant for diagnosis, treatment and analyzing of neuropsychiatric disorders. Automatic segmentation is an active research field. Previous state-of-the-art hippocampus segmentation methods train their methods on healthy or Alzheimer's disease patients from public datasets. It arises the question whether these methods are capable for recognizing the hippocampus in a different domain. Therefore, this study proposes a precise computational method for hippocampus segmentation from MRI of brain to assist physicians in the diagnosis of Alzheimer's disease (HCS-MRI-DAD-LBP). Initially, the input images are pre-processed by Trimmed mean filter for image quality enhancement. Then the pre-processed images are given to ROI detection, ROI detection utilizes Weber's law which determines the luminance factor of the image. In the region extraction process, Chan-Vese active contour model (ACM) and level sets are used (UACM). Finally, local binary pattern (LBP) is utilized to remove the erroneous pixel that maximizes the segmentation accuracy. The proposed model is implemented in MATLAB, and its performance is analyzed with performance metrics, like precision, recall, mean, variance, standard deviation and disc similarity coefficient. The proposed HCS-MRI-DAD-LBP method attains in OASIS dataset provides high disc similarity coefficient of 12.64%, 10.11% and 1.03% compared with the existing methods, like HCS-DAS-MLT, HCS-DAS-RNN and HCS-DAS-GMM and in ADNI dataset provides high precision of 20%, 9.09% and 1.05% compared with existing methods like HCS-MRI-DAD-CNN-ADNI, HCS-MRI-DAD-MCNN-ADNI and HCS-MRI-DAD-CNN-RNN-ADNI, respectively. 2022 World Scientific Publishing Europe Ltd. -
A precise method for gender cataloguing using a minimum distance classifier /
The International Journal of Engineering and Science, Vol-3 (2), pp. 1-4. ISSN (p)-2319-1805 ISSN (e)-2319-1813 -
A prediction technique for heart disease based on long short term memory recurrent neural network
In recent years, heart disease is one of the leading cause of death for both women and men. So, heart disease prediction is considered as a significant part in the clinical data analysis. Standard data mining techniques like Support Vector Machine (SVM), Naive Bayes and other machine learning techniques used in the earlier research for heart disease prediction. These methods are not sufficient for effective heart disease prediction due to insufficient test data. In this research, Bi-directional Long Short Term Memory with Conditional Random Field (BiLSTM-CRF) has been proposed to increase the efficiency of heart disease prediction. The input medical data were analyzed in a bidirectional manner for effective analysis, and CRF provided the linear relationship between the features. The BiLSTMCRF method has been tested on the Cleveland dataset to analyze the performance and compared with existing methods. The results showed that the proposed BiLSTM-CRF outperformed the existing methods in heart disease prediction. The average accuracy of the proposed BiLSTM-CRF is 90.04%, which is higher than the existing methods. 2020 by the authors. -
A Predictive Framework for Sustainable Human Resource Management Using tNPS-Driven Machine Learning Models
This study proposes a predictive framework that integrates machine learning techniques with Transactional Net Promoter Score (tNPS) data to enhance sustainable Human Resource management. A synthetically generated dataset, simulating real-world employee feedback across divisions and departments, was used to classify employee performance and engagement levels. Six machine learning models such as XGBoost, TabNet, Random Forest, Support Vector Machines, K-Nearest Neighbors, and Neural Architecture Search were applied to predict high-performing and at-risk employees. XGBoost achieved the highest accuracy and robustness across key performance metrics, including precision, recall, and F1-score. The findings demonstrate the potential of combining real-time sentiment data with predictive analytics to support proactive HR strategies. By enabling early intervention, data-driven workforce planning, and continuous performance monitoring, the proposed framework contributes to long-term employee satisfaction, talent retention, and organizational resilience, aligning with sustainable development goals in human capital management. 2025 by the authors. -
A predictive model on post-earthquake infrastructure damage
Disaster management initiatives are employed to mitigate the effects of catastrophic events such as earthquakes. However, post-disaster expenses raise concern for both the government and the insurance companies. The paper provides insights about the key factors that add to the building damage such as the structural and building usage properties. It also sheds light on the best model that can be adopted in terms of both accuracy and ethical principles such as transparency and accountability. From the performance perspective, random forest model has been suggested. From the perspective of models with ethical principles, the decision tree model has been highlighted. Thus, the paper fulfills to propose the best predictive model to accurately predict the building damage caused by earthquake for incorporation by the insurance companies or government agency to minimize the post-disaster expenses involved in such catastrophic event. The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021. -
A Predictive Modelling of Factors Influencing Job Satisfaction Through a CNN-BiGRU Algorithm
The fields of humanities, psychology, and sociology are where the word 'job satisfaction' originated. According to psychology, it is a condition in which a worker experiences his circumstances emotionally and responds by experiencing either pleasure or suffering. It is regarded as a variable in various sociological categories pertaining to how each employee assesses and thinks about his work. Because a satisfied employee contributes to and builds upon an organization's success, job satisfaction is intimately tied to an employee's performance and the quality of the work they do. As a result, job satisfaction directly correlates to an organization's success. The proposed strategy incorporates data preprocessing, feature selection, and model training. The missing value is a common feature of data preparation. Feature selection is chosen using the ANOVA F-Test Filter, the Chi-Square Filter, and the full Data Set Construction procedure. The model's efficacy can be evaluated with the help of CNN-BiGRU. The proposed technique is compared to two more models: BiGRU and CNN. It has been shown that our proposed technique outperforms two other models. 2023 IEEE. -
A primary study on the degradation of low-density polyethylene treated with select oxidizing agents and starch
Polyethylene has become an integral part of our contemporary lives. The neoteric versatile nature of polyethylene is used in constructing various applications. Out of the plastic waste discarded, 60% of the plastic waste enters landfills. The polyethylene discarded in the soil and water on exposure to the environment forms macroplastics (>2.5 cm), mesoplastics (5 mm-2.5 cm) and microplastics (<5 mm). Microplastics in the water and soil are observed to have lethal and ecotoxicological effects on aquatic and terrestrial organisms. They enter the food chain and permeate into the food that one eats. In order to address this impending concern, the present study aimed to treat plastics to form a degradable, safe and earthy material. The dissolved polyethylene was treated with starch and was made to react with oxidizing agents such as hydrogen peroxide, nitric acid and acetic acid to lower its inert ability to withstand its degradation. The effect of starch and oxidizing agents on dissolved low density polyethylene was subsequently analysed. The analysis of treated polyethylene showed a decrease in its crystallinity percentage by 6.19 and an increase in its functional groups on reaction with solvent trichloroethylene made to react with starch and oxidizing agents. In the present research, tests were conducted to obtain the various methods that can be utilized to reverse the inert ability of polyethylene. The prevailing recycling model that uses antioxidation techniques is counterproductive since it was found that such techniques appeared to make the polyethylene more resistant to further degradation. In this study, the polyethylene was dissolved in the solvents, such as xylene and trichloroethylene, to make the polyethylene more susceptible to reactants and hence a viable model for treating polyethylene. : Author (s). Publishing rights @ ANSF. -
A Privacy-Preserving Federated Learning Protocol for Secure Analytics of IoT Sensor Data Using Homomorphic Encryption
The proliferation of Internet of Things (IoT) devices has led to massive amounts of sensitive data generation, making data security a paramount concern. Existing methods often struggle with protecting heterogeneous IoT data efficiently, particularly during model training and communication. In this work, we propose a federated learning framework integrated with secure encryption mechanisms to safeguard IoT data during model training and aggregation. Each client device trains a local model using its own sensor data, encrypts the model parameters, and sends them to a server. The server aggregates the encrypted models and sends back the global model for decryption by the clients, ensuring data privacy throughout the process. The proposed framework reduces the unauthorized access risks and also the experimental results demonstrate that the model results in an accuracy of 92% during prediction tasks. The system's encryption overhead was minimal, with only a 7.5% increase in computation time compared to unencrypted federated learning methods. Future work will focus on optimizing the encryption techniques for resource-constrained IoT devices and exploring adaptive security mechanisms powered by machine learning to detect emerging threats dynamically. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026. -
A privatised approach in enhanced spam filtering techniques using TSAS over cloud networks
Major problem over cloud networks is the effect of malicious code that protrudes its own activity without intend of network user in resource sharing. One such activity is the spam-filtering techniques which assumes the data with training and testing sets and also rely on fundamental classification through distribution. A privatised spam filtering approach is a classic problem which automatically recognises user context and incoming mail information relevance. To filter mail contents learning based methods, probabilistic based method trying to improve their accuracy but they cannot attain an improvement in identifying suspicious contents and also in segregating legitimate mail entries. Here a novel representation of structured abstraction scheme (SAS) used to generate abstraction in e-mail process using HTML tag content in e-mail and its algorithm for filtering such process of spam filtering is depicted. In this SAS methodology near duplicate matching process with HTML tag ordering will be processed and newly assigned position ordering were deliberated. The experimental setup shows that there will be a great improvement while filtering spam in accuracy of e-mail content while sharing in cloud networks. Copyright 2022 Inderscience Enterprises Ltd. -
A probabilistic inference algorithm for early detection of age related macular degeneration
Age Related Macular Degeneration or ARMD is a retinal disorder that causes blindness over people of older age group. ARMD is associated with age and is a leading cause of blindness around the world. There is no specific medicine to fully cure ARMD but its development can be controlled by regular exercises and a healthy lifestyle if it is detected early. With a rising population of old age group of people, it becomes important to detect ARMD as early as possible in order to contain its development further. This research attempts to develop an algorithm based on probabilistic inference through Bayesian Network by analyzing large datasets collected from previous cases where datasets include elements of risk factors that could cause ARMD along with eye images. Unlike most of the approaches in detecting ARMD this work not only analyses eye images but also includes analysis of various factors causing the disorder. To include the study and analysis of the presence of factors causing ARMD is sensible because those factors are good indicators when the need is an early detection. 2020, Engg Journals Publications. All rights reserved. -
A Progressive UNDML Framework Model for Breast Cancer Diagnosis and Classification; [Un modelo marco progresivo UNDML para el diagntico y clasificaci del ccer de mama]
According to recent research, it is studied that the second most common cause of death for women worldwide is breast cancer. Since it can be incredibly difficult to determine the true cause of breast cancer, early diagnosis is crucial to lowering the diseases fatality rate. Early cancer detection raises the chance of survival by up to 8 %. Radiologists look for irregularities in breast images collected from mammograms, X-rays, or MRI scans. Radiologists of all levels struggle to identify features like lumps, masses, and micro-calcifications, which leads to high false-positive and false-negative rates. Recent developments in deep learning and image processing give rise to some optimism for the creation of improved applications for the early diagnosis of breast cancer. A methodological study was carried out in which a new Deep U-Net Segmentation based Convolutional Neural Network, named UNDML framework is developed for identifying and categorizing breast anomalies. This framework involves the operations of preprocessing, quality enhancement, feature extraction, segmentation, and classification. Preprocessing is carried out in this case to enhance the quality of the breast picture input. Consequently, the Deep U-net segmentation methodology is applied to accurately segment the breast image for improving the cancer detection rate. Finally, the CNN mechanism is utilized to categorize the class of breast cancer. To validate the performance of this method, an extensive simulation and comparative analysis have been performed in this work. The obtained results demonstrate that the UNDML mechanism outperforms the other models with increased tumor detection rate and accuracy. 2024; Los autores. -
A Prompt Study on Recent Advances in the Development Of Colorimetric and Fluorescent Chemosensors for Nanomolar Detection of Biologically Important Analytes
Fluorescent and colorimetric chemosensors for selective detection of various biologically important analytes have been widely applied in different areas such as biology, physiology, pharmacology, and environmental sciences. The research area based on fluorescent chemosensors has been in existence for about 150years with the development of large number of fluorescent chemosensors for selective detection of cations as metal ions, anions, reactive species, neutral molecules and different gases etc. Despite the progress made in this field, several problems and challenges still exist. The most important part of sensing is limit of detection (LOD) which is the lowest concentration that can be measured (detected) with statistical significance by means of a given analytical procedure. Although there are so many reports available for detection of millimolar to micromolar range but the development of chemosensors for the detection of analytes in nanomolar range is still a challenging task. Therefore, in our current review we have focused the history and a general overview of the development in the research of fluorescent sensors for selective detection of various analytes at nanomolar level only. The basic principles involved in the design of chemosensors for specific analytes, binding mode, photophysical properties and various directions are also covered here. Summary of physiochemical properties, mechanistic view and type of different chemosensors has been demonstrated concisely in the tabular forms. Graphical Abstract: In our current review we have focused the history and a general overview of the development in the research of fluorescent sensors for selective detection of various analytes at nanomolar level only. [Figure not available: see fulltext.] 2024, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature. -
A proof of concept implementation of a mobile based authentication scheme without password table for cloud environment
Cloud computing is a fast growing technology offering a wide range of software and infrastructure services on a pay-per-use basis. Many small and medium businesses (SMB's) have adopted this utility based Computing Model as it contributes to reduced operational and capital expenditure. Though the resource sharing feature adopted by Cloud service providers (CSP's) enables the organizations to invest less on infrastructure, it also raises concerns about the security of data stored at CSP's premises. The fact that data is prone to get accessed by the insiders or by other customers sharing the storage space is a matter of concern. Regulating access to protected resources requires reliable and secure authentication mechanism, which assures that only authorized users are provided access to the services and resources offered by CSP. This paper proposes a strong two-factor authentication mechanism using password and mobile token. The proposed model provides Single Sign-on (SSO) functionality and does not require a password table. Besides introducing the authentication scheme, the proof of concept implementation is also provided. 2015 IEEE.