Sv. Shri Bharathi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Sv. Shri Bharathi has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
Angelina Geetha
Hindustan Institute of Technology and Science
2 shared publications- Arivazhagan Natesan
SRM Institute of Science and Technology
1 shared publication Bolla Raghu Ram Reddy
Hindustan Institute of Technology and Science
1 shared publication- Cheolhee Yoon
Laboratory of Autonomous Vehicle and Blockchain, Korean National Police University, Asan, Republic of Korea
1 shared publication - Dharmalingam Sundaram Manoj Kumar
Vel Tech High Tech Dr.Rangarajan Dr.Sakunthala Engineering College
1 shared publication - Divya G
Department of Data Science and Business Systems, School of Computing, SRM Institute of Science and Technology, Kattankalathur Campus, Chennai, India
1 shared publication
Latest Publications
- Open Access
- 11 August 2025
- Published by Springer Nature in Research Square
Towards Blockchain-Based Cybersecurity Framework for Internet of Things-Enabled Smart Infrastructure Using Feature Engineering with Quantum Deep Learning Model
The fast propagation of Internet of Things (IoT) networks in smart cities have presented many advantages such as improving urban efficacy, automation, and sustainability. However, these unified methods also pose important cybersecurity tasks, such as unauthorized access, data breaches, and cyberattacks that can deal with crucial infrastructure. As smart cities trust deeply real-time data automation and exchange, certifying the integrity, security, and confidentiality of these methods is vital. Conventional security devices, like firewalls and traditional encryption models frequently drop short owing to the spread nature and resource restraints of IoT devices. To tackle these tasks, industry specialists and researchers have developed advanced cybersecurity tactics, with blockchain (BC) technology and artificial intelligence (AI)-based models. Recently, the AI-driven security solutions permit real-world anomaly detection (AD) by analyzing cyber-attack designs and mechanizing risk mitigation events. This study presents the Secure Cybersecurity Framework for Smart Infrastructure Using Feature Engineering and Quantum Deep Learning Model (SCFSI-FEQDLM) methodology. This paper aims to propose a robust cybersecurity model for protecting IoT-based smart infrastructure using BC technology. Initially, the BC technology-based IoT is applied to enhance security, transparency, and efficiency in smart infrastructure systems. Furthermore, the data standardization applies min-max normalization to convert input data into a suitable format. Moreover, the feature selection (FS) process is performed by correlation analysis model, mutual information method, and principal component analysis technique to select the most significant features from a dataset. For the classification process, the SCFSI-FEQDLM method implements quantum long short-term memory with stochastic gradient descent optimizer for improving the classification performance. The incorporatin of feature engineering process with hyperparameter tuned quantum classification model helps in the accomplishment of improved detection rate. The experimental evaluation of the SCFSI-FEQDLM model was examined on a benchmark dataset. The extensive results highlight the significant solution of the SCFSI-FEQDLM approach when compared to recent models.
- 11 August 2025
- Published by Springer Nature in Research Square
Open AccessTowards Blockchain-Based Cybersecurity Framework for Internet of Things-Enabled Smart Infrastructure Using Feature Engineering with Quantum Deep Learning Model
The fast propagation of Internet of Things (IoT) networks in smart cities have presented many advantages such as improving urban efficacy, automation, and sustainability. However, these unified methods also pose important cybersecurity tasks, such as unauthorized access, data breaches, and cyberattacks that can deal with crucial infrastructure. As smart cities trust deeply real-time data automation and exchange, certifying the integrity, security, and confidentiality of these methods is vital. Conventional security devices, like firewalls and traditional encryption models frequently drop short owing to the spread nature and resource restraints of IoT devices. To tackle these tasks, industry specialists and researchers have developed advanced cybersecurity tactics, with blockchain (BC) technology and artificial intelligence (AI)-based models. Recently, the AI-driven security solutions permit real-world anomaly detection (AD) by analyzing cyber-attack designs and mechanizing risk mitigation events. This study presents the Secure Cybersecurity Framework for Smart Infrastructure Using Feature Engineering and Quantum Deep Learning Model (SCFSI-FEQDLM) methodology. This paper aims to propose a robust cybersecurity model for protecting IoT-based smart infrastructure using BC technology. Initially, the BC technology-based IoT is applied to enhance security, transparency, and efficiency in smart infrastructure systems. Furthermore, the data standardization applies min-max normalization to convert input data into a suitable format. Moreover, the feature selection (FS) process is performed by correlation analysis model, mutual information method, and principal component analysis technique to select the most significant features from a dataset. For the classification process, the SCFSI-FEQDLM method implements quantum long short-term memory with stochastic gradient descent optimizer for improving the classification performance. The incorporatin of feature engineering process with hyperparameter tuned quantum classification model helps in the accomplishment of improved detection rate. The experimental evaluation of the SCFSI-FEQDLM model was examined on a benchmark dataset. The extensive results highlight the significant solution of the SCFSI-FEQDLM approach when compared to recent models.
- Open Access
- 5 October 2023
- Published by Anapub Publications in Journal of Machine and Computing
- p. 505-516
Multiple Object Detection on Surveillance Videos for Improving Accuracy Using Enhanced Faster R-CNN
Computer vision is a dynamic and rapidly evolving field within the broader domain of artificial intelligence. Within surveillance monitoring systems, one of the central tasks is object detection, which involves identifying and localizing objects of interest in video sequences to provide safety and security of the people. Detection of multiple objects is a challenging task in video sequences which interprets less accuracy and false Bounding box regression. In this paper, enhanced faster R-CNN model is proposed and trained to compute regional proposal through Convolutional layers on the different scene of the sequences in term of lighting, motion capture related to spatial analysis. These enhancements could encompass architectural improvements, novel training strategies, or the incorporation of additional data sources to improve the model's overall performance. Proposed model is experimented on pedestrian video gives an improved accuracy detection rate than single detector techniques.
- 5 October 2023
- Published by Anapub Publications in Journal of Machine and Computing
- p. 505-516
Open AccessMultiple Object Detection on Surveillance Videos for Improving Accuracy Using Enhanced Faster R-CNN
Computer vision is a dynamic and rapidly evolving field within the broader domain of artificial intelligence. Within surveillance monitoring systems, one of the central tasks is object detection, which involves identifying and localizing objects of interest in video sequences to provide safety and security of the people. Detection of multiple objects is a challenging task in video sequences which interprets less accuracy and false Bounding box regression. In this paper, enhanced faster R-CNN model is proposed and trained to compute regional proposal through Convolutional layers on the different scene of the sequences in term of lighting, motion capture related to spatial analysis. These enhancements could encompass architectural improvements, novel training strategies, or the incorporation of additional data sources to improve the model's overall performance. Proposed model is experimented on pedestrian video gives an improved accuracy detection rate than single detector techniques.
- Open Access
- 1 January 2023
- Published by River Publishers
- p. 39-42
Mental Health Condition and Sentiment Based Chatbot
In today's hectic lifestyle, over time, has arisen the need of looking after one's mental well being. However, many medical, social, financial and personal issues have turned out to be obstacles in timely detection and treatment of mental health diseases which may even cost the life of the sufferer. The presented work proposes a transformer-based mental health condition classification model and a sentiment-oriented response system to interact with the patient, giving human-like replies. Although numerous chatbots have been developed, very few have concentrated on the issues of mental health but that too with irrelevant responses. This approach aims at delivering positive chat system to the patients, using specific Reddit datasets for each category. The classification model resulted in an accuracy upto 93% and the response system has proven to be quite sensible and relevant.
- 1 January 2023
- Published by River Publishers
- p. 39-42
Open AccessMental Health Condition and Sentiment Based Chatbot
In today's hectic lifestyle, over time, has arisen the need of looking after one's mental well being. However, many medical, social, financial and personal issues have turned out to be obstacles in timely detection and treatment of mental health diseases which may even cost the life of the sufferer. The presented work proposes a transformer-based mental health condition classification model and a sentiment-oriented response system to interact with the patient, giving human-like replies. Although numerous chatbots have been developed, very few have concentrated on the issues of mental health but that too with irrelevant responses. This approach aims at delivering positive chat system to the patients, using specific Reddit datasets for each category. The classification model resulted in an accuracy upto 93% and the response system has proven to be quite sensible and relevant.
- Open Access
- 1 January 2023
- Published by River Publishers
- p. 794-797
Efficient Prediction of Infectious and Non- Infectious Diseases using Decision Tree ClassifierAlgorithm
Disease prediction system is devolved to predict the diseases. This is done from the information/symptoms provided by the user; these symptoms are considered as input to the system. The proposed system analysis helps to discover the symptoms that are provided from the users and shows a probability of the disease as an output. This process of prediction of diseases is done using a decision tree classifier. With data present on these disease, the user can predict the disease bases on systems and also book an appointment based on the result from prediction.
- 1 January 2023
- Published by River Publishers
- p. 794-797
Open AccessEfficient Prediction of Infectious and Non- Infectious Diseases using Decision Tree ClassifierAlgorithm
Disease prediction system is devolved to predict the diseases. This is done from the information/symptoms provided by the user; these symptoms are considered as input to the system. The proposed system analysis helps to discover the symptoms that are provided from the users and shows a probability of the disease as an output. This process of prediction of diseases is done using a decision tree classifier. With data present on these disease, the user can predict the disease bases on systems and also book an appointment based on the result from prediction.
- 29 July 2022
- Published by Springer Nature
- p. 284-295
A Hybrid Approach to Facial Recognition for Online Shopping Using PCA and Haar Cascade
- 29 July 2022
- Published by Springer Nature
- p. 284-295
A Hybrid Approach to Facial Recognition for Online Shopping Using PCA and Haar Cascade
- Open Access
- 1 November 2019
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 16 (2) , 882-889
Determination of news biasedness using content sentiment analysis algorithm
Nowadays, identifying news biases in the social media is one of the most fundamental problems. News bias is a complex process that comprises several dimensions to be taken into account and it is interlinked with social, political and economic problems. In general, news bias has the ability to reflect opinion of people about a topic or government policies and actions. The proposed algorithm develops a system which can detect the biasedness of news topics from different news Websites.This approach automatically collects the news contents from various online news media portals and then consolidates them for the determination of news biasedness. In the experimental study, the news topics are gathered from various Websites of U.S., U.K., and India. For training dataset 3265 news sentences were collected under various news topics from 20 different news Websites. The effectiveness of classification of algorithm is proved by the extensive experimental study. The proposed algorithm provides a method improves the determination of news biasedness, which in turn may help in providing impartial, unbiased and reliable information.
- 1 November 2019
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 16 (2) , 882-889
Open AccessDetermination of news biasedness using content sentiment analysis algorithm
Nowadays, identifying news biases in the social media is one of the most fundamental problems. News bias is a complex process that comprises several dimensions to be taken into account and it is interlinked with social, political and economic problems. In general, news bias has the ability to reflect opinion of people about a topic or government policies and actions. The proposed algorithm develops a system which can detect the biasedness of news topics from different news Websites.This approach automatically collects the news contents from various online news media portals and then consolidates them for the determination of news biasedness. In the experimental study, the news topics are gathered from various Websites of U.S., U.K., and India. For training dataset 3265 news sentences were collected under various news topics from 20 different news Websites. The effectiveness of classification of algorithm is proved by the extensive experimental study. The proposed algorithm provides a method improves the determination of news biasedness, which in turn may help in providing impartial, unbiased and reliable information.