Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM69056
Volume:
10 |
11 |
Nov 2026
Pages:
1-5
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Department of Data Science
Aditya College of Engineering and Technology

Department of Data Science
Aditya College of Engineering and Technology

Department of Data Science
Aditya College of Engineering and Technology

Department of Data Science
Aditya College of Engineering and Technology
Department of Data Science
Aditya College of Engineering and Technology
The use of data in research, machine learning, and other applications has increased the need for datasets that can be shared and analysed safely. However, real-world datasets often contain sensitive information, which makes sharing them a privacy concern. Synthetic data provides an alternative by creating artificial records based on patterns found in the original data. This paper presents a web-based application for generating and evaluating synthetic tabular data. Users can upload CSV files, examine and preprocess their datasets, and generate synthetic records using three techniques: Conditional Tabular Generative Adversarial Networks (CTGAN), Variational Autoencoders (TVAE), and Gaussian Copula. The generated data is evaluated using SDMetrics, which examines column distributions and relationships between columns to assess statistical quality. The application also uses the NewRowSynthesis metric to identify exact matches between original and synthetic records as part of its privacy-related assessment. The application provides visual comparisons of the original and generated datasets and allows users to download synthetic data in CSV format and evaluation reports in PDF format. The main objective is to make synthetic data generation and evaluation easier while helping users understand the balance between data quality and potential privacy risks.
KEYWORDS
This paper presents a web-based application for generating and evaluating synthetic tabular data using different data generation techniques. The application supports CTGAN, TVAE, and Gaussian Copula, allowing users to upload datasets, prepare their data, and generate artificial records through a single platform. The generated datasets are evaluated using SDMetrics to examine column distributions and relationships between different attributes. The application also uses the NewRowSynthesis metric to identify exact matches between original and synthetic records. Along with these evaluation features, users can view graphical comparisons and download the generated datasets and evaluation reports. The main contribution of this work is bringing synthetic data generation and evaluation together in an accessible web-based application. It provides a practical way to explore different generation techniques and understand how the generated data compares with the original dataset. However, generating synthetic data does not automatically guarantee privacy. The evaluation methods used in this application provide only limited information about potential disclosure risks and do not offer a formal differential privacy guarantee. Further improvements and experimentation are needed to enhance data quality, generation performance, and privacy assessment.
[1] N. Patki, R. Wedge, and K. Veeramachaneni, “The Synthetic Data Vault,” in Proceedings of the IEEE International Conference on Data Science and Advanced Analytics (DSAA), 2016, pp. 399–410, doi: 10.1109/DSAA.2016.49. [2] L. Xu, M. Skoularidou, A. Cuesta-Infante, and K. Veeramachaneni, “Modeling Tabular Data using Conditional GAN,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 32, 2019, pp. 7333–7343. [3] J. Fonseca and F. Bacao, “Tabular and latent space synthetic data generation: a literature review,” Journal of Big Data, vol. 10, article 115, 2023, doi: 10.1186/s40537-023- 00792-7. [4] J. Yoon, L. N. Drumright, and M. van der Schaar, “Anonymization Through Data Synthesis Using Generative Adversarial Networks (ADS-GAN),” IEEE Journal of Biomedical and Health Informatics, vol. 24, no. 8, pp. 2378–2388, 2020, doi: 10.1109/JBHI.2020.2980262. [5] SDMetrics, “Quality Report,” SDMetrics Documentation. [Online]. Available: https://docs.sdv.dev/sdmetrics/data-metrics/quality. Accessed: Nov. 15, 2026. [6] SDMetrics, “NewRowSynthesis,” SDMetrics Documentation. [Online]. Available: https://docs.sdv.dev/sdmetrics/data-metrics/metrics-in-beta/newrowsynthesis. Accessed: Nov. 15, 2026. [7] Z. S. Yu, “Synthetic Data Disclosure Risk Assessment: A Literature Review,” in Proceedings of Statistics Canada Symposium 2024: The Future of Official Statistics, Statistics Canada, 2025. [8] L. Pilgram et al., “A consensus privacy metrics framework for synthetic data,” Patterns, vol. 6, no. 10, article 101320, 2025, doi: 10.1016/j.patter.2025.101320.
Publication Date
11/17/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
1
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...