CAREER: Data Preparation for Trusted and Fair Data Science
Sponsor: National Science Foundation
This material is based upon work supported by the National Science Foundation under Grant No: IIS-2237149
Principal Investigator: Romila Pradhan
Project Summary: Machine learning systems are increasingly being used in a variety of data science applications that involve automated decision-making. While learning-enabled systems have the potential to eliminate some undesirable aspects of human decision-making, they are known to reinforce systemic biases and discrimination reflected in their training data. This project will develop novel technologies to realize the potential of robust, fair, and explainable data-driven decision-making systems. Toward this goal, the project centers on data preparation and debugging techniques to ensure that the underlying training data and data handling processes are devoid of unexpected errors. Comprehensive and efficient solutions that demonstrate the importance of data quality as a tool for understanding and debugging undesired behavior of data science applications will be developed.
Project Publications:
- PipeLens: Identifying Interventions for Resolving Malfunctioning Data Science Pipelines
Jahid Hasan, Stanley Jiang, Tejendra Singh, Sainyam Galhotra, Romila Pradhan, and Divesh Srivastava. In Proceedings of the 52nd VLDB Conference, Boston, USA, 2026. - Scalable Goal-Oriented Source Selection
Ambarish Singh, and Romila Pradhan. In Proceedings of the 15th International Workshop on Quality in Databases (QDB) at the 52nd VLDB Conference, Boston, USA, 2026. - Selective Data Expansion for Model Performance
Jahid Hasan, and Romila Pradhan. In Proceedings of the 29th International Conference on Extending Database Technology (EDBT), Tampere, Finland, 2026. - SPLICE: An Efficient Framework for Selecting Sources for Machine Learning Tasks
Ambarish Singh, and Romila Pradhan. In Proceedings of the 42nd IEEE International Conference on Data Engineering (ICDE) (Lightning Talk), Montreal, Canada 2026. - Evaluating Pipeline Robustness to Distributional Shifts
Jahid Hasan, Jingya Wang, and Romila Pradhan. In Proceedings of the 42nd IEEE International Conference on Data Engineering (ICDE) (Lightning Talk), Montreal, Canada 2026. - SENTINEL: Evaluating Pipeline Robustness to Distributional Shifts
Jahid Hasan, Jingya Wang, and Romila Pradhan. In Proceedings of the 42nd IEEE International Conference on Data Engineering (ICDE) (Demonstration), Montreal, Canada 2026. - Label Flipping for Group Fairness
Shashank Thandri, and Romila Pradhan. In Proceedings of the 14th International Workshop on Quality in Databases (QDB) at the 51st VLDB Conference, London, UK, 2025. - Explanations for Machine Learning Pipelines under Data Drift
Jahid Hasan, and Romila Pradhan. In Proceedings of the Human-in-the-Loop Data Analytics (HILDA) Workshop at the 2025 ACM International Conference on Management of Data (SIGMOD), Berlin, Germany, 2025. - Explaining Fairness Violations using Machine Unlearning
Tanmay Surve, and Romila Pradhan. In Proceedings of the 28th International Conference on Extending Database Technology (EDBT), Barcelona, Spain, 2025. - Valuation-based Data Acquisition for Machine Learning Fairness
Ekta, and Romila Pradhan. In Proceedings of the 13th International Workshop on Quality in Databases (QDB) at the 50th VLDB Conference, Guangzhou, China, 2024.
Disclaimer: Any opinions, findings, and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).