MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202521059836 A) filed by Parul University Parul Institute Of Engineering; Dr Sanjay Agal; and Ms Krishna Raulji on June 23, 2025, for A System And Method For Automated Data Cleaning And Feature Engineering In Large Scale Data Science Pipelines.
Inventors include Dr Sanjay Agal; and Ms Krishna Raulji.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: A SYSTEM AND METHOD FOR AUTOMATED DATA CLEANING AND FEATURE ENGINEERING IN LARGE-SCALE DATA SCIENCE PIPELINES A comprehensive system, applicable to large-scale data science pipelines, integrates a data ingestion module for receiving heterogeneous raw data from multiple sources, a data profiling engine that performs statistical analyses and generates metadata, and a data cleaning subsystem which employs both statistical measures and machine learning algorithms for missing value imputation and outlier detection. Downstream, a feature engineering engine transforms cleaned data into predictive variables via various encoding and transformation techniques while leveraging a feature selection submodule. A workflow orchestrator manages processing tasks through a DAG- based approach, complemented by a feedback and optimization loop using AutoML to fine-tune parameters, and an API and integration layer providing REST and gRPC endpoints for external interfacing, thereby streamlining and automating data preparation in diverse computing environments.
Disclaimer: Curated by HT Syndication.
