MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621082031 A) filed by Symbiosis International Deemed University on July 02, 2026, for An Automated Mlops Pipeline System For Sustainable Development Goal Text Classification.
Inventors include Rohan Ingle; Mayur Gaikwad; and Dr. Aniket K. Shahade.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT AN AUTOMATED MLOPS PIPELINE SYSTEM FOR SUSTAINABLE DEVELOPMENT GOAL TEXT CLASSIFICATION The present invention discloses an automated machine learning operations pipeline system (100) for classifying research paper abstracts into Sustainable Development Goal categories. The system (100) comprises a data ingestion module (110), a preprocessing module (120) that sequentially normalizes, cleans, and stems textual data, a feature engineering module (130) employing TF-IDF vectorization with configurable vocabulary constraints and n-gram generation, a training module (140) utilizing a random forest classifier with balanced class weighting, an evaluation module (150) for computing classification metrics, a configuration management module (170) for externalizing pipeline parameters, a version control module (180) with hash-based cache management for selective stage re-execution ensuring reproducibility, and a deployment module (160) for containerizing and deploying the trained model to a cloud container orchestration platform with zero downtime rolling updates. The system achieves a weighted F1-score of 0.66 across sixteen goal categories. [
Disclaimer: Curated by HT Syndication.
