Intellectual Property India Publishes Patent Application for System And Method For Machine Learning-Based Analysis Of Student Competencies And Career Pathway Prediction

MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641077351 A) filed by Anurag University; Gunaganti Sravanthi; K. Rashmi; Meenakshi Simha; Kanmathareddy Likhitha; Dingari Acharya Kiran Kumar; Munimanda Premchander; Amita Mishra; and Rachakonda Sravani on June 23, 2026, for System And Method For Machine Learning-Based Analysis Of Student Competencies And Career Pathway Prediction.

Inventors include Anurag University; Gunaganti Sravanthi; K. Rashmi; Meenakshi Simha; Kanmathareddy Likhitha; Dingari Acharya Kiran Kumar; Munimanda Premchander; Amita Mishra; and Rachakonda Sravani.

The application for the patent was published on July 03, 2026, under issue no. 27/2026.

Abstract: A computer-implemented, data-driven system and method for student communication competencies evaluation and prediction of optimized career pathways in higher education, the system comprises a data acquisition layer for aggregating multidimensional student datasets including demographics, academic achievement matrices, continuous engagement metrics, and behavioral profile indicators. A multilingual preprocessing module carries out the transformation of unstructured text streams by the use of cross-lingual lemmatization and root parsing by affix-stripping methods tailored to the different languages (e.g. English Spanish, and Arabic), On top of the emoji-tagging tracking method that changes UTF-8 graphic emoticons into structured semantic text tags to keep the behavioral context. A two-stage statistical feature selection filter first passes the normalized dataset through a Correlation-based Feature Subset (CFS) evaluator and then through a Gain-Ratio split-entropy attribute evaluator to identify an optimal, non-redundant student attribute feature set. An optimized C4.5 Decision Tree inductive classification structure is used for re-mapping of the selected high-impact attribute feature set to specific professional trajectory classifications, leading to definitive model training loss and accuracy convergence at or before 40 training epochs with a dynamic prediction accuracy rate of more than 97.21%.

Disclaimer: Curated by HT Syndication.