MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641125938 A) filed by National Engineering College on October 22, 2026, for A System For Cross-Model Assessment Of Language-Proficiency.
Inventors include Dr. S. Kalaiselvi; B. Ram Balakumaran; Aathi Narayanan; and M. Pon Kathir.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A system (100) for cross-modal assessment of language proficiency comprises a presentation layer (101), an authentication and access-control module (102), an assessment management module (103), a listening assessment engine (104), a speaking assessment engine (105), a reading assessment engine (106), a writing assessment engine (107), an artificial-intelligence processing layer (108), an evaluation-routing engine (113), a cross-modal competency embedding engine (114), a reliability-weighted fusion engine (115), a cross-modal inconsistency detection engine (116), a cross-modal remediation engine (117), a multi-tenant management module (111), a unified data repository (112), a real-time feedback engine (109), a centralized analytics engine (110) and a privacy-preserving cross-tenant federated adaptation engine (118). The engines generate modality parameter vectors and reliability signals, transform the vectors into a common competency space, reliability- weight modality contributions, detect cross-modal discrepancies and selectively re-evaluate lower-reliability modalities. Tenant-scoped data is maintained while perturbed tenant-local model updates are aggregated without transferring raw learner data, thereby providing reliability-controlled assessment, targeted reprocessing and privacy-preserving model adaptation.
Disclaimer: Curated by HT Syndication.
