MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641086148 A) filed by Vallurupalli Nageswara Rao Vignana Jyothi Institute Of Engineering And on July 14, 2026, for An Nlp-Based Voice Answer Evaluation System For Automated Assessment Of Spoken Responses Using Semantic And Coherence Analysis.
Inventors include Dr. Anjusha Pimpalshende; and Dr. V. Baby.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: ABSTRACT [0029] The purpose of the Voice-Based Answer Evaluation System is to go toward the assistance of the physically disabled people who suffer while manually writing answers and uses Natural Language Processing (NLP). Traditional evaluation methods tend to be time-consuming, biased, and labor-intensive, often leading to inconsistency in grading. With this system, students can verbalize their answers; these are then put to text, analyzed, and graded against a certain set of criteria [0030] The evaluation of answers for relevance, completeness, and coherence is based on speech-to-text conversion, semantic analysis, and machine-learning evaluation techniques. Evaluation was improved through algorithms such as Latent Dirichlet Allocation (LDA) for topic extraction, T5 Transformer for answer completeness, and Rhetorical Structure Theory (RST) for logical coherence. With all findings considered, it shows that automated voice-based assessment eliminates grading biases and aesthetics, saves time, and ensures fair evaluation with much more accessibility to such students with disabilities. These together enhance the efficiency, consistency, and inclusivity of academic assessments.
Disclaimer: Curated by HT Syndication.
