MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202621075994 A) filed by Dr. Yogita Maruti Randhavane; and Mr. Abhimanyu Bhaskar Gaikwad on June 19, 2026, for System And Method For Automated Recognition And Standardization Of Marathi Dialects Using Artificial Intelligence.
Inventors include Dr. Yogita Maruti Randhavane; and Mr. Abhimanyu Bhaskar Gaikwad.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention relates to a system and method for automated recognition and standardization of Marathi dialects using artificial intelligence. The invention is designed to receive Marathi language input in spoken or written form and convert dialectal expressions into standard Marathi while preserving the original meaning, context, and communication intent of the user. The system comprises an input acquisition module, preprocessing module, feature extraction module, dialect recognition module, contextual interpretation module, dialect-to-standard mapping module, output generation module, validation module, and adaptive learning module. The input acquisition module receives Marathi dialect input through a mobile device, web platform, chatbot, microphone, or digital communication interface. The preprocessing module performs speech-to-text conversion, noise reduction, spelling correction, tokenization, sentence segmentation, and text normalization. The dialect recognition module identifies dialectal variations by analysing vocabulary, pronunciation patterns, phonetic features, grammar structure, sentence endings, regional expressions, and contextual usage. The dialect-to-standard mapping module converts dialect-specific words, phrases, idioms, and sentence structures into corresponding standard Marathi forms using machine learning, deep learning, natural language processing, and language normalization techniques. The contextual interpretation module ensures that the standardized output maintains semantic accuracy, grammatical correctness, and natural readability. The adaptive learning module receives feedback from users or language experts and updates the dialect database and artificial intelligence models for continuous improvement. The invention supports Marathi dialect speakers in education, governance, healthcare, banking, translation, speech-to-text systems, chatbots, and digital communication platforms. Thus, the invention provides an inclusive and intelligent solution for bridging the gap between regional Marathi dialects and standard Marathi language.
Disclaimer: Curated by HT Syndication.
