MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641097528 A) filed by G. Elavel Visuvanathan; S. Murugaveni; and K. Harisudha on August 12, 2026, for Ai-Powered Natural Language Interface For Complex Database Querying.
Inventors include G. Elavel Visuvanathan; S. Murugaveni; and K. Harisudha.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT A heterogeneous database query system using natural-language query is disclosed that includes a schema-aware system powered by artificial intelligence. The natural-language understanding engine identifies intent, entities, attributes, relationships and conditions in a user's request. The intent extracted from the databases is translated into a common semantic representation using schema information from the several databases. A query planner is responsible for finding the necessary databases and creating the appropriate database operations for each query. For each query generated, the validation of the query against real schema information and access constraints are performed prior to execution. The schema context and validation feedback can be used to correct invalid queries. Validated queries are executed via adapters specific to the database, while results coming from several databases are normalized and integrated. The system then provides a reliable way of accessing the heterogeneous data sources using natural language, as well as lowering the number of invalid database operations, mismatches of schemas, unnecessary retries, and cross-database query complexity.
Disclaimer: Curated by HT Syndication.
