MUMBAI, India, Feb. 13 -- Intellectual Property India has published a patent application (202641007487 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Jan. 26, for 'system for repairing ambiguous user queries using counterfactual embedding ranking.'
Inventor(s) include Akash K; S. Durga Prasad; Rampriya R S; and Hariharan J.
The application for the patent was published on Feb. 13, under issue no. 07/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a system (100) for repairing incomplete or ambiguous natural language queries. The system (100) comprises an input processing layer (104) configured to extract semantic features from a user query (102), a missing-context detection layer (106) configured to classify the query as complete or incomplete, and a counterfactual context generation layer (112) configured to generate multiple alternative query interpretations. A context window manager (108) maintains short-term conversational context. An embedding similarity ranking layer (110) computes similarity scores between embeddings of the query, context, and counterfactual interpretations to rank the interpretations. A query reconstruction layer (114) converts a selected counterfactual interpretation into a coherent and contextually complete query. The invention provides an integrated technical framework for automated query repair based on counterfactual generation and embedding-based ranking."
Disclaimer: Curated by HT Syndication.
