MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202511088884 A) filed by Primus Partners Solutions Pvt. Ltd. on September 18, 2025, for Ai Assisted Entity Resolution System With Family Clustering And Method Relating Thereto.
Inventors include Avijit Dhar; Sameer Jain; and Shaurya Mehta.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The present invention relates to an AI assisted entity resolution system and method for creating a unified, high-integrity dataset by resolving duplicates and inconsistencies across multiple heterogeneous demographic databases which at least include five data source databases (105, 120, 130, 140, 150), a Pre-Processing unit (200) including a Data Analysis Module (205), a Data Cleaning Module (210) and a Data Deduplication Module (215) employing dual parallel matching strategies (232A, 232B). An Entity Resolution unit (300) comprises a Data Selection Module (305) and a Core Identity Matching Module (310) that encodes records using a domain-specific fine-tuned Dense Vector Representation Model (626), retrieves candidate pairs via a Semantic Grouping Module (627) and approximate nearest- neighbor search, applies deterministic filters (323), and executes multi-mode text comparison (324) to produce categorized match results (325). A Post-Processing unit (400) aggregates resolved records into household groups (406) and constructs structured Family Relationship Graphs (416) encoding head-of-household, spousal, parent-child, and extended family relationships through sequential inference. A Unified Data unit (500) receives the resulting unified dataset. An AI Assistant (600) acting as Master Assistant supports all processing stages through a Natural Language to Data Processing system (605), a Data Quality Reporting Agent (615), a Data Cleaning Agent (610), a Data Profiling score Module (620), and an AI-Assisted Search Module (625). Fig. 1
Disclaimer: Curated by HT Syndication.
