一种基于大语言模型和多路召回的多源数据对齐方法、系统、设备及介质
By constructing keyword and semantic vector indexes using a large language model and multi-way recall strategy, and combining multi-way recall and decision rules, the problem of automated entity alignment of heterogeneous data sources is solved, achieving efficient and accurate data alignment, reducing labor costs and improving data analysis efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING LE MA SHI INTELLIGENT TECH CO LTD
- Filing Date
- 2026-04-23
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies struggle to automatically and accurately identify and align data records pointing to the same real-world entity from heterogeneous data sources without relying on pre-defined unified identifiers or manual mapping rules. This results in severe data silos, impacting the efficiency of data analysis and intelligent decision-making.
By employing a large language model and a multi-path recall strategy, and by constructing keyword indexes and semantic vector indexes, combined with multi-path recall and decision rules, we achieve automated multi-source data alignment.
It achieves high-precision and high-recall entity alignment, reduces labor costs, improves the automation level of data governance and the credibility of alignment results, and adapts to complex and heterogeneous data scenarios.
Smart Images
Figure CN122064737B_ABST