对实体链接模型的实体嵌入的强化及多义词歧义消除方法
By leveraging knowledge base metadata and adapter networks to enhance entity embedding, the problem of insufficient semantic specificity of entity embedding in entity linking is solved, thereby improving the accuracy of entity linking and the effect of polysemous word ambiguity resolution.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- UNIV OF ELECTRONICS SCI & TECH OF CHINA
- Filing Date
- 2023-06-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for entity linking suffer from insufficient semantic specificity in entity embedding, which increases the difficulty of resolving ambiguities of polysemous words and affects the effectiveness of entity linking.
By obtaining the entity embeddings from the entity linking model, utilizing knowledge base metadata and text embedding models, saliency coefficients are calculated and entity embeddings are adjusted. An adapter network is then used for reinforcement processing, which is then applied to the entity linking model to improve the semantic specificity of the entity embeddings.
It improves the accuracy of entity linking, reduces misjudgment of similar candidate entities, lowers the difficulty of entity linking, and enhances the effect of polysemous ambiguity resolution.
Smart Images
Figure CN116702790B_ABST