Word sense disambiguation method based on quasi-quantum non-homologous representation constraint
Through the method based on quantum-like non-homologous representation constraints, the problem of long-tail word meaning disambiguation task under low resources is solved, and the effective recognition and accurate representation of long-tail word meanings is achieved, which significantly improves the performance of word meaning disambiguation model.
Patent Information
- Application Number
- CN202411892033.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-20
- Publication Date
- 2025-04-15
AI Technical Summary
The prior art is difficult to effectively solve the long-tail word meaning disambiguation task under low resources, especially when the number of training samples is insufficient and the word meaning annotation text is limited.
Using a method based on the constraints of a class of quantum non-homologous representation, the traditional and quantum word meaning recognition scores are calculated by obtaining the target word representation and word meaning definition representation, and adding them as the final word meaning disambiguation result, to improve the accuracy of word meaning definition representation and distinguish boundaries.
It significantly improves the accuracy of long-tail word meaning recognition under low resources and can be applied to text representation problems under other low resources, verifying the effectiveness of this method in the long-tail word meaning disambiguation task.