A dialogue question generation method, device and equipment and storage medium
By identifying and predicting dialogue intent and combining it with reference documents to generate answers and questions for the next round of dialogue, the problem of low efficiency and high cost in existing technologies is solved, and automated, low-cost and stable dialogue question generation is achieved.
Patent Information
- Application Number
- CN202211418251.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-14
- Publication Date
- 2026-07-24
- Estimated Expiration
- 2042-11-14
AI Technical Summary
Existing human-based dialogue question generation solutions are inefficient and costly, while deep learning-based methods require laborious answer labeling and lack generalization ability.
By acquiring reference documents and historical dialogue data in a specified scenario, the intent of each round of dialogue is identified, and a pre-trained intent recognition model is used to predict the intent of the next round of dialogue. The answers and questions for the next round of dialogue are generated by combining the reference documents and target intent information, and an automated method is used without the need for known answers.
It achieves efficient and low-cost automatic generation of questions and answers for the next round of dialogue, and has strong generalization ability and stable question-answer pair generation effect.