机器问答方法以及装置
By acquiring the question text and reference text, and using a question-and-answer model to generate the answer text, the problem of answer generation not meeting user needs in existing technologies is solved, and the accuracy and controllability of the answer text are improved.
CN116756278BActive Publication Date: 2026-07-17ALIBABA (CHINA) CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ALIBABA (CHINA) CO LTD
- Filing Date
- 2023-05-15
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing technologies that generate answers based on question text do not meet user needs and lack accuracy and controllability.
Method used
By acquiring the question text and reference text, an answer is generated using a question-answering model. The question-answering model is trained based on multiple sample pairs. The sample question groups carry weight labels, which are determined based on the priority and relevance of the sample reference statements. The answer text is generated by combining encoding and decoding layers.
Benefits of technology
It improves the accuracy and controllability of the answer text, ensuring that the generated answers better meet user needs.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116756278B_ABST
Abstract
本说明书实施例提供机器问答方法以及装置，其中所述机器问答方法包括：获取问句文本和参考文本，其中，所述参考文本包含与所述问句文本相关联的知识信息；将所述问句文本和所述参考文本输入问答模型，获得答案文本，其中，所述问答模型基于多个样本对训练获得，所述样本对包括样本问句组和样本答案，所述样本问句组包括样本问句和样本参考语句，所述样本问句组携带有权重标签，所述权重标签基于所述样本参考语句的优先级以及所述样本参考语句与所述样本问句之间的相关性确定。通过基于包含权重标签以及样本答案的样本对训练问答模型，从而提升由问答模型输出的，问句文本和参考文本对应的答案文本的准确度和可控性。
Need to check novelty before this filing date? Find Prior Art