Question-answer pair generation method and device, electronic equipment and readable storage medium

Abstract

The invention relates to an intelligent decision, and discloses a question-answer pair generation method which comprises the steps of obtaining a corresponding word list from a first database according to a domain category to which an initial text belongs, and performing word segmentation processing on the initial text based on the word list to obtain a word sequence; inputting the word sequence into a trained question-answer pair generation model to obtain initial answer information of a question-answer pair expected to be generated, and performing filtering processing on the initial answer information to obtain a plurality of target answers; and inputting the word sequence and the target answers into the question-answer pair generation model to obtain a target question corresponding to each target answer, and forming a question-answer pair by each target answer and the corresponding target question to obtain a plurality of question-answer pairs. The invention further provides a question-answer pair generation device, electronic equipment and a readable storage medium. The invention can be applied to the field of digital medical treatment, the question-answer pair generation efficiency is improved, and the question-answer pair matching accuracy is improved.

Description

technical field [0001] The invention relates to the field of intelligent decision-making, in particular to a question-answer pair generation method, device, electronic equipment and readable storage medium. Background technique [0002] Open domain (that is, unlimited domain) question answering system is a hot topic in natural language processing, which can be applied to various aspects such as intelligent customer service, user guidance, and user FQA (question answering). For example, a company has released a new product. According to the product manual, multiple question-answer pairs can be generated. When a user inquires about the product, the intelligent customer service can match the user's questions according to the question-answer pair to determine the target question and answer. To answer users' questions. Currently, Q&A systems are usually built by manually sorting out Q&A data, which is inefficient, and when the Q&A data spans domains, there may be mismatches betw...

AI Technical Summary

Problems solved by technology

At present, the Q&A system is usually established by manually sorting out the Q&A data, which is inefficient, and when the Q&A data crosses domains, there may be cases where the manually sorted Q&A pairs do not match

Images