Question-answer dynamic matching method, device and electronic device

Abstract

This application proposes a question-and-answer dynamic matching method, device, and electronic equipment, which relate to the field of intelligent search, wherein the method includes: obtaining query sentences and candidate paragraphs; performing entity recognition on query sentences and candidate paragraphs to obtain multiple entity vectors, and according to multiple The entity vector constructs a knowledge graph; updates each entity vector in the knowledge graph to generate an entity vector to be matched; performs matching processing on the entity vector to be matched to obtain the matching degree between the query statement and the candidate paragraph, and determines the target paragraph according to the matching degree. To solve the technical problem that the matching degree between the question and answer paragraphs cannot be accurately calculated in the prior art, resulting in inaccurate target paragraphs, by dynamically updating the representation of the entity vectors in the question and answer paragraphs, based on the updated entity vector Perform matching degree calculations to improve the accuracy of target paragraphs and meet user needs.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a question-answer dynamic matching method, device and electronic equipment in the field of intelligent search. Background technique [0002] Generally, related question-answering tasks such as text similarity, question-answer matching, and answer extraction can be processed by pre-trained models. In related technologies, the pre-trained knowledge vectors are directly introduced into the matching model to obtain the matching degree of questions and answers. However, the above-mentioned pre-trained knowledge vectors are knowledge vectors trained based on general knowledge graphs, resulting in a matching degree of questions and answers that is different from the actual The situation does not match, which ultimately affects the accuracy of the target answer. Contents of the invention [0003] The present application aims to solve one of the above-mentioned technical pr...

AI Technical Summary

Problems solved by technology

[0004] For this reason, the first purpose of this application is to propose a question-and-answer dynamic matching method, which solves the technical problem that the matching degree between the question and answer paragraphs cannot be accurately calculated in the prior art, resulting in the determined target paragraphs being inaccurate. By dynamically updating the representation of the entity vectors in the question and answer paragraphs, and calculating the matching degree based on the updated entity vectors, the accuracy of the target paragraphs is improved to meet the needs of users

Images