Deep learning-based question matching method and apparatus

Abstract

The invention discloses a deep learning-based question matching method and apparatus. The method comprises the steps of obtaining an input question of a user, and performing semantic analysis on the input question to obtain words of the input question; obtaining a similarity matching matrix between the words in the input question and words in a standard question, and performing hierarchical convolution based on the similarity matching matrix to obtain a semantic granularity matching feature between the input question and the standard question; based on a convolutional neural network, performing accurate matching on the input question and the standard question according to feature information of the input question and feature information of the standard question to obtain a fine granularity matching feature between the input question and the standard question; and obtaining a matching degree between the input question and the standard question according to the semantic granularity matching feature and the fine granularity matching feature. According to the method, the extracted features better meet actual demands; large cost for manual extraction of the features is omitted, and the accuracy of a matching result is improved.

Description

technical field [0001] The present invention relates to the fields of information technology and data business, in particular to a method and device for question matching based on deep learning. Background technique [0002] The current question answering system faces defects such as large consumption of human resources and untimely response. The goal of an automatic question answering system is to give short and precise answers to a given question. Whether it is an industry application or an academic research, the analysis of the true intention of the question and the identification of the matching relationship between the question and the answer are still constraints on automatic question answering. Two key puzzles in the performance of question answering systems. As we all know, question matching is to judge the matching relationship between user input questions and standard questions, or user questions and answers, and accurately locate the standard questions and answer...

AI Technical Summary

Problems solved by technology

[0004] However, the current problem is that machine learning requires manual extraction of features, and the accuracy of the model is very dependent on the design of features. In addition, the user's input questions in the question answering system are usually short texts with diverse, random, and irregular words. It is difficult to extract rich semantic features, and the designed features are not universal, resulting in inaccurate problem matching results, which in turn affects the positioning of standard problems and leads to poor user experience

Images