Data retrieval method based on semantic features, electronic device and storage medium

Abstract

The invention discloses a data retrieval method based on semantic features applied to an electronic device. The method comprises the steps of creating a question-answer database; when user problem data is received, processing the user question data to obtain a third word set composed of fourth words, inputting the user problem data into a sliding window model; sequentially sliding a window and calculating an average value of at least two fourth word vectors arranged side by side in the window; selecting the word vector with the maximum average value as a first sentence vector of the user question data; and meanwhile, converting each piece of preset question data in the question and answer database into a corresponding second sentence vector, selecting a question with the highest similaritywith the question data of the user by utilizing the first sentence vector and all the second sentence vectors, finding a preset answer corresponding to the question from the question and answer database, and feeding back the preset answer to the user. According to the invention, the historical questions whose expression meaning better conforms to the questions proposed by the user can be selected, and more accurate and reasonable answers can be obtained.

Description

technical field [0001] The present invention relates to the technical field of data processing, and in particular, to a data retrieval method, electronic device and storage medium based on semantic features. Background technique [0002] With the rapid development of computer technology and artificial intelligence, the intelligent question answering system has brought great convenience to life. The answers to the questions with the highest matching degree are found from the pre-created question and answer library through the questions raised by the users, and then fed back to the users, which replaces labor and reduces labor costs. However, most of the current intelligent question answering systems simply compare the similarity between the questions raised by the user and the historical questions in the question answering database, and then select the answer corresponding to the historical question with the highest similarity and feed it back to the user. It uses the litera...

AI Technical Summary

Problems solved by technology

However, most of the current intelligent question answering systems simply compare the similarity between the questions raised by the user and the historical questions in the question and answer database, and then select the answer to the question corresponding to the historical question with the highest similarity and give it back to the user. The literal features of the words contained in the question are used, and the semantic feature information of the words is ignored. As a result, even though the selected historical questions have a high degree of similarity, they do not necessarily have similar meanings.

Images