Question and answer recommendation method based on artificial intelligence, device and computer device

Abstract

The application presents a question and answer recommendation method based on artificial intelligence, a question and answer recommendation device and a computer device. The question and answer recommendation method based on artificial intelligence includes steps of receiving search terms input by a user, wherein the search terms are questions; matching the search terms with questions having answers in a question and answer knowledge base; if the terms in question and answer knowledge base are not matched with the search terms completely, calculating the semantic similarity between the search terms and the questions having answers in the question and answer knowledge base; judging if the question and answer knowledge base has a question of which semantic similarity with the search terms is larger than a preset threshold value; if it does, using the answer of which semantic similarity with the search terms is more than the preset threshold value as the answer of the search term, and recommending to a user. Answer recommendation is realized according to the semantic similarity, thus the meaning transfer risk is reduced and the interference by meaningless oral expression element is reduced.

Description

technical field [0001] The present application relates to the field of search technology, and in particular to an artificial intelligence-based question-and-answer recommendation method, device and computer equipment. Background technique [0002] With the rapid development of the Internet, some question answering systems with a large number of human participation have gradually begun to transform into automation and manual combination. Using automated question and answer recommendations to solve some problems can reduce manual participation and quickly respond to user needs. [0003] In the existing related technologies, words are generally used for simple matching, and recommendations are made according to the degree of overlap between the words in the target search word (query) and the questions in the existing knowledge base. However, this solution only uses words for simple matching, which is prone to Escaping, and unable to distinguish meaningless colloquial expression...

AI Technical Summary

Problems solved by technology

[0003] In the existing related technologies, words are generally used for simple matching, and recommendations are made according to the degree of overlap between the words in the target search word (query) and the questions in the existing knowledge base. However, this solution only uses words for simple matching, which is prone to Escaping, and unable to distinguish meaningless colloquial expressions, resulting in recommended answers that cannot meet user needs and poor user experience

Images