Answer retrieval method and device based on in-depth question answering

Abstract

The present invention proposes an answer retrieval method and device based on in-depth question and answer, wherein the method includes: receiving an input query sentence; retrieving a web page containing candidate answers according to the query sentence; based on a web page analysis method, a deep learning method, and a template matching method Obtaining the first candidate answer, the second candidate answer and the third candidate answer respectively; determining the final answer according to the first candidate answer, the second candidate answer and the third candidate answer; presenting the final answer. The invention can directly display the answers corresponding to the user's query questions on the search result page, improve the relevance and accuracy of answer retrieval, and improve the presentation effect of search results.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to an answer retrieval method and device based on in-depth question and answer. Background technique [0002] Deep Question and Answer (DeepQA) refers to understanding human language, intelligently identifying the meaning of questions, and extracting answers to questions from massive Internet data. [0003] With the rapid development of the Internet, the functions of search engines are becoming more and more powerful, and users have higher and higher expectations for search engines, starting to change from basic relevant web page recall to intelligent question answering. When a user enters a question to be queried through a search engine, the search result that is expected to be obtained is no longer a relevant web page, but an answer to the question can be obtained directly. [0004] However, the existing search engine technology can only provide users with highly relevant we...

AI Technical Summary

Problems solved by technology

[0004] However, the existing search engine technology can only provide users with highly relevant web pages as search results through information retrieval and document summarization technology. , and extract the desired answer by itself

Therefore, the existing search engine technology does not have an in-depth question and answer function, and cannot directly provide users with answers to questions, and the presentation effect of search results is poor.

Images