Method and device for obtaining answer to question

Abstract

The invention provides a method and a device for obtaining answers to questions, and the method comprises the steps: obtaining question information inputted by a user; conducting word segmentation processing on the problem information based on a multi-modal model, determining a word segmentation result, and the multi-modal model comprising at least two of a word model, a character model, a pinyinmodel and a glyph model; establishing a dependency relationship between words according to a word segmentation result, and converting the question information into a query statement in a machine language form according to the dependency relationship; and querying a corresponding database according to the query statement, and displaying a corresponding query result. By means of the method and device for obtaining the answers to the questions, semantic analysis can be conducted more accurately based on multiple modes, so that the word segmentation result is more accurate. Semantic information ofsentences can be described more completely and comprehensively based on the dependency relationship between words, so that query sentences are more accurate, more accurate results can be queried, andthe query accuracy is improved.

Description

technical field [0001] The present invention relates to the technical field of natural language processing, in particular, to a method and device for obtaining answers to questions. Background technique [0002] At present, the mainstream financial problem search method is the database retrieval technology based on keyword matching. A large number of fields and data are stored in the database. When a user asks a question, the traditional word segmentation algorithm is used to extract the keywords of the question, and the query is entered into the database according to the keywords to find the result. [0003] This search technology mainly has the following problems and shortcomings: [0004] The result based on keyword matching will display a large number of documents containing keywords, which needs to be manually read and screened from the answers, which is inefficient. Moreover, based on keyword matching, it is impossible to understand the question, and it is difficult ...

AI Technical Summary

Problems solved by technology

[0004] The result based on keyword matching will display a large number of documents containing keywords, which needs to be manually read and screened from the answers, which is inefficient

And based on keyword matching, it is impossible to understand the question, and it is difficult to display an accurate answer. The returned results are often only relevant, but cannot answer the question exactly

Images