KR-20260064680-A - QUERY PROCESSING METHOD AND SYSTEM

KR20260064680AKR 20260064680 AKR20260064680 AKR 20260064680AKR-20260064680-A

Abstract

The present invention relates to a query processing method and system for outputting a response according to a user's query intent. A query processing method according to the present invention may include the steps of receiving a user query from a user, generating a normalized query corresponding to the user query using a query normalization model trained to generate a normalized sentence or word with at least one sentence as input, generating a response corresponding to the normalized query, and providing the response to the user.

Inventors

이유영
김혜영
김선라
인수교
문기윤
김경덕
서수빈
남경민
김현욱

Assignees

네이버 주식회사

Dates

Publication Date: 20260507
Application Date: 20260427

Claims (14)

In a query processing method performed by at least one processor, Step of receiving a user query; A step of generating a normalized query from the user query using a query normalization model trained to output a normalized query; A step of obtaining a plurality of responses including a response to the normalized query and a response to the user query; and A query processing method characterized by including the step of determining, based on a predetermined standard, one of the plurality of responses that reflects the query intent of the user query as the final response, and providing the final response to the user.
In paragraph 1, The step of determining the above final response is, A query processing method characterized by determining the response with the highest score among the plurality of responses as the final response, based on a score indicating the degree to which each of the plurality of responses reflects the query intent of the user query.
In paragraph 2, The step of determining the above final response is, A query processing method characterized by further including the step of determining the response to the user query as the final response when the score of the response with the highest score among the plurality of responses is less than a preset threshold.
In paragraph 1, The step of generating the above-mentioned normalized query is, It includes the step of generating a plurality of normalized queries from the above user query, and The step of obtaining the above multiple responses is, A query processing method characterized by including the step of obtaining a response to each of the above-mentioned plurality of normalized queries and a response to the above-mentioned user query as the above-mentioned plurality of responses.
In paragraph 1, The step of generating the above-mentioned normalized query is, If the above user query is related to a conversation context prior to the above user query, the normalized query is generated by combining some of the multiple words constituting the prior conversation and some of the multiple words constituting the above user query, and A query processing method characterized by generating the normalized query using the user query when the user query is not related to the previous conversation context.
In paragraph 5, A query processing method characterized by generating the normalized query by filling in the omitted component using the previous conversation when at least one component among the subject, object, and predicate indicating the query intent is omitted in the user query.
In paragraph 1, The normalization model of the above query is, A query processing method characterized by being trained to receive a conversation sequence containing two or more sentences, and to generate a normalized sentence for the last sentence by considering the context of the sentences preceding the last sentence in the conversation sequence.
In paragraph 1, The step of providing the above final response to the user is, The method further includes the step of providing the normalized query on a search page provided to the user, A query processing method characterized in that the normalized query is provided in different areas of the search page depending on whether the reliability of the normalized query satisfies a preset standard.
In paragraph 8, The above search page includes a query input area, and A query processing method characterized by providing the normalized query to the query input area and providing search results related to the normalized query to the search page when the reliability of the normalized query satisfies the previously set criteria.
In Paragraph 9, If the reliability of the normalized query above does not satisfy the pre-set criteria, the normalized query is provided as an icon in an area different from the query input area, and the user query is provided in the query input area. A query processing method characterized in that search results related to the above normalized query are provided on the search page when the above icon is selected.
In paragraph 1, The step of receiving the above user query is, A step of receiving a user voice from a user terminal; and A query processing method characterized by including the step of obtaining the user query from the user voice based on STT (Speech-to-Text) conversion.
In paragraph 1, The step of generating the above-mentioned normalized query is, A step of determining the query type of the above user query as one of a plurality of query types; and A query processing method characterized by including the step of generating the normalized query using the query normalization model learned based on different training data, based on the above-determined query type.
As a query processing system, Memory; and At least one processor connected to the memory and configured to execute at least one computer-readable program contained in the memory. Includes, The above at least one program is, Receive user queries from the user, Using a query normalization model trained to receive one or more sentences as input and output a normalized query, a normalized query is generated from the above user query, and Obtaining a plurality of responses including a response to the above normalized query and a response to the above user query, A query processing system characterized by including commands for determining, based on a predetermined standard, one of the plurality of responses that reflects the query intent of the user query as the final response, and providing the final response to the user.
A program that is executed by one or more processes in an electronic device and stored on a computer-readable recording medium, The above program is, Step of receiving a user query from a user; A step of generating a normalized query from the user query using a query normalization model trained to receive one or more sentences as input and output a normalized query; A step of obtaining a plurality of responses including a response to the normalized query and a response to the user query; and A program characterized by including instructions that, based on a pre-set standard, determine one of the plurality of responses that reflects the intent of the user query as the final response, and perform the step of providing the final response to the user.

Description

Query Processing Method and System The present invention relates to a query processing method and system for outputting a response according to the user's query intent. The dictionary definition of artificial intelligence is a technology that realizes human learning, reasoning, perception, and natural language understanding abilities through computer programs. This artificial intelligence has achieved rapid development through deep learning. In particular, driven by the advancement of artificial intelligence, various language models have been developed. These models have reached a level where they not only recognize text and understand its meaning but also extract and classify information from vast amounts of text-based data, such as documents, and even generate text directly. These language models are actively utilized in various fields, and there are diverse areas where text-based operations can be performed, such as search services, document creation (e.g., resume writing, report writing, posting, etc.), free conversation on various topics, data parsing from given text (e.g., data summarization, classification, etc.), provision of expertise, programming, and converting given sentences into sentences of an appropriate style. Meanwhile, in the case of search services, it is crucial to receive queries from users and derive accurate answers (or responses) that match their intent. In particular, as technology advances, the means and situations for receiving user queries have also diversified. Recently, much research has been conducted on technologies that utilize conversational agents to identify user queries and understand their intent within a dialogue between the agent and the user, thereby deriving appropriate answers. At this time, various studies are being conducted to identify the user's query intent, and research is needed on methods to accurately determine the user's query intent by actively utilizing language models composed of vast amounts of data. FIGS. 1 and 2 are conceptual diagrams for explaining a query processing method and system using a language model according to the present invention. Figures 3 and 4 are conceptual diagrams for explaining normalized queries. Figures 5 to 8 are conceptual diagrams for explaining training data. FIGS. 11 to 14 are conceptual diagrams for explaining examples of use of a query processing method and system according to the present invention. Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are assigned the same reference number regardless of the drawing symbols, and redundant descriptions thereof will be omitted. The suffixes “module” and “part” for components used in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the present invention. Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another. When it is stated that one component is “connected” or “connected” to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is “directly connected” or “directly connected” to another component, it should be understood that there are no other components in between. A singular expression includes a plural expression unless the context clearly indicates otherwise. In this application, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. FIGS. 1 and 2 are conceptual diagrams for explaining a query processing system according to the present invention, and FIGS. 3 and 4 are conceptual diagrams for explaining a normalized query. Furthermore, FIGS. 5 to 8 are conceptual