KR-20260063322-A - METHOD AND SYSTEM FOR PROCESSING QUERY PLUG-IN RECOMMENDATION BASED ON GENERATIVE AI

KR20260063322AKR 20260063322 AKR20260063322 AKR 20260063322AKR-20260063322-A

Abstract

A method and system for recommending query processing plug-ins based on a generative artificial intelligence service are provided. A query processing plug-in recommendation system according to one embodiment acquires a query, receives judgment criteria information for a plug-in from a generative artificial intelligence service, and can receive a suitability rating for each plug-in from the generative artificial intelligence service using the judgment criteria information. The query processing plug-in recommendation system can select a recommended plug-in corresponding to the query using the received suitability rating and output information regarding the selected recommended plug-in.

Inventors

김광석

Assignees

삼성에스디에스 주식회사

Dates

Publication Date: 20260507
Application Date: 20241030

Claims (20)

In a method performed by a computing device, Step of obtaining a query; A step of automatically generating a first prompt for obtaining judgment criterion information for the above query and transmitting it to a generative artificial intelligence service, wherein the judgment criterion information includes one or more judgment criteria and their weights; A step of receiving the judgment criteria information in response to the transmission of the first prompt from the generative artificial intelligence service; A step of automatically generating a second prompt for obtaining a suitability based on the judgment criterion information for each of the previously registered multiple plug-ins using the above judgment criterion information and transmitting it to the generative artificial intelligence service, wherein the second prompt includes functional information for each of the multiple plug-ins; A step of receiving the suitability of each of the plurality of plug-ins in response to the transmission of the second prompt from the generative artificial intelligence service; A step of selecting some of the plurality of plug-ins as recommended plug-ins corresponding to the query using the above suitability; and A step comprising outputting information about the selected recommended plug-in, Recommended method for query processing plug-ins.
In Article 1, The above judgment criteria information is, Including one or more evaluation items for each of the above judgment criteria, Recommended method for query processing plug-ins.
In Article 1, The step of automatically generating the above-mentioned first prompt and transmitting it to a generative artificial intelligence service is: A step comprising automatically generating the first prompt including the above query and user information requesting the above query, Recommended method for query processing plug-ins.
In Article 1, The step of automatically generating the above-mentioned first prompt and transmitting it to a generative artificial intelligence service is: A step comprising automatically generating the first prompt including the above query and the adjacent dialogue of the above query, Recommended method for query processing plug-ins.
In Article 1, The step of automatically generating the second prompt and transmitting it to the generative artificial intelligence service is A step comprising generating the second prompt, which further includes the received judgment criterion information, Recommended method for query processing plug-ins.
In Article 1, The step of automatically generating the second prompt and transmitting it to the generative artificial intelligence service is A step comprising generating the second prompt, which further includes a request to generate the suitability of each of the plurality of plug-ins using the judgment criteria information generated by the generative artificial intelligence service according to the first prompt. Recommended methods for query processing plug-ins.
In Article 1, Function information for each of the above plurality of plug-ins is, including the function text and user review text of the above plug-in, Recommended methods for query processing plug-ins.
In Article 1, The step of obtaining the above query is, It includes the step of receiving the above query from a user terminal, and The step of outputting information about the selected recommended plug-ins above is, A step of generating a plug-in selection request message including summary information and suitability information for the recommended plug-in above; and The method comprising the step of transmitting the above plug-in selection request message to the user terminal. Recommended methods for query processing plug-ins.
In Article 8, The step of transmitting the above plug-in selection request message to the user terminal is: A step comprising transmitting the plug-in selection request message such that the plug-in selection request message is displayed as a reply message immediately following the message of the query in the dialog box where the query is entered. Recommended methods for query processing plug-ins.
In Article 8, The above plug-in selection request message is, Includes a selection item for each of the above recommended plug-ins and a selection item indicating not to use the plug-in, Recommended methods for query processing plug-ins.
In Article 10, When a result is received from the user terminal in which a selection item indicating that the plug-in is not used is selected, the method includes the step of adding negative feedback previously designated as user review information for each of the recommended plug-ins. Recommended method for query processing plug-ins.
In Article 8, The step of outputting information about the selected recommended plug-ins above is, A step of determining a plug-in to be applied to the query using plug-in designation information received from the user terminal; and A method further comprising the step of transmitting the query to the service server of the determined plug-in. Recommended method for query processing plug-ins.
A communication interface connected to the first user's user terminal; Memory for loading query processing plug-in recommendation programs and; Includes one or more processors that execute the above-mentioned query processing plug-in recommendation program, The above query processing plug-in recommendation program is, Instruction to obtain a query; An instruction that automatically generates a first prompt for obtaining judgment criterion information for the above query and transmits it to a generative artificial intelligence service, wherein the judgment criterion information includes one or more judgment criteria and their weights; An instruction for receiving the judgment criterion information in response to the transmission of the first prompt from the generative artificial intelligence service; Instructions for automatically generating a second prompt for obtaining suitability based on the judgment criterion information for each of the previously registered multiple plug-ins using the above judgment criterion information and transmitting it to the generative artificial intelligence service, wherein the second prompt includes functional information for each of the multiple plug-ins; An instruction to receive the suitability of each of the plurality of plug-ins in response to the transmission of the second prompt from the generative artificial intelligence service; Instructions for selecting some of a plurality of plug-ins as recommended plug-ins corresponding to the query using the above suitability; and Includes an instruction that outputs information about the selected recommended plug-in, Query processing plug-in recommendation system.
In Article 13, The above judgment criteria information is, Including one or more evaluation items for each of the above judgment criteria, Query processing plug-in recommendation system.
In Article 13, The instruction that automatically generates the above-mentioned first prompt and transmits it to a generative artificial intelligence service is, Including an instruction that automatically generates the first prompt, which includes the above query and user information requesting the above query. Query processing plug-in recommendation system.
In Article 13, The instruction that automatically generates the above-mentioned first prompt and transmits it to a generative artificial intelligence service is, Instructions for automatically generating the first prompt including the above query and the adjacent dialogue of the above query, Query processing plug-in recommendation system.
In Article 13, The instruction that automatically generates the above second prompt and transmits it to the above generative artificial intelligence service is, Including an instruction that generates the second prompt, which further includes the received judgment criterion information. Query processing plug-in recommendation system.
In Article 13, The instruction that automatically generates the above second prompt and transmits it to the above generative artificial intelligence service is, Including an instruction for generating the second prompt, which further includes a request to generate the suitability of each of the plurality of plug-ins using judgment criterion information generated by the generative artificial intelligence service according to the first prompt. Query processing plug-in recommendation system.
In Article 13, Function information for each of the above plurality of plug-ins is, including the function text and user review text of the above plug-in, Query processing plug-in recommendation system.
Query processing plug-in recommendation system; and Includes generative AI services, The above query processing plug-in recommendation system is, Instruction to obtain a query; An instruction that automatically generates a first prompt for obtaining judgment criterion information for the above query and transmits it to the generative artificial intelligence service, wherein the judgment criterion information includes one or more judgment criteria and their weights; An instruction for receiving the judgment criterion information in response to the transmission of the first prompt from the generative artificial intelligence service; Instructions for automatically generating a second prompt for obtaining suitability based on the judgment criterion information for each of the previously registered multiple plug-ins using the above judgment criterion information and transmitting it to the generative artificial intelligence service, wherein the second prompt includes functional information for each of the multiple plug-ins; An instruction to receive the suitability of each of the plurality of plug-ins in response to the transmission of the second prompt from the generative artificial intelligence service; Instructions for selecting some of a plurality of plug-ins as recommended plug-ins corresponding to the query using the above suitability; and Includes an instruction that outputs information about the selected recommended plug-in, Query processing plug-in recommendation system.

Description

Method and System for Producing Query Plug-in Recommendation Based on Generative AI The present disclosure relates to a computing system and method for recommending a plug-in to a user that processes a user's query using a generative artificial intelligence service. More specifically, the disclosure relates to a computing system and method that improves the accuracy of responses to user queries by recommending an accurate plug-in using a generative artificial intelligence service. An AI-based question-answering service is provided. For example, a Large Language Model (LLM) is a technology that generates new content based on given input data, and when combined with Natural Language Processing (NLP) technology, it significantly enhances interaction with users. However, existing generative AI-based systems have limitations in the process of generating responses to user queries. For example, specialized fields such as medicine, law, and engineering require accurate information specific to those areas, but LLM systems often fail to fully understand and provide such information. Due to these issues, users may not be able to obtain necessary information in a reliable manner, which can lead to a decline in the reliability and usefulness of the service. Furthermore, conventional LLM-based services suffer from the problem that users must manually select the necessary plug-ins to obtain the information they desire. While plug-ins are additional tools that extend the functionality of the basic LLM or assist in handling specific tasks, users must fully understand the functions and usage of each plug-in to utilize them properly. However, most users are not sufficiently aware of how these plug-ins operate, resulting in instances where they incorrectly select a plug-in or fail to select one at all. These issues can lead to a decrease in the accuracy and efficiency of the information users seek, negatively impacting the user experience. Therefore, to solve the aforementioned problem, a system is required that automatically recommends appropriate plugins even if the user lacks in-depth knowledge of how they operate, and enables the user to obtain optimal information from the recommended plugins. This system can maximize user convenience by analyzing the user's query to determine which plugin is most suitable and providing a function to automatically select or recommend the appropriate plugin. Furthermore, it is particularly useful in situations involving complex query processing or requiring specific domain knowledge, and it can serve to complement the limitations of existing LLM services. FIG. 1 is a configuration diagram of a query processing plug-in recommendation system according to one embodiment of the present disclosure. FIG. 2 is a flowchart of a query processing plug-in recommendation method according to another embodiment of the present disclosure. FIG. 3 is a diagram illustrating an example of judgment criteria information for a query received from an LLM in some embodiments of the present disclosure. Figure 4 is a flowchart for explaining in more detail some operations of the query processing plug-in recommendation method described with reference to Figure 2. FIG. 5 is a diagram illustrating an example of obtaining a suitability based on a plurality of plug-in judgment criteria information from a generative artificial intelligence service in some embodiments of the present disclosure. FIG. 6 is a hardware configuration diagram of a computing system described in some embodiments of the present disclosure. Preferred embodiments of the present disclosure are described in detail below with reference to the accompanying drawings. The advantages and features of the present disclosure and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present disclosure is not limited to the embodiments described below but may be implemented in various different forms. The embodiments are provided merely to make the present disclosure complete and to fully inform those skilled in the art of the scope of the invention, and the present disclosure is defined only by the scope of the claims. It should be noted that when assigning reference numerals to the components of each drawing, the same components are given the same reference numeral whenever possible, even if they are shown in different drawings. Furthermore, in describing the present disclosure, if it is determined that a detailed description of related known components or functions could obscure the essence of the present disclosure, such detailed description is omitted. Unless otherwise defined, all terms used herein (including technical and scientific terms) may be used in a meaning commonly understood by those skilled in the art to which this disclosure pertains. Furthermore, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively