Generative artificial intelligence-based information provision apparatus, information provision system including same, and information provision method

Abstract

A generative artificial intelligence-based information provision system according to an embodiment of the present invention may comprise: a product database that stores product information according to a type name system; and an information provision apparatus that receives a prompt including search request information and weight information for the search request information, searches for, on the basis of the prompt, product information corresponding to the search request information from among the product information stored in the product database, acquires result information corresponding to the prompt by inputting the prompt and the found product information into a generative artificial intelligence model, and transmits the result information to a user terminal.

Description

A generative artificial intelligence-based information providing device, an information providing system including the same, and an information providing method

[0001] The present invention relates to a generative artificial intelligence-based information providing device, an information providing system including the same, and an information providing method.

[0002] Generative artificial intelligence (GAI) is a type of AI technology that generates or creates new data based on given input data. Because generative AI models generate new data based on their understanding of training data, they can be used in a variety of fields, including document writing, art creation, voice synthesis, game character creation, and engineering design.

[0003] In particular, chatbots, a key application of generative AI models, are systems that provide users with necessary information through conversation. They are being utilized in various fields for purposes such as user support, information provision, and task automation. However, simple Natural Language Processing (NLP) technology alone has difficulties accurately understanding and interpreting all user input. Furthermore, the so-called hallucination phenomenon, which provides incorrect information about unstudied content, has raised concerns about the accuracy and usefulness of generated content.

[0004] To address this, a new architecture called Retrieval-Augmented Generation (RAG) was developed, which improves the performance of text generation by integrating the Large Language Model (LLM) of a generative AI model with an information retrieval model.

[0005] Meanwhile, with the recent advancements in related technologies like the Internet of Things (IoT) and Artificial Intelligence (AI), smart factories are becoming a hot topic. A smart factory is a manufacturing system that operates by applying Information and Communications Technology (ICT) combined with digital automation solutions throughout the production process, from design and development to manufacturing.

[0006] One of the key requirements in operating a smart factory is to understand information that is closely related to products or businesses and is highly important and practical.

[0007] For example, it is necessary to identify products or parts of the company that have the same or similar rating conditions as a specific product or part, but are price competitive or have high product stability.

[0008] To this end, it is necessary to establish an information provision system utilizing the RAG architecture so that relevant parties, including smart factory operations managers and production planners, can quickly and accurately find the information they need.

[0009] The purpose of the present invention is to provide an advanced generative artificial intelligence-based information providing device that recommends products with a high degree of compatibility with user requests, an information providing system including the same, and an information providing method.

[0010] In one embodiment of the present invention, a generative artificial intelligence-based information providing system may include: a product database storing product information according to a type designation system; an information providing device receiving a prompt including search request information and weight information for the search request information, searching for product information corresponding to the search request information among product information stored in the product database based on the prompt, inputting the prompt and the searched product information into a generative artificial intelligence model to obtain result information corresponding to the prompt, and transmitting the result information to a user terminal.

[0011] The above search request information may include comparative product information including at least one of a product type, a product manufacturer, and a product model name, and at least one rating condition of the comparative product information.

[0012] The above weight information may include priorities between the rating conditions when there are multiple rating conditions or priorities between the multiple rating conditions and the product price.

[0013] The above information providing device can construct a vector database by vector embedding the above product information.

[0014] The above information providing device can generate the result information based on the similarity between the embedding vector of the prompt and the embedding vector in the vector database using a generative artificial intelligence model.

[0015] The above information providing device can calculate the similarity based on the weight information.

[0016] The above information providing device can build a user database by storing the search request information, the weight information, the result information, and user information corresponding to the search request information.

[0017] In one embodiment of the present invention, a generative artificial intelligence-based information providing device may include a processor that receives a prompt including search request information and weight information for the search request information, searches for product information corresponding to the search request information among product information stored in a product database according to a type system based on the prompt, inputs the prompt and the searched product information into a generative artificial intelligence model to obtain result information corresponding to the prompt, and transmits the result information to a user terminal.

[0018] A generative artificial intelligence-based information providing method performed by an information providing device according to one embodiment of the present invention may include the steps of: receiving a prompt including search request information and weight information for the search request information; searching for product information corresponding to the search request information among product information stored in a product database according to a type name system based on the prompt; inputting the prompt and the searched product information into a generative artificial intelligence model to obtain result information corresponding to the prompt; and transmitting the result information to a user terminal.

[0019] According to one embodiment of the present invention, it is possible to clearly understand a user's intention within a short period of time and recommend a product more suitable for the user's requirements, and cost savings are expected for operating a product-related consultation center.

[0020] According to one embodiment of the present invention, an independent database of product information within a smart factory can be built to provide information that is closely related to a target product or company and has high importance and practicality.

[0021] According to one embodiment of the present invention, the user's intention can be clearly identified through a chain structure with a large language model (LLM), thereby improving the reliability and validity of information.

[0022] FIG. 1 is a schematic diagram illustrating an information provision system according to one embodiment of the present invention.

[0023] FIG. 2 is a block diagram illustrating the configuration of an information providing device according to one embodiment of the present invention.

[0024] FIG. 3 is a diagram illustrating an operation flow chart of an information providing device according to one embodiment of the present invention.

[0025] FIG. 4 is a diagram illustrating a product naming system according to one embodiment of the present invention.

[0026] FIG. 5 is a diagram illustrating the operation of an information provision system according to one embodiment of the present invention.

[0027] FIG. 6 is a drawing illustrating a result information generation process of an information providing device according to one embodiment of the present invention.

[0028] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The detailed description set forth below, together with the accompanying drawings, is intended to explain exemplary embodiments of the present invention and is not intended to represent the only embodiments in which the present invention may be practiced. In the drawings, portions irrelevant to the description may be omitted for clarity in describing the present invention, and the same reference numerals may be used throughout the specification for identical or similar components.

[0029] FIG. 1 is a schematic diagram illustrating an information provision system according to one embodiment of the present invention.

[0030] An information provision system (1) (hereinafter referred to as system (1)) according to one embodiment of the present invention may include a database (10), an information provision device (100), and a user terminal (200).

[0031] A database (10) is a device that stores information required in the process of providing information by an information providing device (100). The database (10) may be implemented by dividing into two or more databases as needed, and the implementation method is not limited to either being implemented as a separate device from the information providing device (100) or being implemented within the information providing device (100).

[0032] The information providing device (100) is a device that provides information required during the operation of a smart factory, and can be implemented as a computer, server, smart phone, tablet PC, smart pad, laptop, etc.

[0033] The user terminal (200) is a device that inputs a prompt including search request information and weight information into a chatbot program implemented for product search and receives result information from an information providing device (100), and may be implemented as a computer, a PLC (Programmable Logic Controller), a server, a smart phone, a tablet PC, a smart pad, a laptop, etc. In this case, the user terminal (200) may be a terminal of a person in charge of a smart factory, a smart factory server, or equipment within a smart factory. The user terminal (200) may be a device that has a Manufacturing Execution System (MES), an Employee Assistance Program (EAP), a Recipe Parameter Management System (RPMS), etc. built in.

[0034] In addition, although not shown, the system (1) may further include a web server for collecting information required by the information providing device (100). The information providing device (100) may perform crawling to collect information from the web server.

[0035] In the present invention, an information provision system (1) utilizing the RAG architecture is proposed so that relevant parties, including smart factory operation managers, sales representatives, and production planners, can more quickly and accurately supply or receive necessary information.

[0036] Hereinafter, the configuration and operation of an information providing device (100) according to one embodiment of the present invention will be specifically described with reference to the drawings.

[0037] FIG. 2 is a block diagram illustrating the configuration of an information providing device according to one embodiment of the present invention.

[0038] An information providing device (100) according to one embodiment of the present invention may include an input unit (110), a communication unit (120), a display unit (130), a storage unit (140), and a processor (150).

[0039] The input unit (110) generates input data in response to user input of the information providing device (100). For example, the user input may be a user input that initiates the operation of the information providing device (100), a user input that stores product information according to a type naming system, a user input required to build a database, etc. In addition, the input may be applied without limitation to a user input required to transmit result information corresponding to search request information to a user terminal (200).

[0040] The input unit (110) includes at least one input means. The input unit (110) may include a keyboard, a key pad, a dome switch, a touch panel, a touch key, a mouse, a menu button, etc.

[0041] The communication unit (120) can perform communication with external devices such as a user terminal (200) and a web server to transmit and receive search request information, weight information, user information, product information, result information, etc.

[0042] To this end, the communication unit (120) can perform wireless communication such as 5G (5th generation communication), LTE-A (Long Term Evolution-Advanced), LTE (Long Term Evolution), Wi-Fi (Wireless Fidelity), Bluetooth, or wired communication such as LAN (Local Area Network), WAN (Wide Area Network), and power line communication.

[0043] The display unit (130) displays display data according to the operation of the information providing device (100). The display unit (130) can display a screen displaying search request information and weight information, a screen displaying product information, a screen displaying result information, a screen receiving user input, etc.

[0044] The display unit (130) includes a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, a micro electro mechanical systems (MEMS) display, and an electronic paper display. The display unit (130) may be implemented as a touch screen by being combined with the input unit (110).

[0045] The storage unit (140) stores the operation programs of the information providing device (100). The storage unit (140) includes a non-volatile storage that can preserve data (information) regardless of whether power is supplied, and a volatile memory that loads data to be processed by the processor (150) and cannot preserve data if power is not supplied. The storage includes a flash memory, a hard-disc drive (HDD), a solid-state drive (SSD), a read-only memory (ROM), etc., and the memory includes a buffer, a random access memory (RAM), etc.

[0046] The storage unit (140) may be implemented by including any one of the databases (10), as described above. The storage unit (140) may store search request information, user information, product information, result information, generative artificial intelligence models, etc., and may store operation programs required in the process of collecting and searching product information, vector embedding, operating generative artificial intelligence models, obtaining result information, etc.

[0047] The processor (150) can control at least one other component (e.g., hardware or software component) of the information providing device (100) by executing software such as a program, and can perform various data processing or operations.

[0048] A processor (150) according to one embodiment of the present invention may receive a prompt including search request information and weight information for the search request information, search for product information corresponding to the search request information among product information stored in the product database based on the prompt, input the prompt and the searched product information into a generative artificial intelligence model to obtain result information corresponding to the prompt, and transmit the result information to a user terminal.

[0049] At this time, the processor (150) may build a database and / or a generative artificial intelligence model, or may receive and store a previously built database and / or a generative artificial intelligence model from the outside and use it, but is not limited to either one.

[0050] Meanwhile, the processor (150) may perform at least a portion of the data analysis, processing, and result information generation for performing the above operations using at least one of a machine learning, neural network, or deep learning algorithm as a rule-based or artificial intelligence (AI) algorithm. Examples of the neural network may include models such as a convolutional neural network (CNN), a deep neural network (DNN), a recurrent neural network (RNN), and a transformer.

[0051] FIG. 3 is a diagram illustrating an operation flow chart of an information providing device according to one embodiment of the present invention.

[0052] A processor (150) according to one embodiment of the present invention can receive a prompt including search request information and weight information for the search request information (S10).

[0053] Search request information is the basic information required to perform a search at the request of an internal representative or client. It may include comparative product information and at least one rating condition. Comparative product information refers to information about products to be compared with product information stored in the product database, and may include at least one of the following: product type, product manufacturer, and product model name. Weighting information may include information indicating the priority between rating conditions when there are multiple rating conditions, as well as information indicating the priority between multiple rating conditions and product price.

[0054] For example, search request information may include rating conditions for a circuit breaker product from circuit breaker manufacturer A, with a rated voltage of 7.2 kV, a rated current of 1250 A, and an H-type drawer type. In this case, weighting information that prioritizes the product in the order of rated voltage, rated current, and type may be included. As another example, weighting information that prioritizes the product in the order of type, product price, rated voltage, and rated current may be included. This is because even if a product has a higher rated voltage and rated current than required, it may be adopted if it is reasonably priced.

[0055] The processor (150) can receive a prompt including search request information and weight information from the user terminal (200) through a chatbot program.

[0056] A prompt is a query input into a generative artificial intelligence model, for example, "Please recommend a vacuum circuit breaker of circuit breaker manufacturer A with a similar rated voltage and rated current to model VH-06-H-25-B-13." At this time, the weight information can be set in various ways, such as setting the priority in text format within the prompt, or setting the priority through a user interface implemented to select in order of importance among the received rating conditions. The weight information can be, for example, information selecting in order the three main factors that must be considered among the product model names, or information selecting in order of importance the three main factors and the product price.

[0057] A prompt according to one embodiment of the present invention may be a voice input in addition to a text input, and the format of the prompt is not limited to any one.

[0058] A processor (150) according to one embodiment of the present invention can search for product information corresponding to search request information among product information stored in a product database based on a prompt (S20).

[0059] The processor (150) can build a product database by storing product information according to a model name system. The model name system is designed to enable easy identification of product characteristics, such as rating conditions and options, based solely on the model name. An example of the model name system is illustrated in FIG. 4.

[0060] Meanwhile, the processor (150) can collect product information based on search request information via a web server. For example, product information can be collected by searching for "Case A" on a search portal or the like, or product information can be collected within Company A's corporate website.

[0061] At this time, the processor (150) can collect product information using a separate program that performs crawling and scraping. Specifically, the processor (150) can retrieve an HTML page, parse the HTML / CSS, etc., and extract the necessary product information. Alternatively, the processor (150) can extract the necessary product information by calling an Open API (Rest API) to a service that provides an Open API. In addition, any technology that collects product information can be applied without limitation.

[0062] Meanwhile, as described above, the present invention adopts the RAG architecture to improve the performance of text generation of a generative artificial intelligence model, and the RAG architecture is largely composed of a generative artificial intelligence model and a vector database.

[0063] A generative AI model can be a large-scale language model, and a vector database is constructed by vector-embedding information that serves as a reference for the large-scale language model to provide answers to queries. This reference information can be product information or information retrieved from product information based on a prompt. This will be described in detail with reference to Figure 6.

[0064] A processor (150) according to one embodiment of the present invention can input a prompt and searched product information into a generative artificial intelligence model to obtain result information corresponding to the prompt (S30).

[0065] The processor (150) can store the searched product information by vector embedding it in a vector database. The processor (150) can generate result information based on the similarity between the embedding vector of the prompt and the embedding vector in the vector database. The similarity can be determined by a distance measurement method such as Euclidean distance or cosine similarity, and the similarity can also be measured in various other known ways. The processor (150) can generate result information by using the embedding vectors with the highest similarity through a generative artificial intelligence model. At this time, the processor (150) can obtain result information by forming an enhanced prompt that includes both the embedding vector of the prompt and the embedding vector in the vector database, and inputting the enhanced prompt into the generative artificial intelligence model.

[0066] A processor (150) according to one embodiment of the present invention can transmit result information to a user terminal (200) (S40).

[0067] The processor (150) can display result information on the user terminal (200) through a chatbot program.

[0068] The processor (150) can provide a plurality of result information listed according to the similarity between embedding vectors to the user terminal (200), and can receive a user input for selecting one result information from the user terminal (200).

[0069] The processor (150) can list multiple result information in order of high vector similarity with weights reflected, and can list multiple result information in order of low price, high product stability, etc., depending on the request.

[0070] The result information is detailed information about the product searched through the prompt, and may include the product name, product rating conditions, drawings of the searched product information, product price, etc.

[0071] According to one embodiment of the present invention, it is possible to clearly understand a user's intention within a short period of time and recommend a product more suitable for the user's requirements, and cost savings are expected for operating a product-related consultation center.

[0072] FIG. 4 is a drawing illustrating a product type naming system according to one embodiment of the present invention.

[0073] The product type naming system is designed to make it easy to identify product characteristics, such as rating conditions and options, using only the product model name.

[0074] For example, Fig. 4 illustrates a product type naming system of vacuum circuit breakers (VCB). The type naming system of a vacuum circuit breaker can be composed of a representative type name, rated voltage, type distinction, breaking current, phase-to-phase distance, external type distinction, and rated current.

[0075] The product name VH-06-H-25-B-13 is a vacuum circuit breaker with a rated voltage of 7.2 kV, an H-type draw-out type, a breaking current of 25 kA, a phase-to-phase distance of 210 mm, and a rated current of 1250 A.

[0076] The information providing device (100) can store product information in a product database based on a product type name system, and can provide a user interface implemented to set a priority based on the product type name system when receiving weight information, and a user interface implemented to set a priority among the product type name system and product price.

[0077] For example, when a user searches for a vacuum circuit breaker product that meets his / her requirements, he / she may receive weight information that sets priorities for at least one of the representative model name, rated voltage, type classification, breaking current, phase-to-phase distance, external classification, and rated current of the vacuum circuit breaker, and the price of the vacuum circuit breaker.

[0078] FIG. 5 is a diagram illustrating the operation of an information provision system according to one embodiment of the present invention.

[0079] Figure 5 specifically describes the operation process of the information provision device (100) described above with reference to Figure 3. Any content that overlaps with the content described with reference to Figure 3 will be omitted.

[0080] As described above, the database (10) of FIG. 1 may be implemented with two or more databases as needed. The database may include a user database (11), a product database (12), and a vector database (13).

[0081] At this time, each database can be implemented separately, but two or more databases can be combined and implemented as needed. For example, the user database (11) and the product database (12) can be implemented as a single database.

[0082] According to one embodiment of the present invention, the processor (150) may construct a user database (11) by storing search request information, weight information, and user information corresponding to the search request information. The user information may include the user's account information and affiliation information. The user's account information may include an email account, team channel address, etc.

[0083] The processor (150) can match the search request information and weight information received from the user terminal (200) with the result information selected by the user terminal (200) and store them as search history in the user database (11).

[0084] The processor (150) can collect product information and build a product database (12). The product database (12) can be implemented as a proprietary product database and a third-party product database. The proprietary product database can store product information including product names, product ratings, type designation systems, and product drawings. The third-party product database can similarly store product names, product ratings, and type designation systems.

[0085] The processor (150) can store product information in the product database (12) according to the type system illustrated in FIG. 4.

[0086] The processor (150) can extract and collect product information from a pre-existing product catalog or internal documents, and can collect product information based on search request information via a web server. In addition, product information can be collected in various ways, and the format or collection method of product information is not limited to any one.

[0087] The processor (150) can build a vector database (13) by vector-embedding product information.

[0088] According to one embodiment of the present invention, an independent database of product information within a smart factory can be built to provide information that is closely related to a target product or company and has high importance and practicality.

[0089] FIG. 6 is a drawing illustrating a result information generation process of an information providing device according to one embodiment of the present invention.

[0090] The information providing device (100) can perform product information search-importance evaluation-final selection based on a generative artificial intelligence model (610) to select product information most suitable for search request information among the collected product information.

[0091] When the information providing device (100) receives the prompt (620), it searches for the same product group in its product database based on the comparative product information described in the search request information, and compares at least one rating condition described in the search request information with at least one rating condition of the product searched in its product database.

[0092] Specifically, the information providing device (100) can compare the straight-line distance using the embedding vector of the rating conditions of its own product and the rating conditions of the comparison product to select the similarity, and can assign a weight to the similarity according to the priority indicated in the weight information. At this time, the information providing device (100) can use the embedding vector of the vector database (13) for the rating conditions of its own product, and can use the embedding vector of the prompt (620) for the rating conditions of a third-party product. Meanwhile, the information providing device (100) can obtain more specific third-party product information from the third-party product database based on the comparison product information, and compare the third-party product information with the information of its own product.

[0093] The information providing device (100) can compare weighted similarities and provide the top products in order of high similarity to the user terminal (200).

[0094] The information providing device (100) can store the final result information selected from among the plurality of result information provided to the user terminal (200) in the user database (11).

[0095] According to one embodiment of the present invention, the user's intention can be clearly identified through a chain structure with a large language model (LLM), thereby improving the reliability and validity of information.

Claims

1. In the information provision system based on generative artificial intelligence, A product database that stores product information according to a nomenclature system; Receive a prompt including search request information and weighting information for said search request information; Based on the above prompt, product information corresponding to the search request information is searched among product information stored in the product database, By inputting the above prompt and the above searched product information into the generative artificial intelligence model, the result information corresponding to the above prompt is obtained, An information providing system including an information providing device that transmits the above result information to a user terminal.

2. In paragraph 1, An information providing system wherein the above search request information includes comparative product information including at least one of a product type, a product manufacturer, and a product model name, and at least one rating condition of the comparative product information.

3. In paragraph 2, The above weight information is an information providing system including priorities between the above rating conditions when there are multiple rating conditions or priorities between the multiple rating conditions and product prices.

4. In paragraph 1, The above information providing device, An information providing system that constructs a vector database by vector embedding the above product information.

5. In paragraph 4, The above information providing device, An information providing system that generates the result information based on the similarity between the embedding vector of the prompt and the embedding vector in the vector database using a generative artificial intelligence model.

6. In paragraph 5, The above information providing device, An information providing system that calculates the similarity based on the above weight information.

7. In paragraph 5, The above information providing device, An information providing system that constructs a user database by storing the search request information, the weight information, the result information, and user information corresponding to the search request information.

8. In a generative artificial intelligence-based information providing device, Receive a prompt including search request information and weighting information for said search request information; Based on the above prompt, product information corresponding to the above search request information is searched among product information stored in the product database according to the nomenclature system, By inputting the above prompt and the above searched product information into the generative artificial intelligence model, the result information corresponding to the above prompt is obtained, An information providing device including a processor that transmits the above result information to a user terminal.

9. In paragraph 8, The above search request information is an information providing device including comparative product information including at least one of a product type, a product manufacturer, and a product model name, and at least one rating condition of the comparative product information.

10. In paragraph 9, The above weight information is an information providing device including priorities between the above rating conditions when there are multiple rating conditions or priorities between the multiple rating conditions and product prices.

11. In paragraph 8, The above processor, An information providing device that constructs a vector database by vector embedding the above product information.

12. In paragraph 11, The above processor, An information providing device that generates the result information based on the similarity between the embedding vector of the prompt and the embedding vector in the vector database using a generative artificial intelligence model.

13. In paragraph 12, The above processor, An information providing device that calculates the similarity based on the above weight information.

14. In paragraph 12, The above processor, An information providing device that constructs a user database by storing the search request information, the weight information, the result information, and user information corresponding to the search request information.

15. In a method for providing information based on artificial intelligence, which is performed by an information providing device, A step of receiving a prompt including search request information and weighting information for the search request information; A step of searching for product information corresponding to the search request information among product information stored in a product database according to a nomenclature system based on the above prompt; A step of inputting the above prompt and the above searched product information into a generative artificial intelligence model to obtain result information corresponding to the above prompt; An information providing method comprising a step of transmitting the above result information to a user terminal.

Images