KR102648300B1 - Method and system for generating product groups based on user's intent of search queries - Google Patents

Abstract

This disclosure relates to a method for generating a product group based on user intent of a search query. The product group generation method includes receiving data associated with a plurality of product queries, extracting a trend associated with at least one of a specific group or a specific period based on the data associated with the plurality of product queries, and one associated with the extracted trend. It includes the step of creating a product group including the above products.

Description

Method and system for generating product groups based on user intent of search queries {METHOD AND SYSTEM FOR GENERATING PRODUCT GROUPS BASED ON USER'S INTENT OF SEARCH QUERIES}

The present disclosure relates to a method and system for generating a product group based on user intent of a search query, and specifically, a method for extracting a trend based on data associated with a plurality of product queries and generating a product group including one or more products associated with the trend. and systems.

Recently, various online product information provision methods and product sales methods have been developed to help consumers' shopping convenience, such as television home shopping broadcasts, Internet shopping, and catalog shopping. Product sellers, shopping mall operators, open market operators, etc. can plan product promotions to enable products to be purchased cheaper or on better terms as a means of promoting or marketing online shopping services. For example, a product seller can target specific consumers, collect products expected to be needed by the target consumers, and provide an event where the products can be purchased at a discounted price. Alternatively, a product seller, etc. can collect products expected to be needed by consumers at a specific time and provide information about the products to consumers at that time (or earlier than that time). Through these product promotions, consumers at a specific time or belonging to a specific group can check and purchase the products they need at a glance. In addition, product sellers, shopping mall operators, open market operators, etc. can improve sales by changing the status of stagnant product sales through product promotions and entice potential customers to use related online shopping services.

According to conventional technology, product sellers, shopping mall operators, open market operators, etc. who want to plan product promotions must passively analyze and identify consumers' shopping trends and create product groups corresponding to the shopping trends. In other words, operators who want to create product groups that respond to trends face the inconvenience of having to directly identify shopping trends and find products that fit the shopping trends. In addition, consumers are also provided with a product group where multiple products are simply classified by product category, similarity in product name, etc., so they cannot receive the products they actually need or want to purchase together at the same time.

The present disclosure provides a method for generating a product group, a computer program stored in a recording medium, and a device (system) to solve the above problems.

The present disclosure may be implemented in various ways, including as a method, device (system), or computer program stored in a readable storage medium.

According to one embodiment of the present disclosure, a method of generating a product group based on user intent of a search query performed by at least one computing device includes receiving data associated with a plurality of product queries, Based on this, it includes extracting a trend associated with at least one of a specific group or a specific period, and generating a product group including one or more products associated with the extracted trend.

According to another embodiment of the present disclosure, a method of learning a product extraction model performed by at least one computing device includes receiving data on a learning product group, receiving information on each of a plurality of products, and learning. Based on data on the product group and information on each of the plurality of products, as a trend keyword is input, it includes learning a product extraction model to output one or more products matching the trend keyword.

According to another embodiment of the present disclosure, a computer program stored in a computer-readable recording medium is provided to execute the above-described method on a computer.

According to another embodiment of the present disclosure, the information processing system includes a communication module, a memory, and at least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory, and at least one The program receives data associated with a plurality of product queries, extracts a trend associated with at least one of a specific group or a specific period based on the data associated with the plurality of product queries, and selects one or more products associated with the extracted trend. Contains commands for creating a product group.

According to another embodiment of the present disclosure, the information processing system includes a communication module, a memory, and at least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory, and at least one The program receives data about the learning product group, receives information about each of the plurality of products, and based on the data about the learning product group and information about each of the plurality of products, as the trend keyword is input, the trend keyword Includes commands for learning a product extraction model to output one or more products that match.

In various embodiments of the present disclosure, product queries searched by consumers with similar characteristics and/or product queries searched by consumers at similar times are determined to be related to the same shopping purpose or intention, and the shopping purpose or intention is the same. Product queries or keywords can be clustered into multiple clusters. By extracting shopping trends from a plurality of clustered groups, shopping trends that have not been discovered by users (for example, operators of online shopping malls or product promotions) can be identified.

In various embodiments of the present disclosure, by entering a keyword for a shopping trend (e.g., a product query corresponding to a shopping trend, a keyword indicating the information of the consumer who entered the product query, a time period, etc.), the trend is automatically identified. You can create phrases or sentences that represent a topic.

In various embodiments of the present disclosure, a plurality of products may be clustered into a plurality of product groups by reflecting shopping purposes or search intentions recently shared by many consumers, rather than simply clustering them based on category similarities between products. Accordingly, the diversity and suitability of products included in one product group can be improved at the same time. Additionally, users (for example, operators of online shopping malls or product promotions) can be provided with information about products according to shopping purposes or search intentions shared by multiple consumers.

The effects of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be explained by those skilled in the art in the technical field pertaining to the present disclosure from the description of the claims (hereinafter referred to as 'those skilled in the art'). can be clearly understood.

Embodiments of the present disclosure will be described with reference to the accompanying drawings described below, in which like reference numerals indicate like elements, but are not limited thereto.
1 is a diagram illustrating an example in which a user is provided with shopping content through a screen of a user terminal according to an embodiment of the present disclosure.
Figure 2 is a schematic diagram showing a configuration in which an information processing system according to an embodiment of the present disclosure is connected to enable communication with a plurality of user terminals.
Figure 3 is a block diagram showing the internal configuration of a user terminal and an information processing system according to an embodiment of the present disclosure.
Figure 4 is a flowchart showing a method for creating a product group according to an embodiment of the present disclosure.
Figure 5 is a diagram illustrating an example of extracting trend keywords from data associated with a plurality of product inquiries according to an embodiment of the present disclosure.
Figure 6 is a diagram illustrating an example of determining the title of a special exhibition representing the theme of a trend based on a trend keyword according to an embodiment of the present disclosure.
Figure 7 is a diagram illustrating an example of extracting one or more products matching the trend keyword and/or trend topic from a plurality of products based on the trend keyword and/or trend topic according to an embodiment of the present disclosure. .
Figure 8 is a diagram showing an example of learning a product extraction model according to an embodiment of the present disclosure.
Figure 9 is a diagram illustrating an example in which a user (eg, an operator) creates a product promotion according to an embodiment of the present disclosure.
Figure 10 is a diagram illustrating an example in which a user (eg, an operator) is provided with one or more product promotion candidates according to an embodiment of the present disclosure.

Hereinafter, specific details for implementing the present disclosure will be described in detail with reference to the attached drawings. However, in the following description, detailed descriptions of well-known functions or configurations will be omitted if there is a risk of unnecessarily obscuring the gist of the present disclosure.

In the accompanying drawings, identical or corresponding components are given the same reference numerals. Additionally, in the description of the following embodiments, overlapping descriptions of identical or corresponding components may be omitted. However, even if descriptions of components are omitted, it is not intended that such components are not included in any embodiment.

Advantages and features of the disclosed embodiments and methods for achieving them will become clear by referring to the embodiments described below in conjunction with the accompanying drawings. However, the present disclosure is not limited to the embodiments disclosed below and may be implemented in various different forms. The present embodiments are merely provided to ensure that the present disclosure is complete and that the present disclosure does not convey the scope of the invention to those skilled in the art. It is provided only for complete information.

Terms used in this specification will be briefly described, and the disclosed embodiments will be described in detail. The terms used in this specification are general terms that are currently widely used as much as possible while considering the function in the present disclosure, but this may vary depending on the intention or precedent of a technician working in the related field, the emergence of new technology, etc. In addition, in certain cases, there are terms arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the relevant invention. Accordingly, the terms used in this disclosure should be defined based on the meaning of the term and the overall content of the present disclosure, rather than simply the name of the term.

In this specification, singular expressions include plural expressions, unless the context clearly specifies the singular. Additionally, plural expressions include singular expressions, unless the context clearly specifies plural expressions. When it is said that a certain part includes a certain element throughout the specification, this does not mean excluding other elements, but may further include other elements, unless specifically stated to the contrary.

Additionally, the term 'module' or 'unit' used in the specification refers to a software or hardware component, and the 'module' or 'unit' performs certain roles. However, 'module' or 'unit' is not limited to software or hardware. A 'module' or 'unit' may be configured to reside on an addressable storage medium and may be configured to run on one or more processors. Thus, as an example, a 'module' or 'part' refers to components such as software components, object-oriented software components, class components and task components, processes, functions and properties. , procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, or variables. Components and 'modules' or 'parts' may be combined into smaller components and 'modules' or 'parts' or further components and 'modules' or 'parts'. Could be further separated.

According to an embodiment of the present disclosure, a 'module' or 'unit' may be implemented with a processor and memory. 'Processor' should be interpreted broadly to include general-purpose processors, central processing units (CPUs), microprocessors, digital signal processors (DSPs), controllers, microcontrollers, state machines, etc. In some contexts, 'processor' may refer to an application-specific integrated circuit (ASIC), programmable logic device (PLD), field programmable gate array (FPGA), etc. 'Processor' refers to a combination of processing devices, for example, a combination of a DSP and a microprocessor, a combination of a plurality of microprocessors, a combination of one or more microprocessors in combination with a DSP core, or any other such combination of configurations. You may. Additionally, 'memory' should be interpreted broadly to include any electronic component capable of storing electronic information. 'Memory' refers to random access memory (RAM), read-only memory (ROM), non-volatile random access memory (NVRAM), programmable read-only memory (PROM), erasable-programmable read-only memory (EPROM), May also refer to various types of processor-readable media, such as electrically erasable PROM (EEPROM), flash memory, magnetic or optical data storage, registers, etc. A memory is said to be in electronic communication with a processor if the processor can read information from and/or write information to the memory. The memory integrated into the processor is in electronic communication with the processor.

In this disclosure, 'shopping content' refers to online shopping services connected to the Internet, broadcasts about shopping products provided through social media, information on product promotions, shopping product information, shopping payment information, shopping seller or manufacturer information, etc. It can refer to information or content related to online shopping. For example, shopping content may include, but is not limited to, content containing text, images or videos for descriptions, advertisements, and promotions of products subject to online shopping, live broadcast content, recorded broadcast content, etc. . As another example, shopping content may include information about one or more products or product groups in which a large number of consumers in a specific period or specific group are interested, the title of the product promotion, tags related to the product promotion, and information about the special products of the product promotion. It may include information, etc.

In the present disclosure, 'special sales events' may generally refer to an event that advertises, displays, or sells one or more products or product groups related to a specific theme during a specific period. For example, a special exhibition may be provided through an online shopping service, social media, etc., and may be provided to a specific group of users during a specific time (e.g., a season, a specific month of the year, a specific date or time). In another example, a special exhibition may be created or provided to advertise, display, or sell one or more products or product groups related to a specific product topic, such as summer vacation clothing, children's toys, camping supplies, etc. In one embodiment, the special exhibition includes the theme of the special exhibition, the title of the special exhibition indicating the product to be sold, etc., special product information (e.g., product image, product price, product description, etc.), the validity period and/or sales conditions of the special exhibition. It may be expressed as data including, but is not limited to this.

FIG. 1 is a diagram illustrating an example in which a user 100 is provided with shopping content through the screen 120 of the user terminal 110 according to an embodiment of the present disclosure. The screen 120 shown in FIG. 1 is displayed when a user (eg, consumer) 100 executes an application that provides shopping content on the user terminal (eg, smartphone) 110 to display shopping content. This shows an example of receiving information about a product group or product promotion. For example, as shown, the user 100 uses the user terminal 110 as shopping content to select a trending topic related to a specific product group (for example, 'How about decorating your home with lighting?', which is the title of a product exhibition). , hashtags related to the topic of the trend (e.g., trend keywords '#table lamp', '#floor lamp', '#furniture', '#lighting') and information about products related to the topic of the trend. (For example, product images, product names, product prices, etc. of special products included in the product planning exhibition) can be provided.

In order to provide product group information to the user 100 in this way, the information processing system communicating with the user terminal 110 generates product group information including one or more products related to the current trend (or trend at a specific point in time). This can be transmitted to the user terminal 110. To this end, the information processing system can extract one or more products associated with a specific trend. That is, the information processing system can create product groups that include one or more products related to a specific trend. For example, an information processing system may automatically generate a product group containing one or more products based on various search queries (e.g., product search queries).

The information processing system may receive data associated with a plurality of product queries to create a product group that includes one or more products. Here, 'product inquiry' may refer to a search query entered by a user through an online shopping application or web page, search application or web page run by the user terminal, for the purpose of searching or purchasing a specific product. . In addition, data related to a product inquiry includes the product query itself (e.g., search query word, text representing the search keyword, image, etc.), the date and time the product query was entered, and information on the user terminal where the product query was entered. (eg, location of the user terminal, IP information, etc.), and information of the user who entered the product inquiry (eg, user account or profile information, user age, gender, occupation, etc.).

The information processing system may extract a trend associated with at least one of a specific group or a specific period based on data associated with a plurality of product queries, and create a product group including one or more products associated with the extracted trend. Here, a specific group may refer to a group created by classifying a plurality of users based on one or more criteria (eg, age, gender, interests, occupation, etc.). For example, specific groups include groups classified by age such as teenagers, 20s, and 30s, groups classified by occupation such as office workers, self-employed, and college students, groups classified by gender such as men and women, married people, and It may include groups classified according to married status, such as single people. Additionally, a specific period may include a season, an anniversary, a weekday or weekend, a public holiday, a holiday period, a specific year, a specific month of the year, a specific week of the month, a section between specific points in the day, etc.

The information processing system clusters data associated with a plurality of product inquiries received into a plurality of clusters, and extracts one or more trend keywords associated with at least one of a specific group or a specific period from one or more of the plurality of clusters. You can. For example, the information processing system divides data associated with a plurality of product queries into a plurality of clusters based on the query intention, topic, information of the user who searched for the product query, the date and time the product query was entered, etc. It can be clustered. Additionally or alternatively, the information processing system may cluster data associated with multiple product queries into multiple clusters using a statistical or probability-based clustering model, such as a statistical theme embedding model. . Here, trend keywords may include product queries included in one or more clusters, groups associated with one or more clusters, periods, etc.

Then, the information processing system may extract one or more products associated with one or more trend keywords from the plurality of products and create a product group including the one or more extracted products. In one embodiment, the information processing system determines a topic of a trend (e.g., common characteristics among product queries included in one or more clusters, etc.) based on one or more trend keywords, and based on the determined trend topic, One or more products related to one or more trend keywords can be extracted from a plurality of products. For example, the information processing system may combine one or more trend keywords to generate a phrase or sentence representing the topic of the trend. Alternatively or additionally, the information processing system may use a language model, such as a language model based on statistics or a language model based on an artificial neural network, to convert (or generate) one or more trend keywords into phrases or sentences representing the topic of the trend. can do. The information processing system can extract one or more products associated with one or more trend keywords from a plurality of products based on the phrase or sentence representing the topic of the trend created in this way. In another embodiment, the information processing system may use a product extraction model to extract one or more products matching one or more trend keywords from a plurality of products.

In one embodiment, the product extraction model is based on information about each of a plurality of products (e.g., product name, product description, category, related product, tag, introduction, review, image, price, brand, manufacturer, etc.) Calculate a matching score with one or more trend keywords (e.g., product query included in one or more clusters, topic of trend, title of trend, phrase or sentence representing topic of trend, etc.) for each of the plurality of products. , Based on the calculated matching score, it may correspond to a model learned to extract a predetermined number of products from a plurality of products and output the extracted predetermined number of products as one or more products matching one or more trend keywords. there is. Additionally or alternatively, the product extraction model may include an embedding generation model trained to generate embedding data for each of a plurality of products and embedding data for one or more trend keywords. In this case, the product extraction model calculates a matching score with one or more trend keywords for each of the plurality of products based on the embedding data for each of the plurality of products and the embedding data for one or more trend keywords, and the calculated matching Based on the score, it may correspond to a model learned to extract a predetermined number of products from a plurality of products and output the extracted predetermined number of products as one or more products matching one or more trend keywords.

Additionally, the information processing system may generate a product promotion that includes at least one product as a special product among products included in the generated product group. In this case, the information processing system may determine a sentence or phrase representing the topic of the trend as the title of the special exhibition. In one embodiment, the information processing system outputs the generated product group through the user terminal, and in response to the user's selection of at least one product from the output product group, a product plan including at least one product as a special product. can be created. In another embodiment, the information processing system may automatically generate a special exhibition that includes at least one product from the generated product group as a special product.

According to the embodiments described above, trends related to a specific group and/or a specific period can be extracted by reflecting information on a plurality of product inquiries actually entered by consumers, and product groups and/or products that fit the extracted trends. You can create special exhibitions. Accordingly, the operator of the product promotion can create a product group and/or product promotion by identifying trends that were not previously thought of in a timely manner, and the product group and/or product promotion created in this way include various products that fit the newly identified trend. may be included. In addition, as trend keywords, trend topics, product groups, special exhibition topics, special exhibition titles, and/or product special exhibitions that fit the extracted trend are automatically extracted/generated, the operator's work burden can be reduced. In addition, as trend keywords, trend topics, product groups, special exhibition topics, special exhibition titles, and/or special products that match the current product sales trend are automatically recommended, operators can efficiently create/publish product special exhibitions.

Figure 2 is a schematic diagram showing a configuration in which the information processing system 230 according to an embodiment of the present disclosure is connected to communicate with a plurality of user terminals 210_1, 210_2, and 210_3. As shown, a plurality of user terminals 210_1, 210_2, and 210_3 may be connected to the information processing system 230 that can provide a product group creation/provision service through the network 220. Here, the plurality of user terminals 210_1, 210_2, and 210_3 may include terminals of users (eg, planners or operators of online shopping malls or product promotions, consumers, etc.) who will be provided with product group creation/provision services. In one embodiment, the information processing system 230 is one or more server devices and/or databases capable of storing, providing, and executing computer-executable programs (e.g., downloadable applications) and data related to product line creation/provision services, etc. , or may include one or more distributed computing devices and/or distributed databases based on cloud computing services.

The product group creation/provision service provided by the information processing system 230 includes a product group creation/provision application installed on each of the plurality of user terminals 210_1, 210_2, and 210_3, a product promotion creation application, a shopping application, a search application, and a mobile It may be provided to the user through a browser application or web browser. For example, the information processing system 230 provides information or performs processing in response to a request for product group creation/provision received from the user terminals 210_1, 210_2, and 210_3 through a product group creation/provision application, etc. can do.

A plurality of user terminals 210_1, 210_2, and 210_3 may communicate with the information processing system 230 through the network 220. The network 220 may be configured to enable communication between a plurality of user terminals 210_1, 210_2, and 210_3 and the information processing system 230. Depending on the installation environment, the network 220 may be, for example, a wired network such as Ethernet, a wired home network (Power Line Communication), a telephone line communication device, and RS-serial communication, a mobile communication network, a wireless LAN (WLAN), It may consist of wireless networks such as Wi-Fi, Bluetooth, and ZigBee, or a combination thereof. The communication method is not limited, and may include communication methods utilizing communication networks that the network 220 may include (e.g., mobile communication networks, wired Internet, wireless Internet, broadcasting networks, satellite networks, etc.) as well as user terminals (210_1, 210_2, 210_3). ) may also include short-range wireless communication between

In Figure 2, the mobile phone terminal (210_1), tablet terminal (210_2), and PC terminal (210_3) are shown as examples of user terminals, but they are not limited thereto, and the user terminals (210_1, 210_2, 210_3) use wired and/or wireless communication. This is possible and may be any computing device on which a product group creation/provision application, a product promotion creation application, a shopping application, a search application, a mobile browser application, or a web browser, etc. can be installed and executed. For example, user terminals include AI speakers, smartphones, mobile phones, navigation, computers, laptops, digital broadcasting terminals, PDAs (Personal Digital Assistants), PMPs (Portable Multimedia Players), tablet PCs, game consoles, It may include wearable devices, IoT (internet of things) devices, VR (virtual reality) devices, AR (augmented reality) devices, set-top boxes, etc. In addition, in Figure 2, three user terminals (210_1, 210_2, 210_3) are shown as communicating with the information processing system 230 through the network 220, but this is not limited to this, and a different number of user terminals are connected to the network ( It may be configured to communicate with the information processing system 230 through 220).

According to one embodiment, the information processing system 230 may receive data associated with a plurality of product inquiries from a plurality of user terminals 210_1, 210_2, and 210_3 (eg, a consumer's user terminal). Then, the information processing system 230 generates a product group based on data associated with the plurality of product queries thus received, and provides information about the generated product group (e.g., about at least some products included in the product group). Information, trend keywords of the product group, trend topics of the product group, etc.) can be provided to a plurality of user terminals 210_1, 210_2, 210_3 (e.g., user terminals of operators and/or consumers of an online shopping mall or product promotion). .

Figure 3 is a block diagram showing the internal configuration of the user terminal 210 and the information processing system 230 according to an embodiment of the present disclosure. The user terminal 210 may refer to any computing device capable of executing a product group creation/provision application, a product promotion creation application, a shopping application, a search application, a mobile browser application, or a web browser, and capable of wired/wireless communication, For example, it may include the mobile phone terminal 210_1, tablet terminal 210_2, and PC terminal 210_3 of FIG. 2 . As shown, the user terminal 210 may include a memory 312, a processor 314, a communication module 316, and an input/output interface 318. Similarly, information processing system 230 may include memory 332, processor 334, communication module 336, and input/output interface 338. As shown in FIG. 3, the user terminal 210 and the information processing system 230 are configured to communicate information and/or data through the network 220 using respective communication modules 316 and 336. It can be. Additionally, the input/output device 320 may be configured to input information and/or data to the user terminal 210 through the input/output interface 318 or to output information and/or data generated from the user terminal 210.

Memories 312 and 332 may include any non-transitory computer-readable recording medium. According to one embodiment, the memories 312 and 332 are non-perishable mass storage devices such as random access memory (RAM), read only memory (ROM), disk drive, solid state drive (SSD), and flash memory. It may include a (permanent mass storage device). As another example, non-perishable mass storage devices such as ROM, SSD, flash memory, disk drive, etc. may be included in the user terminal 210 or the information processing system 230 as a separate persistent storage device that is distinct from memory. Additionally, the memories 312 and 332 may store an operating system and at least one program code (for example, code for a product group creation/provision application installed and running on the user terminal 210).

These software components may be loaded from a computer-readable recording medium separate from the memories 312 and 332. This separate computer-readable recording medium may include a recording medium directly connectable to the user terminal 210 and the information processing system 230, for example, a floppy drive, disk, tape, DVD/CD- It may include computer-readable recording media such as ROM drives and memory cards. As another example, software components may be loaded into the memories 312 and 332 through a communication module rather than a computer-readable recording medium. For example, at least one program is loaded into memory 312, 332 based on a computer program installed by files provided over the network 220 by developers or a file distribution system that distributes installation files for applications. It can be.

The processors 314 and 334 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. Instructions may be provided to the processors 314 and 334 by memories 312 and 332 or communication modules 316 and 336. For example, the processors 314 and 334 may be configured to execute instructions received according to program codes stored in a recording device such as the memory 312 and 332.

The communication modules 316 and 336 may provide a configuration or function for the user terminal 210 and the information processing system 230 to communicate with each other through the network 220, and may provide a configuration or function for the user terminal 210 and/or information processing. The system 230 may provide a configuration or function for communicating with other user terminals or other systems (for example, a separate cloud system, etc.). For example, a request or data generated by the processor 314 of the user terminal 210 according to a program code stored in a recording device such as the memory 312 (e.g., a request for product group creation/provision, product promotion creation) A request for a product promotion, a request for the issuance of a product promotion, data associated with a plurality of product inquiries, etc.) may be transmitted to the information processing system 230 through the network 220 under the control of the communication module 316. Conversely, a control signal or command provided under the control of the processor 334 of the information processing system 230 is transmitted through the communication module 316 of the user terminal 210 through the communication module 336 and the network 220. It may be received by the user terminal 210. For example, the user terminal 210 receives trend keywords, trend topics, information about at least one product included in the product group, special exhibition topic candidates, and special exhibition targets from the information processing system 230 through the communication module 316. Candidates for consumers, candidates for trend keywords for special exhibitions, candidates for titles for special exhibitions, and/or candidates for special products to be included in special exhibitions can be received.

The input/output interface 318 may be a means for interfacing with the input/output device 320. As an example, input devices may include devices such as cameras, keyboards, microphones, mice, etc., including audio sensors and/or image sensors, and output devices may include devices such as displays, speakers, haptic feedback devices, etc. You can. As another example, the input/output interface 318 may be a means for interfacing with a device that has components or functions for performing input and output, such as a touch screen, integrated into one. For example, the processor 314 of the user terminal 210 uses information and/or data provided by the information processing system 230 or another user terminal when processing instructions of a computer program loaded in the memory 312. A service screen, etc. constructed by doing so may be displayed on the display through the input/output interface 318. In FIG. 3 , the input/output device 320 is shown not to be included in the user terminal 210, but the present invention is not limited to this and may be configured as a single device with the user terminal 210. Additionally, the input/output interface 338 of the information processing system 230 may be connected to the information processing system 230 or means for interfacing with a device (not shown) for input or output that the information processing system 230 may include. It can be. In FIG. 3, the input/output interfaces 318 and 338 are shown as elements configured separately from the processors 314 and 334, but the present invention is not limited thereto, and the input/output interfaces 318 and 338 may be configured to be included in the processors 314 and 334. there is.

The user terminal 210 and information processing system 230 may include more components than those in FIG. 3 . However, there is no need to clearly show most prior art components. According to one embodiment, the user terminal 210 may be implemented to include at least some of the input/output devices 320 described above. Additionally, the user terminal 210 may further include other components such as a transceiver, a global positioning system (GPS) module, a camera, various sensors, and a database. For example, if the user terminal 210 is a smartphone, it may include components generally included in a smartphone, such as an acceleration sensor, a gyro sensor, a camera module, various physical buttons, and a touch screen. Various components such as buttons using a panel, input/output ports, and a vibrator for vibration may be implemented to be further included in the user terminal 210. According to one embodiment, the processor 314 of the user terminal 210 may be configured to operate an application that provides a product group creation/provision service. At this time, code associated with the corresponding application and/or program may be loaded into the memory 312 of the user terminal 210.

While a program for an application providing a product group creation/provision service is running, the processor 314 inputs input such as a touch screen, keyboard, camera including an audio sensor and/or an image sensor, and a microphone connected to the input/output interface 318. Text, images, videos, voices and/or actions input or selected through the device can be received, and the received text, images, videos, voices and/or actions can be stored in the memory 312 or stored in the communication module 316. ) and can be provided to the information processing system 230 through the network 220. For example, the processor 314 receives a user input entering a product query through the input device 320, and processes the product query and data associated with the product query through the network 220 and communication module 316. It can be provided to the system 230. As another example, the processor 314 may include a trend keyword or a topic candidate for a trend, at least one product from a product group, a topic candidate for a special exhibition, a target consumer candidate for a special exhibition, a trend keyword candidate for a special exhibition, a title candidate for a special exhibition, and/or a product to be included in the special exhibition. By receiving input indicating the user's selection of promotional product candidates, etc. It can be provided to the information processing system 230 through the communication module 316 and the network 220.

The processor 314 of the user terminal 210 manages, processes, and/or stores information and/or data received from the input device 320, other user terminals, the information processing system 230, and/or a plurality of external systems. It can be configured to do so. Information and/or data processed by processor 314 may be provided to information processing system 230 via communication module 316 and network 220. The processor 314 of the user terminal 210 may transmit information and/or data to the input/output device 320 through the input/output interface 318 and output the information. For example, the processor 314 may display the received information and/or data on the screen of the user terminal.

The processor 334 of the information processing system 230 may be configured to manage, process, and/or store information and/or data received from a plurality of user terminals 210 and/or a plurality of external systems. Information and/or data processed by the processor 334 may be provided to the user terminal 210 through the communication module 336 and the network 220. In one embodiment, the processor 334 of the information processing system 230 generates a trend associated with at least one of a specific group or a specific period based on data associated with a plurality of product inquiries received from a plurality of user terminals 210. You can extract and create a product group containing one or more products related to the extracted trend. In another embodiment, the processor 334 of the information processing system 230 includes the at least one product as a special product in response to the user's selection of at least one product from the product group, which is received from the user terminal 210. You can create a special product exhibition.

The processor 334 of the information processing system 230 uses the output device 320, such as a display output capable device (e.g., touch screen, display, etc.), an audio output capable device (e.g., speaker), of the user terminal 210. It may be configured to output processed information and/or data. For example, the processor 334 of the information processing system 230 transmits the generated product group (e.g., information about the product group, information about products included in the product group, etc.) to the communication module 336 and the network 220. ) and may be configured to output the product group through a display output capable device of the user terminal 210.

Figure 4 is a flowchart showing a product group creation method 400 according to an embodiment of the present disclosure. In one embodiment, the product group creation method 400 may be performed by a processor (eg, at least one processor of an information processing system). The product group creation method 400 may be initiated by the processor receiving data associated with a plurality of product queries (S410). For example, the processor may receive data associated with a plurality of product inquiries from a plurality of user terminals. Additionally or alternatively, the processor may receive data associated with a plurality of product queries from internal and/or external storage devices (e.g., a product query database, etc.). Here, the data associated with the product query includes the product query itself (e.g., search query, text representing the search keyword, image, etc.), the date and time the product query was entered, and information on the user terminal where the product query was entered ( For example, the location of the user terminal, IP information, etc.), and information of the user who entered the product inquiry (for example, user account or profile information, user age, gender, occupation, etc.).

The processor may extract a trend associated with at least one of a specific group or a specific period based on data associated with a plurality of product inquiries (S420). In one embodiment, the processor clusters data associated with a plurality of product queries into a plurality of clusters, and from one or more of the plurality of clusters, one or more trend keywords associated with at least one of a specific group or a specific time period (e.g., Search queries, topics, dates, etc.) can be extracted. For example, the processor may use a statistical topic embedding model to cluster data associated with multiple product queries into multiple clusters.

Then, the processor may create a product group including one or more products associated with the extracted trend (S430). For example, the processor may extract one or more products associated with one or more trend keywords (e.g., search queries included in one or more clusters, etc.) from a plurality of products and create a product group containing the one or more extracted products. You can. In one embodiment, the processor determines a topic of a trend (e.g., a common topic among search queries included in one or more clusters, etc.) based on one or more trend keywords, and based on the topic of the trend, determines a plurality of trend keywords. Among products, one or more products related to one or more trend keywords can be extracted. To this end, the processor may use a language model to convert one or more trend keywords into phrases or sentences representing the topic of the trend. The phrase or sentence representing the theme of the converted trend may include a phrase or sentence for use as the title of a special exhibition. Additionally or alternatively, the processor may use a product extraction model to extract one or more products from a plurality of products that match one or more trend keywords (or the topic of the trend, a phrase or sentence representing the topic of the trend, etc.) there is.

In one embodiment, the product extraction model calculates a matching score with one or more trend keywords for each of the plurality of products, based on information about each of the plurality of products, and calculates a matching score for at least one or more trend keywords for each of the plurality of products based on the calculated matching score. It may correspond to a model learned to extract and output the corresponding product list. Furthermore, the product extraction model may correspond to a model learned to extract a predetermined number of products from a plurality of products and output the extracted predetermined number of products as one or more products matching one or more trend keywords. Additionally or alternatively, the product extraction model may include an embedding generation model trained to generate embedding data for each of a plurality of products and embedding data for one or more trend keywords. In this case, the product extraction model calculates a matching score with one or more trend keywords for each of the plurality of products based on the embedding data for each of the plurality of products and the embedding data for one or more trend keywords, and the calculated matching Based on the score, it may correspond to a model learned to extract a predetermined number of products from a plurality of products and output the extracted predetermined number of products as one or more products matching one or more trend keywords.

Additionally or alternatively, the product extraction model may include embedding data for learning trend keywords associated with the learning product family (e.g., topics of trends in the learning product family, phrases or sentences representing the topic of trends in the learning product family, etc.) and learning trend keywords associated with the learning product family. Learning using a triplet loss function that forces the matching score between the embedding data for products included in the learning trend keyword to be greater than the matching score between the embedding data for the learning trend keyword and the embedding data for products not included in the learning product group. It may correspond to a model that has been used. Here, the learning product group may refer to a product group actually created previously and/or a product group artificially created to learn a product extraction model.

In one embodiment, the processor outputs the generated product group through the user terminal, and in response to the user's selection of at least one product from the output product group, generates a product special exhibition including at least one product as a special product. can do.

FIG. 5 is a diagram illustrating an example of extracting a trend keyword 540 from data 510 associated with a plurality of product inquiries according to an embodiment of the present disclosure. Groups of users, including users with similar characteristics (e.g., age, occupation, interests, etc.), are grouped together at similar times (e.g., within a week, within a month, during certain seasons such as spring, or during certain periods such as holidays or anniversaries). Product queries searched for (period, etc.) may include the same or similar user intent (e.g., shopping intent, search intent) shared among the corresponding users. Product queries entered or searched according to the same or similar user intent may be clustered into one or more clusters, and users' product search or product shopping trends may be extracted from the clusters containing the same or similar user intent.

For example, a processor (e.g., at least one processor of an information processing system) generates one or more clusters containing product queries with the same or similar user intent (e.g., search intent, shopping intent, etc.); , trends of a specific group and/or a specific period can be extracted from one or more generated clusters. Here, the specific group may refer to the group of users themselves who entered or searched product queries included in one or more groups. Additionally or alternatively, a specific group may refer to the gender, age, occupation, etc. of a user who entered or searched multiple product queries included in one or more groups. Additionally, product queries may include meaningful queries (eg, refined, filtered, and/or preprocessed shopping queries) among queries entered or searched by users for shopping.

In one embodiment, the processor may generate a plurality of clusters 530 by clustering data 510 associated with a plurality of product queries into a plurality of clusters 530 . Then, the processor may extract one or more trend keywords 540 associated with at least one of a specific group or a specific period from one or more clusters among the plurality of clusters 530. For example, the processor extracts one or more trend keywords 540 from one or more clusters among the plurality of clusters 530, and features for one or more extracted trend keywords (e.g., a group that searches for the corresponding trend keyword in large numbers, Based on gender, age, time period, etc.), a specific group and/or specific period of time associated with one or more keywords 540 may be determined. That is, the processor may extract a trend associated with at least one of a specific group or a specific period based on the data 510 associated with the received plurality of product inquiries.

In addition, the processor uses the statistical topic embedding model 520 to cluster data 510 associated with a plurality of product queries into a plurality of clusters 530, and to select one or more trends from one or more clusters among the plurality of clusters 530. Keywords 540 can be extracted. Here, the statistical topic embedding model 520 may be a statistical or probability-based clustering model and/or an embedding model learned to cluster product queries with common shopping intent (or search intent, query intent) into one or more clusters. . As shown, the processor inputs data 510 associated with a plurality of product queries into the statistical subject embedding model 520, thereby generating a plurality of embedding data corresponding to each of the data 510 associated with the plurality of product queries. You can. The processor may cluster the generated plurality of embedding data into a plurality of clusters 530. Then, the processor may extract one or more trend keywords 540 from one or more of the plurality of clusters 530.

In one embodiment, the processor may use a statistical topic embedding model 520 to embed each product query into an embedding space based on data 510 associated with a plurality of product queries. Then, the processor may cluster the multiple product queries into a plurality of clusters 530 based on the distance between each product query in the embedding space. For example, the processor generates a cluster containing product queries that are close in distance based on a standard product query (e.g., a seed keyword) in the embedding space, thereby dividing the plurality of product queries into a plurality of clusters (530). ) can be clustered. Here, the standard product query may include a product query or keyword indicating a trend.

The processor may determine a reference product query to cluster the plurality of product queries into a plurality of clusters 530 . That is, the processor may determine a standard product query (or central product query) indicating a trend based on data 510 associated with a plurality of product queries. In one embodiment, the processor extracts popular keywords for a specific time of year (e.g., summer, winter, etc.) and/or for a specific group (e.g., a specific age, gender) based on data 510 associated with a plurality of product queries. And the extracted popular keywords can be determined as a standard product query. For example, the processor may discover keywords reflecting temporal characteristics based on data 510 associated with a plurality of product queries and determine them as standard product queries (e.g., seed keyword, main intent keyword). Here, the reference product query may include popular keywords corresponding to a specific point in time or period in the present or future, and accordingly, the processor can create a special exhibition corresponding to a specific point in time or period in the present or future desired by the operator. .

To determine baseline product inquiries, the processor organizes and analyzes time series data based on the inflow of product inquiries by gender/age group, and identifies products with high seasonality, trends, and inflow of product inquiries. You can select a query. In the case of product queries (or keywords) that are currently highly seasonal, the seasonal pattern may remain similar in the past and future, which are temporally adjacent to the present. Therefore, instead of time series data of product inquiries from a future point in time (e.g., 1 week later, 2 weeks later) where the inflow of product inquiries cannot be observed, the processor uses data from the same period in the past (e.g., the same date a year ago). , same season, etc.), the standard product query can be determined by analyzing time series data collected.

For example, the processor inputs valid product queries obtained through refining and/or filtering into the statistical topic embedding model 520, embeds each product query in the embedding space, and uses a standard product query (or keyword) (e.g. For example, a cluster containing product queries (or keywords) (e.g., 'maternity bag', 'newborn car seat', 'baby carrier', etc.) that are close to each other in the embedding space can be created around 'baby spacesuit'). You can. The processor then extracts one or more trend keywords 540 from the cluster and determines a specific population and/or specific time period associated with the cluster and/or the one or more trend keywords 540, thereby determining the specific population and/or Trends related to a specific period can be extracted.

One or more trend keywords 540 extracted from the cluster include keywords included in product queries included in the cluster, themes represented by the cluster, common characteristics of product queries contained in the cluster, the cluster, and It can include the associated user group, the time period associated with the group, etc. For example, one or more trend keywords 540 extracted from the 'baby spacesuit' cluster include 'baby bag' and 'newborn car seat', which are product queries (i.e., query sets of the cluster) included in the 'baby spacesuit' cluster. , 'baby carrier', etc. Additionally or alternatively, one or more trend keywords 540 extracted from the 'Baby Spacesuit' cluster may be selected from the topic of the 'Baby Spacesuit' cluster (e.g., common characteristics, categories, and , relevance, query intent, etc.), such as 'birth of a newborn', 'childcare', 'newborn', etc. Additionally or alternatively, one or more trend keywords 540 extracted from the 'baby spacesuit' cluster may include 'April', which is the time when many product inquiries included in the 'baby spacesuit' cluster were input. Additionally or alternatively, one or more trend keywords 540 extracted from the 'baby spacesuit' cluster include attributes related to age, gender, and marital status of a group of users who entered multiple product queries included in the 'baby spacesuit' cluster. May include people in their 30s, ‘women’, ‘married people’, etc. Based on one or more trend keywords 540 extracted from the cluster in this way, a trend topic (or special exhibition title) (for example, 'newborn childbirth theme') for the cluster may be created.

Data 510 associated with multiple product queries may correspond to various types of data. For example, data 510 associated with a plurality of product queries may include various types of data such as text data, vector data, etc.

FIG. 6 is a diagram illustrating an example of determining a title 630 of a special exhibition indicating the theme of a trend based on a trend keyword 610 according to an embodiment of the present disclosure. A processor (e.g., at least one processor of an information processing system) may determine a topic of a trend based on one or more trend keywords 610. In one embodiment, the processor may simply combine one or more trend keywords 610 to generate a phrase or sentence representing the topic of the trend. In another embodiment, the processor may use a language model to convert one or more trend keywords 610 into phrases or sentences representing the topic of the trend. Here, a phrase or sentence representing the theme of the trend can be used as the title of a special exhibition. For example, when creating a special exhibition, the processor may determine a title 630 of the special exhibition indicating the subject of the trend based on one or more trend keywords 610. To this end, the processor may include a special exhibition title generator (or sales event title generator) 620 that generates the title of the special exhibition.

In one embodiment, the special exhibition title generator 620 may simply combine one or more input trend keywords 610 to generate a special exhibition title 630 indicating the theme of the trend. In another embodiment, the special exhibition title generator 620 may use a language model to convert one or more trend keywords 610 into a title of a special exhibition indicating the topic of the trend. In this case, the special exhibition title generator 620 may include a language model, and the language model generates a phrase or sentence in which the intention of the input keyword (for example, a topic related to the keyword, etc.) is expressed based on the input keyword. It may be an artificial neural network model and/or a statistical model trained to do so. In other words, the language model may be a model learned to recognize the relationships and meanings between a plurality of input trend keywords and generate phrases or sentences representing the topic of the trend.

As shown, the special exhibition title generator 620 receives trend keywords 610 including 'rash guard', 'swimsuit', 'beachwear', and 'June 27' as input, and creates the title of the special exhibition 'Water'. ‘Park, wait!’ (630) can be created. In one embodiment, the special exhibition title generator 620 may receive one or more trend keywords 610 as input and generate a plurality of phrases or sentences. In this case, the processor may determine at least one of the plurality of phrases or sentences generated by the special exhibition title generator 620 as the title 630 of the special exhibition. For example, the processor outputs a plurality of phrases or sentences generated by the special exhibition title generation unit 620 as title candidates for the special exhibition through a user terminal (e.g., the terminal of the special exhibition operator), and outputs the plurality of sentences output. In response to a user's (for example, special exhibition operator) selection of at least one phrase or sentence, the selected at least one phrase or sentence may be determined as the title 630 of the special exhibition. Alternatively, the processor may receive a phrase or sentence directly entered by the user through the user terminal and determine it as the title of the special exhibition.

7 shows matching with a trend keyword 710 and/or a trend topic 720 among a plurality of products based on a trend keyword 710 and/or a trend topic 720 according to an embodiment of the present disclosure. This is a diagram showing an example of extracting one or more products. A processor (eg, at least one processor of an information processing system) may extract one or more trend keywords 710 and/or one or more products 740 associated with the topic 720 of the trend from the plurality of products. In one embodiment, the processor may use the product extraction model 730 to extract one or more products 740 that match one or more trend keywords 710 from a plurality of products. Additionally or alternatively, the processor may extract one or more products 740 associated with one or more trend keywords 710 from the plurality of products based on the topic 720 of the trend.

Here, the product extraction model 730 may be a model learned to output one or more products 740 that match the input trend keyword 710 and/or trend topic 720. For example, the product extraction model 730 is based on information about each of the plurality of products (e.g., product title, category, tag, introduction, review, image, etc.) for each of the plurality of products. Calculate a matching score (e.g., distance in the embedding space) with one or more trend keywords 710, extract a predetermined number of products from a plurality of products based on the calculated matching score, and extract the extracted predetermined number of products. It may correspond to a model learned to output a number of products as one or more products 740 that match one or more trend keywords 710.

As shown, the processor inputs a trend keyword 710 and/or a trend topic 720 into the product extraction model 730, thereby selecting one or more products 740 output from the product extraction model 730 among a plurality of products. ) can be created. Here, the trend keyword 710 and/or trend topic 720 input to the product extraction model 730 is a combination of trend keywords, a sentence or phrase containing a trend keyword, or a sentence or phrase representing the topic of the trend ( For example, it may include the title of a trend), the title of a special exhibition, etc. Additionally, the trend keyword 710 and/or trend topic 720 input to the product extraction model 730 may include information about consumer groups, periods, product inquiries, etc. related to the trend.

For example, as the processor inputs the title of the special exhibition corresponding to the theme 720 of the trend, ‘Wait for the water park’ into the product extraction model 730, the product extraction model 730 selects the special exhibition title among the plurality of products. Products with a high matching score for the title (i.e. products that match the title of the special exhibition) such as ‘swimsuit’, ‘swimming cap’, ‘aqua shoes’, ‘swimming bag’, and ‘water park season ticket’ can be printed. The processor may generate a product group including at least some of the plurality of products output in this way. In one embodiment, the processor outputs the generated product group through the user terminal, and in response to the user's selection of at least one product from the output product group, creates a product special exhibition including at least one product as a special product. You can. In another embodiment, the processor may automatically determine at least some of the plurality of output products as special products and create a special exhibition including the special products.

Figure 8 is a diagram showing an example of learning a product extraction model according to an embodiment of the present disclosure. The above-described product extraction model may include an embedding generation model learned to generate embedding data for each of a plurality of products and embedding data for one or more trend keywords (or topics of trends). Here, the embedding generation model may generate embedding data for each of a plurality of products based on information about each of the plurality of products being input. For example, the embedding generation model learns to generate embedding data for each of multiple products based on the product name, category, seller, tag, brand, manufacturer, product description, options, price, image, etc. It may be a model. In this case, the product extraction model is based on embedding data for each of a plurality of products and embedding data for one or more trend keywords (or topics of trends), and one or more trend keywords (or trends) for each of a plurality of products. calculate a matching score with a subject), extract a predetermined number of products from a plurality of products based on the calculated matching score, and select the extracted predetermined number of products as one or more trend keywords (or the topic of the trend) It may be an artificial neural network model and/or a statistical model learned to output as one or more products that match.

A processor (e.g., at least one processor of an information processing system) may receive data for a learning product group and/or receive information for each of a plurality of products to generate and/or learn a product extraction model. . Here, the data on the learning product group is data about the product group actually created previously (e.g., planned products from the product planning event that was actually executed) and/or data about the product group artificially created to learn the product extraction model. may include. Additionally, data on the product group may include trend keywords for the product group, trend topics, information on products included in the product group, consumer groups associated with the trend of the product group, and times associated with the trend of the product group. Additionally, data about the learning product group may be received from internal and/or external storage devices of the information processing system.

The processor selects one or more products that match the trend keyword (or topic of the trend) as the trend keyword (or topic of the trend) is input based on data about the learning product group and information about each of the plurality of products. You can learn a product extraction model to output. In one embodiment, as a trend keyword is input, the processor calculates a matching score with the trend keyword for each of a plurality of products, and based on the calculated matching score, a predetermined number (e.g., Learn a product extraction model to extract 10 products (for example, products with matching score rankings from 1st to 10th) and output a predetermined number of products as one or more products that match trend keywords. can do.

In one embodiment, the processor uses an embedding data generation model (e.g., a language model such as Bidirectional Encoder Representations from Transformers (BERT)) to generate learning trend keywords (or topics of learning trends) associated with a set of learning products. Embedding data for products, embedding data for products included in the learning product group, and embedding data for products not included in the learning product group can be generated. For example, the processor may generate embedding data for a learning trend keyword (or a topic of a learning trend) by inputting a learning trend keyword (or topic of a learning trend) of a learning product group into an embedding data generation model. , By entering products included in the learning product group (e.g., information about products included in the learning product group), embedding data for products included in the learning product group can be generated, and products not included in the learning product group ( For example, by entering information about products not included in the learning product group, embedding data for products not included in the learning product group can be generated. Afterwards, the processor determines the matching score between the embedding data for the learning trend keyword (or, the subject of the learning trend) and the embedding data for the products included in the learning product group, for the learning trend keyword (or, the subject of the learning trend). A product extraction model can be learned using a triplet loss function that forces the matching score between the embedding data and the embedding data for products not included in the learning product group to be greater than the matching score. Here, the matching score between embedding data may be inversely proportional to the distance between embedding data in the embedding space.

Additionally, in order to improve the performance of the product extraction model, a sampling process (e.g., negative sampling) may be performed to extract products that are not included in the learning product group as learning data. In one embodiment, the processor may sample products that contain some similar features to products included in the learning product group as learning data (i.e., negative data) from among a plurality of products not included in the learning product group. For example, when learning a product extraction model using the 'Razor' product group as a learning product group, the processor selects the same category as the razor (e.g., 'Digital Category') among a plurality of products not included in the 'Razor' product group. 'Hair dryer', a product included in , can be extracted as a 'product not included in the learning product group' for learning a product extraction model. In this case, the product extraction model can be learned to extract products included in the product group by considering more detailed characteristics of the products.

As shown, the processor generates a learning trend keyword A (also referred to as 'anchor') 812, a product B included in the learning product family (i.e., a product in the learning product family) (also referred to as 'positive') 814, and a learning product group 814. Product C (i.e., any product that is not a product of the learning product group) (also referred to as 'negative') 816 that is not included in Embedding data for each C can be generated. Then, the processor generates the embedding data 822 and The product extraction model can be learned in a direction in which the distance between the embedding data 826 for product C increases.

Figure 9 is a diagram illustrating an example in which a user (eg, an operator) creates a product promotion according to an embodiment of the present disclosure. Based on the product group created according to some embodiments of the present disclosure, the information processing system may create one or more special exhibitions. For example, the information processing system can create a special exhibition in which at least some of the products included in the generated product group are special products. In one embodiment, the information processing system uses the trend keywords, extracted/generated according to some embodiments of the present disclosure, as trend keyword candidates 914 for the special exhibition, and the trending topic as the topic candidate for the special exhibition (or, the special exhibition target consumer candidate). ) 912 and/or as a title candidate 916 for the special exhibition, the product group may be provided as a special product candidate 922 to the operator of the special exhibition. In this case, the information processing system may include a subject candidate 912 for the special exhibition (or a candidate for the special exhibition target consumer), a trend keyword candidate for the special exhibition 914, a title candidate for the special exhibition 916, and/or a special product candidate to be included in the special exhibition ( In response to the operator's selection of at least one of 922), one or more product special exhibitions may be created.

A user terminal (eg, at least one processor of the user terminal) may provide a user (eg, operator) a user interface for creating a product promotion. In one embodiment, the user terminal may display the special exhibition topic candidate 912 on the display. For example, the user terminal may display candidates for a specific group and/or a specific time period targeted by the special exhibition on the display. Afterwards, the user terminal may receive a user input for selecting at least one of the special exhibition topic candidates 912 and transmit a request and/or data related thereto to the information processing system.

In one embodiment, the information processing system selects one from one or more clusters associated with a user-selected promotional topic (i.e., a specific group and/or specific time period selected by the user) from among a plurality of clusters generated by clustering a plurality of product queries. The above trend keywords can be extracted and output as trend keyword candidates 914 through the user terminal. Here, the one or more clusters may be a cluster containing product queries entered in large numbers by a specific group selected by the user and/or a cluster containing product queries entered in large numbers at a specific time. The user terminal may display trend keyword candidates 914 on the display, receive a user input for selecting at least one trend keyword from among the trend keyword candidates 914, and send a request and/or data related thereto to the information processing system. It can be sent to .

In one embodiment, the information processing system may generate one or more sentences based on at least one trend keyword selected by the user and output them through the user terminal as a title candidate 916 for a special exhibition. For example, the information processing system may generate one or more phrases or sentences based on at least one trend keyword selected by the user, as well as keywords associated with the topic of the special exhibition selected by the user, keywords associated with one or more clusters, and/or keywords similar to the trend keyword. can be generated and output through the user terminal as a title candidate 916 for a special exhibition. The user terminal may display title candidates 916 for a special exhibition on a display, receive a user input for selecting at least one title candidate 916 for a special exhibition, and send a request and/or data related thereto to the information processing system. Can be transmitted.

In one embodiment, the information processing system extracts at least one product that matches the title of the special exhibition selected by the user from among a plurality of products, and presents a list containing the extracted at least one product as a special product candidate 922 to the user. It can be printed through the terminal. Accordingly, the user terminal can display a list containing at least one product (i.e., a planned product candidate) on the display, and the user selects at least one product from the list (i.e., a planned product candidate) as a planned product. Input may be received and associated requests and/or data transmitted to an information processing system. Additionally, the information processing system may receive a request for creating and/or issuing a product promotion from the user terminal.

As described above, the information processing system may generate and/or issue a product exhibition including the special exhibition theme, trend keyword, title of the special exhibition, and/or special product selected by the user based on the received request and/or data. You can. Additionally or alternatively, the user may directly input the special exhibition theme, trend keyword, title of the special exhibition, and/or special product without selecting it from the candidate group. In this case, that is, the information processing system may create and/or issue a special product exhibition based on the special exhibition topic, trend keyword, title of the special exhibition, and/or special product input by the user. The information processing system can provide product promotions to consumers by creating and publishing product promotions.

As shown in the first operation 910, the user terminal is a special exhibition topic candidate 912, a period candidate ('this week', 'next week', 'in 2 weeks', 'in 3 weeks') and/ or group candidates ('early teens', 'late teens', 'early 20s', '30s', '40s', '50s', '60s', 'male', 'female') can be output through the user interface. In response to a user input selecting at least one of the printed special exhibition topic candidates 912 (e.g., '2 weeks later', '40s', 'women'), the user terminal selects the selected period and/or group and Trend keyword candidates extracted from the associated first cluster ('home decoration', 'lighting', 'curtain', 'interior') 914 and trend keyword candidates extracted from the second cluster ('hoodie', 'hood') , 'winter clothes') (914) can be output through the user interface. Here, the first cluster and/or the second cluster may be a cluster associated with product queries entered in large numbers during the selected period and/or a cluster associated with product queries entered in large numbers by the selected group.

The user terminal responds to a user input for selecting at least one of the output trend keyword candidates 914 ('home decoration', 'lighting', and 'interior'), and selects a special exhibition title candidate ( ‘How about decorating your home with lighting?’, ‘Interior items for this winter’) (916) can be printed through the user interface. In response to a user input for selecting at least one of the title candidates 916 for the special exhibition output in the first operation 910 ('How about decorating your home with lighting?'), the user terminal is displayed in the second operation 920. As shown, a special product candidate 922 including one or more products ('A stand (floor lamp)', 'B stand (table lamp)', 'C stand (table lamp') matching the title of the selected special exhibition. can be output. Based on a user input for selecting at least one product (‘A stand (floor lamp)’) from among the output special product candidates 922, a product special exhibition including the selected at least one product may be created. Additionally, the corresponding product promotion may be issued and provided to consumers according to a user input requesting the issuance of the created product promotion.

FIG. 10 is a diagram illustrating an example in which a user (eg, an operator) is provided with one or more product promotion candidates 1000 according to an embodiment of the present disclosure. Based on the product group created according to some embodiments of the present disclosure, the information processing system may create one or more special exhibitions. That is, according to the above-described embodiments, the information processing system can extract a trend and create a product promotion related to the extracted trend. For example, the information processing system may create one or more special exhibitions in which at least some of the products from the generated product group are special products. In this case, the information processing system may determine a phrase or sentence representing the trend topic as the title of the special exhibition. The information processing system may provide one or more product promotions created as product promotion candidates 1000 to the user through the user terminal. The user terminal may display information about one or more product promotions corresponding to the product promotion candidate 1000 on the display.

As shown, the user terminal can output 'winter health care food', 'winter school uniform styling', 'prepare New Year's gifts in advance', and 'let's go to the ski resort' as product promotion candidates 1000 through the user interface. For example, the user terminal is the target consumer for each product promotion candidate (e.g., '57-63 year old male', '18 year old female', '45-55 year old female', '20-30 male') , period, image, hashtags ('#health', '#elderly', '#student', etc.), title of special exhibition, etc. can be printed through the user interface. Additionally, the user terminal may display the detailed configuration of the selected product promotion on the display in response to a user input for selecting at least one product promotion from among the product promotion candidates 1000.

In one embodiment, the information processing system provides automatically generated information about one or more product promotions to the operator of the product promotion, and in response to the operator's selection of at least one of the one or more product promotions, selects the selected product promotion. can be issued. Alternatively, the information processing system may immediately publish one or more automatically generated product promotions. As the user selects at least one product promotion among the product promotion candidates 1000 and performs a request to issue the selected product promotion, the corresponding product promotion may be created and issued and provided to consumers. In Figure 10, the user terminal outputs four product promotion candidates, but the present invention is not limited to this and may output a different number of product promotion candidates.

The above-described method may be provided as a computer program stored in a computer-readable recording medium for execution on a computer. The medium may continuously store a computer-executable program, or may temporarily store it for execution or download. In addition, the medium may be a variety of recording or storage means in the form of a single or several pieces of hardware combined. It is not limited to a medium directly connected to a computer system and may be distributed over a network. Examples of media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, And there may be something configured to store program instructions, including ROM, RAM, flash memory, etc. Additionally, examples of other media include recording or storage media managed by app stores that distribute applications, sites or servers that supply or distribute various other software, etc.

The methods, operations, or techniques of this disclosure may be implemented by various means. For example, these techniques may be implemented in hardware, firmware, software, or a combination thereof. Those skilled in the art will understand that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the disclosure herein may be implemented in electronic hardware, computer software, or combinations of both. To clearly illustrate this interchange of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and design requirements imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementations should not be interpreted as causing a departure from the scope of the present disclosure.

In a hardware implementation, the processing units used to perform the techniques may include one or more ASICs, DSPs, digital signal processing devices (DSPDs), programmable logic devices (PLDs). ), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, electronic devices, and other electronic units designed to perform the functions described in this disclosure. , a computer, or a combination thereof.

Accordingly, the various illustrative logical blocks, modules, and circuits described in connection with this disclosure may be general-purpose processors, DSPs, ASICs, FPGAs or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or It may be implemented or performed as any combination of those designed to perform the functions described in. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, such as a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other configuration.

For firmware and/or software implementations, techniques include random access memory (RAM), read-only memory (ROM), non-volatile random access memory (NVRAM), and PROM ( on computer-readable media such as programmable read-only memory (EPROM), electrically erasable PROM (EEPROM), flash memory, compact disc (CD), magnetic or optical data storage devices, etc. It may also be implemented as stored instructions. Instructions may be executable by one or more processors and may cause the processor(s) to perform certain aspects of the functionality described in this disclosure.

Although the above-described embodiments have been described as utilizing aspects of the presently disclosed subject matter in one or more standalone computer systems, the disclosure is not limited thereto and may also be implemented in conjunction with any computing environment, such as a network or distributed computing environment. . Furthermore, aspects of the subject matter of this disclosure may be implemented in multiple processing chips or devices, and storage may be similarly effected across the multiple devices. These devices may include PCs, network servers, and portable devices.

Although the present disclosure has been described in relation to some embodiments in this specification, various modifications and changes may be made without departing from the scope of the present disclosure as can be understood by a person skilled in the art to which the invention pertains. Additionally, such modifications and changes should be considered to fall within the scope of the claims appended hereto.

100: User
110: user terminal
120: screen

Claims (20)

In a method for generating a product group based on user intent of a search query performed by at least one computing device,
Receiving data associated with a plurality of product queries;
extracting a trend associated with at least one of a specific group or a specific period based on data associated with the plurality of product inquiries; and
Creating a product group including one or more products related to the extracted trend
Including,
The step of creating a product group including one or more products related to the extracted trend,
Comprising a step of extracting one or more products related to one or more trend keywords from a plurality of products,
The step of extracting one or more products related to one or more trend keywords from the plurality of products includes,
Using a product extraction model, extracting one or more products matching the one or more trend keywords from the plurality of products,
The product extraction model is,
The matching score between the embedding data for the learning trend keyword associated with the learning product group and the embedding data for the product included in the learning product group is the embedding data for the learning trend keyword and the embedding data for the product not included in the learning product group. A product group generation method learned using a triplet loss function that forces the matching score to be greater than the matching score between them.
According to paragraph 1,
The step of extracting a trend associated with at least one of a specific group or a specific period based on data associated with the plurality of product inquiries includes:
Clustering data associated with the plurality of product queries into a plurality of clusters; and
Extracting one or more trend keywords associated with at least one of the specific group or specific period from one or more clusters among the plurality of clusters.
Method for creating a product group, including.
According to paragraph 2,
The step of clustering the data associated with the plurality of product queries into a plurality of clusters,
A product group generation method comprising clustering data associated with the plurality of product queries into a plurality of clusters using a statistical topic embedding model.
According to paragraph 2,
The step of creating a product group including one or more products related to the extracted trend,
Creating a product group including the extracted one or more products
Method for creating a product group, including.
According to clause 4,
The step of extracting one or more products related to the one or more trend keywords from the plurality of products includes,
Based on the one or more trend keywords, determining a topic of the trend
Method for creating a product group, including.
According to clause 5,
The step of determining the topic of the trend based on the one or more trend keywords,
A product group generation method comprising converting the one or more trend keywords into a phrase or sentence representing the subject of the trend using a language model.
delete According to paragraph 1,
The product extraction model calculates a matching score with the one or more trend keywords for each of the plurality of products, based on information about each of the plurality of products, and based on the calculated matching score, A product group generation method learned to extract a predetermined number of products from among products and output the extracted predetermined number of products as one or more products matching the one or more trend keywords.
According to clause 8,
The product extraction model includes an embedding generation model learned to generate embedding data for each of the plurality of products and embedding data for the one or more trend keywords,
The product extraction model calculates a matching score with the one or more trend keywords for each of the plurality of products based on the embedding data for each of the plurality of products and the embedding data for the one or more trend keywords, and the calculation Based on the matching score, a product group generation method is learned to extract a predetermined number of products from the plurality of products and output the extracted predetermined number of products as one or more products matching the one or more trend keywords. .
delete According to clause 4,
Outputting the generated product group through a user terminal; and
In response to the user's selection of at least one product from the output product group, generating a product special exhibition including the at least one product as a special product.
A method for creating a product group, further comprising:
In a method of learning a product extraction model performed by at least one computing device,
Receiving data about a learning product group;
Receiving information about each of a plurality of products; and
Based on data on the learning product group and information on each of the plurality of products, as a trend keyword is input, learning a product extraction model to output one or more products matching the trend keyword,
The matching score between the embedding data for the learning trend keyword and the embedding data for the product included in the learning product group is the matching score between the embedding data for the learning trend keyword and the embedding data for the product not included in the learning product group. A method of learning a product extraction model, comprising the step of learning a product extraction model using a triplet loss function that is forced to become larger.
According to clause 12,
The step of learning a product extraction model to output one or more products matching the trend keyword as a trend keyword is input based on data about the learning product group and information about each of the plurality of products,
As the trend keyword is input, a matching score with the trend keyword for each of the plurality of products is calculated, and based on the calculated matching score, a predetermined number of products are extracted from the plurality of products, A product extraction model learning method comprising the step of learning the product extraction model to output a predetermined number of products extracted as one or more products matching the trend keyword.
According to clause 12,
Using an embedding data generation model, generating embedding data for learning trend keywords associated with the learning product group, embedding data for products included in the learning product group, and embedding data for products not included in the learning product group. Further including, product extraction model learning method.
delete A computer program stored in a recording medium for executing the method according to any one of claims 1 to 6, 8, 9 and 11 to 14 on a computer.
As an information processing system,
communication module;
Memory; and
At least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory
Including,
The at least one program is,
Receive data associated with a plurality of product queries,
Based on the data associated with the plurality of product inquiries, extract trends associated with at least one of a specific group or a specific period,
Contains instructions for creating a product group including one or more products related to the extracted trend,
Creating a product group containing one or more products related to the extracted trend includes,
Including extracting one or more products associated with one or more trend keywords from a plurality of products,
Extracting one or more products related to one or more trend keywords from the plurality of products,
Using a product extraction model, extracting one or more products matching the one or more trend keywords from the plurality of products,
The product extraction model is,
The matching score between the embedding data for the learning trend keyword associated with the learning product group and the embedding data for the product included in the learning product group is the embedding data for the learning trend keyword and the embedding data for the product not included in the learning product group. An information processing system trained using a triplet loss function that forces the matching scores between
According to clause 17,
The at least one program is,
Further comprising instructions for clustering data associated with the plurality of product queries into a plurality of clusters and extracting one or more trend keywords associated with at least one of the specific group or a specific period from one or more of the plurality of clusters, Information processing system.
As an information processing system,
communication module;
Memory; and
At least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory
Including,
The at least one program is,
Receive data about the learning product range,
Receive information about each of a plurality of products,
Based on data on the learning product group and information on each of the plurality of products, as a trend keyword is input, it includes instructions for learning a product extraction model to output one or more products matching the trend keyword,
The matching score between the embedding data for the learning trend keyword and the embedding data for the product included in the learning product group is the matching score between the embedding data for the learning trend keyword and the embedding data for the product not included in the learning product group. An information processing system that learns a product extraction model using a triplet loss function that forces it to become larger.
According to clause 19,
The at least one program is,
As the trend keyword is input, a matching score with the trend keyword for each of the plurality of products is calculated, and based on the calculated matching score, a predetermined number of products are extracted from the plurality of products, An information processing system further comprising instructions for learning the product extraction model to output a predetermined number of products extracted as one or more products matching the trend keyword.

Images