Method for operating electronic apparatus for supporting selection of influencer suitable for promotion of target product

Abstract

A method of operating an electronic device is disclosed. The method of operating an electronic device according to the present disclosure comprises: receiving product information including at least one of a product name, product type, and product function of a target product through user input received from a user terminal; obtaining follower information including at least one of an age group, gender, and nationality of followers of each of a plurality of influencers active on at least one social platform; calculating a suitability degree indicating the suitability for promotion of the target product for each of the plurality of influencers based on the follower information and product information obtained for each influencer; and selecting at least one influencer among the plurality of influencers based on the calculated suitability degree.

Description

Technology Field

[0001] The present disclosure relates to influencer marketing support technology using artificial intelligence technology, and more specifically, to a method of operation of an electronic device that evaluates and selects influencers based on product information, follower information, etc., and generates email content and promotional scenarios using a generative model. Background Technology

[0002] In the recent digital marketing environment, influencer marketing based on social platforms is rapidly expanding. Influencers can effectively deliver brand messages through trust built with their followers, allowing for higher engagement and conversion rates compared to traditional advertising methods. Consequently, there is a growing demand among companies and brands for technologies that enable them to select influencers suitable for their product characteristics and target consumer bases, as well as to produce customized promotional content.

[0003] Traditionally, the process of selecting influencers often relied on limited metrics such as follower counts and simple view counts. However, this approach had limitations; it failed to adequately reflect an influencer's actual suitability for promotion and did not consider issues such as fatigue from repetitive content production and a decline in long-term engagement. Furthermore, because individual characteristics of influencers were not reflected during campaign execution, collaborations often resulted in uniform and ineffective outcomes.

[0004] Therefore, there is a need for new technological means capable of comprehensively analyzing various factors such as product information, follower information, engagement information, and sales history data, predicting the suitability and sales contribution of influencers based on this analysis, and further utilizing generative artificial intelligence to automatically generate customized email content and promotional scenarios. This disclosure was conceived against this backdrop. Prior art literature

[0005] Registered Patent Publication No. 10-2246387 The problem to be solved

[0006] The present disclosure aims to solve the problems of the prior art described above by comprehensively analyzing product information, follower information, engagement information, sales history, etc., to quantitatively evaluate the promotional suitability of an influencer, and to provide an objective and highly reliable influencer selection technology based on this.

[0007] Furthermore, the present disclosure aims to provide an influencer selection technology that considers long-term marketing effects and potential for actual sales contribution, rather than a limited evaluation based on simple follower counts or views, by integrally utilizing multifaceted performance indicators such as influencer activity persistence, overlap, sales forecasting indicators, and diffusion speed indicators.

[0008] Furthermore, the present disclosure provides a method of operation for an electronic device that supports companies and brands in efficiently designing and executing marketing campaigns by utilizing a generative artificial intelligence model to automatically generate product collaboration request email content and promotional scenarios, thereby enabling the simultaneous achievement of consumer-tailored promotional effects and brand image enhancement effects.

[0009] The purposes of the present disclosure are not limited to those mentioned above, and other purposes and advantages of the present disclosure not mentioned may be understood from the following description and will be more clearly understood from the embodiments of the present disclosure. Furthermore, it will be readily apparent that the purposes and advantages of the present disclosure can be realized by the means and combinations thereof set forth in the claims. means of solving the problem

[0010] A method of operating an electronic device according to the present embodiment comprises: receiving product information including at least one of a product name, product type, and product function of a target product through user input received from a user terminal; obtaining follower information including at least one of an age group, gender, and nationality of followers of each of a plurality of influencers active on at least one social platform; calculating a suitability degree indicating the suitability for promotion of the target product for each of the plurality of influencers based on the follower information and product information obtained for each influencer; and selecting at least one influencer among the plurality of influencers based on the calculated suitability degree.

[0011] The method of operating the electronic device comprises the steps of obtaining script information of at least one past content provided by each of the plurality of influencers to the social platform, and obtaining influencer characteristic information for each of the plurality of influencers based on the follower information and the script information, and the step of calculating the suitability may calculate the suitability for each of the plurality of influencers based on the influencer characteristic information and the product information.

[0012] The method of operating the electronic device described above includes the step of obtaining engagement information including the number of views and interest of followers regarding at least one past content provided by each of the plurality of influencers, and the step of calculating the suitability may include the step of calculating the suitability for each of the plurality of influencers based on the follower information and the product information, and the step of obtaining an activity duration indicating the persistence of the followers' engagement information by time interval during a preset first period based on the engagement information.

[0013] The method of operation of the electronic device described above includes the steps of: obtaining script information for a plurality of past contents uploaded by the plurality of influencers during a second period set by the plurality of influencers; obtaining a similarity between the plurality of past contents based on the script information; and identifying a similar content group in which the similarity is greater than or equal to a preset threshold. The step of calculating the suitability may include the step of calculating the suitability for each of the plurality of influencers based on the follower information and the product information; and the step of obtaining a redundancy indicating the degree of repetition of the past contents based on the frequency with which the similar content group was uploaded within the second period set by the similar content group and the number of similar contents included in the similar content group.

[0014] The method of operating the electronic device described above may include the step of identifying sales history information for each of the plurality of influencers based on at least one of a commerce platform and a social platform, and the step of calculating the suitability, which may include the step of calculating the suitability for each of the plurality of influencers based on follower information and product information, and the step of obtaining a sales forecast indicator based on the sales history information when the sales history information is identified.

[0015] The step of calculating the suitability may include the step of calculating the suitability for each of the plurality of influencers based on the follower information and the product information, and, if the sales history information is not identified, the step of obtaining the sales forecast indicator by referring to a pre-set forecast indicator based on the follower information of the plurality of influencers.

[0016] The step of calculating the suitability may include: a step of calculating the suitability for each of the plurality of influencers based on the follower information and the product information; a step of obtaining the growth rate of the engagement information for each of the past content during a preset third period including the upload time; a step of obtaining a unit diffusion index representing the diffusion speed of each of the past content based on the growth rate; and a step of obtaining a diffusion speed index based on the unit diffusion index.

[0017] The method of operation of the electronic device may further include the steps of generating email content regarding a product introduction and a collaboration request through a generative model based on the script information and product information of the selected influencer, and sending the generated email content to the email address of the selected influencer.

[0018] The method of operation of the electronic device comprises the steps of: obtaining promotional plan information including at least one of a promotional budget, a promotional concept, and content volume information for the promotion of the target product in a generative model for generating a scenario; and inputting the promotional plan information, the product information, and the follower information to obtain a promotional scenario for target content for the promotion of the target product by the selected influencer, wherein the promotional scenario may include at least one of progress sequence information, content volume information, and concept information.

[0019] The method of operation of the electronic device described above includes a step of determining the contribution of the target content to the sales of the target product when the target content for promoting the target product of the selected influencer is uploaded to the social platform, and the step of determining the contribution to the sales of the target product may determine the contribution based on engagement information regarding the target content and sales information of the target product confirmed by time intervals within a certain period including the time of upload of the target content. Effects of the invention

[0020] The method of operation of an electronic device according to the present disclosure receives various data (product information, follower information, engagement logs, sales history) and calculates the suitability, activity persistence, overlap, sales prediction indicators, and diffusion speed indicators of an influencer, thereby enabling the objective and reliable selection of an influencer optimized for the promotion of a target product. Through this, companies can conduct influencer marketing with higher efficiency and performance compared to existing methods.

[0021] Furthermore, according to the present disclosure, influencers who produce only repetitive and similar content can be excluded, and influencers with long-term engagement and rapid diffusion effects can be identified, thereby ensuring content diversity and novelty. This has the effect of significantly contributing to the sustainability of marketing effects and the maintenance of consumer interest.

[0022] Furthermore, since the present disclosure enables the automatic generation of collaboration request email content and promotional scenarios through a generative AI-based model, companies can efficiently deliver customized messages to each influencer and produce individually differentiated promotional content while maintaining a consistent brand image. Consequently, not only is the efficiency of marketing operations significantly improved, but positive effects on collaboration acceptance rates and product sales growth can also be achieved. Brief explanation of the drawing

[0023] Aspects, features, and advantages of specific embodiments of the present disclosure will become more apparent from the following description with reference to the accompanying drawings. FIG. 1 is a diagram illustrating an operation in which an electronic device provides a function to select an influencer based on suitability according to one embodiment of the present disclosure. FIG. 2a is a block diagram illustrating the configuration of an electronic device according to one embodiment of the present disclosure. FIG. 2b is a block diagram illustrating the detailed configuration of an electronic device according to one embodiment of the present disclosure. FIG. 3 is a flowchart illustrating a series of flows for an electronic device to select an influencer based on suitability according to one embodiment of the present disclosure. FIG. 4 is a diagram illustrating the operation of an electronic device providing a list of influencers with high suitability and a chatbot function according to one embodiment of the present disclosure. FIG. 5 is a diagram illustrating the operation of an electronic device providing a search function according to one embodiment of the present disclosure. FIG. 6 is a diagram illustrating the operation of an electronic device generating mail content and sending it to the mail address of a selected influencer according to one embodiment of the present disclosure. FIG. 7 is a diagram illustrating the operation of an electronic device generating a promotional scenario according to one embodiment of the present disclosure. FIG. 8 is a diagram illustrating the operation of an electronic device according to one embodiment of the present disclosure to determine the contribution of an influencer to target content. Specific details for implementing the invention

[0024] Before specifically describing the present disclosure, the method of description in the specification and drawings is described.

[0025] First, the terms used in this specification and claims have been selected based on general terms considering their functions in the various embodiments of this disclosure. However, these terms may vary depending on the intent of those skilled in the art, legal or technical interpretations, and the emergence of new technologies. Additionally, some terms have been arbitrarily selected by the applicant. Such terms may be interpreted according to the meanings defined in this specification; in the absence of specific definitions, they may be interpreted based on the overall content of this specification and common technical knowledge in the relevant field.

[0026] In addition, the same reference numbers or symbols described in each drawing attached to this specification represent parts or components that perform substantially the same function. For convenience of explanation and understanding, the same reference numbers or symbols are used to describe different embodiments. That is, even if components having the same reference number are all depicted in multiple drawings, the multiple drawings do not imply a single embodiment.

[0027] Additionally, in this specification and claims, terms including ordinal numbers, such as "first," "second," etc., may be used to distinguish between components. These ordinal numbers are used to distinguish identical or similar components from one another, and the meaning of the terms should not be limited by the use of such ordinal numbers. For example, the order of use or arrangement of components combined with such ordinal numbers should not be restricted by the number. If necessary, each ordinal number may be used interchangeably.

[0028] In this specification, singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "consisting of" are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0029] In the embodiments of the present disclosure, terms such as "module," "unit," "part," etc. are used to refer to a component that performs at least one function or operation, and such component may be implemented in hardware or software, or in a combination of hardware and software. Additionally, a plurality of "modules," "units," "parts," etc. may be integrated into at least one module or chip and implemented as at least one processor, except where each needs to be implemented in specific individual hardware.

[0030] Furthermore, in the embodiments of the present disclosure, when a part is described as being connected to another part, this includes not only a direct connection but also an indirect connection through another medium. Additionally, the meaning that a part includes a certain component implies that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0031] In the present disclosure, the expression “at least one of a, b, or c” may refer to “a”, “b”, “c”, “a and b”, “a and c”, “b and c”, “a, b, and c all”, or variations thereof.

[0032] The terms used in this disclosure have been selected to be as widely used and general as possible, taking into account their functions within this disclosure; however, these terms may vary depending on the intent of those skilled in the art, case law, the emergence of new technologies, etc. Additionally, in specific cases, terms have been selected at the applicant's discretion, and in such cases, their meanings will be described in detail in the relevant explanatory sections. Therefore, terms used in this disclosure should be defined not merely by their names, but based on their meanings and the overall content of this disclosure.

[0033] Singular expressions may include plural expressions unless the context clearly indicates otherwise. Terms used herein, including technical or scientific terms, may have the same meaning as generally understood by those skilled in the art as described in this specification. Additionally, terms including ordinal numbers, such as "first" or "second," used in this specification may be used to describe various components, but said components should not be limited by said terms. Such terms are used solely for the purpose of distinguishing one component from another.

[0034] When a part of a specification is described as "comprising" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components. Furthermore, terms such as "part" or "module" as used in the specification refer to a unit that processes at least one function or operation, and this may be implemented in hardware or software, or as a combination of hardware and software.

[0035] Embodiments of the present disclosure are described below with reference to the attached drawings so that those skilled in the art can easily implement the invention. However, the present disclosure may be embodied in various different forms and is not limited to the embodiments described herein. Furthermore, in order to clearly explain the present disclosure in the drawings, parts unrelated to the explanation have been omitted, and similar parts throughout the specification are denoted by similar reference numerals.

[0036] FIG. 1 is a diagram illustrating an operation in which an electronic device provides a function to select an influencer based on suitability according to one embodiment of the present disclosure.

[0037] Referring to FIG. 1, the electronic device (100) can receive product information (101) about a target product from a user terminal (200).

[0038] The electronic device (100) may be a device of a service provider that provides a service for selecting an influencer suitable for promoting a target product.

[0039] Product information (101) may include various information such as the product name, product type (e.g., cosmetics, food, clothing, etc.), provider information (e.g., brand name), main ingredients, product function, price, specifications, precautions, expiration date, etc. of the target product.

[0040] In addition, product information (101) may include images, descriptive text, videos, etc. of the target product.

[0041] In addition, product information (101) may further include target consumers (e.g., age group, gender, region, hobbies, etc.), brand positioning (e.g., premium, mass market, MZ target, etc.), marketing objectives (e.g., increasing brand awareness, inducing purchase conversion, penetrating specific targets, etc.).

[0042] Target products may include products belonging to the beauty industry, such as cosmetics, skincare, and makeup products, but may also include, but are not limited to, food and beverages, health functional foods, fashion accessories, clothing, electronic devices, household goods, home appliances, hobby items, exercise equipment, baby products, pet supplies, travel-related products, etc.

[0043] A user terminal (200) is a device for receiving product information (101) and performing ~~, and may be a smartphone, tablet PC, laptop PC, desktop PC, wearable device (e.g., smart watch), smart TV, set-top box, console, a terminal dedicated to virtual reality (VR) and augmented reality (AR) (e.g., HMD (Head Mounted Device), smart home device, etc., but is not limited thereto. The user terminal can perform one or more functions and operations, and may be implemented as a single device or distributed across multiple devices so that each function and operation is implemented.

[0044] The electronic device (100) can obtain follower information (102) including at least one of the age group, gender, and nationality of the followers of each of the multiple influencers active on at least one social platform.

[0045] An influencer refers to an individual or group that creates and posts content based on social platforms, possesses a certain number of followers, and can influence the consumption patterns or interests of their followers.

[0046] A social platform refers to an online-based service that allows users to create and post content in various formats, such as text, images, and videos, and to communicate with other users or share information.

[0047] Social platforms may include, but are not limited to, social network services (SNS) such as Instagram, YouTube, TikTok, Facebook, Twitter, and Snapchat, as well as community-based content platforms such as Naver Blog, Brunch, and Reddit, live broadcasting platforms such as Twitch and AfreecaTV, and review-centric platforms such as Hwahae, Glowpick, and Olive Young review services.

[0048] In this case, a follower refers to a user who can access content posted by an influencer or receive real-time updates by subscribing to or following the account of a specific influencer on a social platform.

[0049] The electronic device (100) can calculate a suitability rating indicating the suitability of promotion for a target product for each of a plurality of influencers and select an influencer based on this (S130).

[0050] For example, the electronic device (100) can analyze past content posted by influencer A with the description “moisture cream effective for dry skin” and determine that there is a high correlation with the product information (101) of the target product corresponding to “moisture cream” and calculate a suitability score of 90 points.

[0051] On the other hand, if influencer B’s past content is about “sunscreen,” the relevance to the product information (101) of the target product corresponding to “moisturizing cream” can be calculated as 82 points, which is relatively lower than that of influencer A.

[0052] In this case, the electronic device (100) can select an influencer A with a higher suitability.

[0053] Thus, the electronic device (100) according to the present disclosure calculates a suitability indicating the suitability for promotion of a target product based on the influencer's past content and product information (101) of the target product, and selects an influencer with a high suitability, thereby objectively selecting an influencer with a high correlation to the target product, and thus has the effect of expecting an improvement in the promotional effect.

[0054] Hereinafter, embodiments of the present disclosure will be described in more detail with reference to the attached drawings.

[0055] FIG. 2a is a block diagram illustrating the configuration of an electronic device according to one embodiment of the present disclosure.

[0056] Referring to FIG. 2a, the electronic device (100) may include a memory (110), a communication interface (120), and a processor (130).

[0057] The electronic device (100) may be implemented as a device or system implemented with at least one computer. In one embodiment, the electronic device (100) may be implemented as a server that communicates with various user terminals (200) via a web page or application to provide a service including an influencer selection function. However, it is also possible to implement it as a terminal device such as a smartphone, tablet PC, or desktop PC.

[0058] When the electronic device (100) is a server, the electronic device may be an FTP server (File Transfer Protocol Server), a web server, a database server, or a cloud server, but is not limited thereto. The server may perform one or more functions or operations, and the electronic device (100) may be implemented as a single device, but may also be implemented as a set or combination of two or more devices.

[0059] The memory (110) stores various programs or data temporarily or non-temporarily and transmits the stored information to the processor (130) upon the processor's call. Additionally, the memory (110) can store various information required for the processor's calculation, processing, or control operations in an electronic format.

[0060] The memory (110) may include, for example, at least one of a main memory and an auxiliary memory. The main memory may be implemented using a semiconductor storage medium such as ROM and / or RAM.

[0061] The memory (110) can store at least one artificial intelligence model, output data obtained based on input data input to the artificial intelligence model, and usage information obtained based on the output data.

[0062] Specifically, the memory (110) can store various data such as product information (101), follower information (102), and suitability for each influencer. In particular, the memory (110) can store at least one artificial intelligence model used for influencer selection (e.g., suitability calculation model, influencer selection model, generative model, etc.).

[0063] However, such artificial intelligence models or related data may be stored inside the electronic device (100), or may be utilized by the electronic device (100) while stored on an external AI server or cloud server capable of communicating with the electronic device (100).

[0064] Referring to FIG. 2a, the memory (110) can store a fit calculation model (111) and an influencer selection model (116).

[0065] The suitability calculation model (111) is a model for calculating the suitability of an influencer for a target product by receiving follower information and product information (101) as input, and may correspond to an artificial intelligence model implemented as an ensemble-based classification and regression model such as Random Forest or LightGBM, or a deep learning-based recommendation model (e.g., neural network recommendation model, Neural Collaborative Filtering).

[0066] The fit calculation model (111) may be a model trained using supervised learning, collaborative filtering-based learning methods, etc., with follower information (age, gender, region), engagement information (views, likes and comments history), product category characteristic information (product type, function, target customer base), and influencer past sponsorship performance data as training data.

[0067] The influencer selection model (116) is a model for selecting at least one influencer among multiple influencers by receiving performance indicators such as suitability, activity duration, overlap, sales forecast, and diffusion speed, and may correspond to an artificial intelligence model implemented as a weight-based ranking model, a multi-criteria decision making model (MCDM), or a learning-to-rank model (LambdaMART, RankNet family).

[0068] The influencer selection model (116) may be a model trained using a Learning-to-Rank method, a Multi-Objective Optimization method, etc., with training data such as influencer performance indicator data (fitness, activity persistence, overlap, sales forecast, spread speed), past campaign performance records, sales efficiency relative to advertising budget (ROAS, ROI), and brand preference survey data.

[0069] Here, ranking learning may correspond to a method of learning using a loss function to rank multiple candidates, and multi-criteria optimization may correspond to a method of deriving an optimal choice by considering the balance among multiple performance indicators.

[0070] The communication interface (120) is a component for performing data transmission and reception between the electronic device (100) and an external device, and can be connected to the user terminal (200) via wired or wireless communication.

[0071] The wireless communication interface can transmit and receive data with an external device using short-range wireless communication technologies such as Bluetooth, Bluetooth Low Energy, Wi-Fi, Wi-Fi Direct, Ultra Wide Band (UWB), Zigbee, Near Field Communication (NFC), and Infrared Data Association (IrDA), or mobile communication technologies such as 3GPP, Wi-Max, LTE (Long Term Evolution), and 5G. The wireless communication interface may be implemented by including an antenna, a communication chip, and a circuit board for transmitting and receiving electromagnetic waves.

[0072] Wired communication interfaces can communicate with external devices based on physical media such as pair cables, coaxial cables, fiber optic cables, or Ethernet cables.

[0073] The input / output interface may include physical interfaces for direct data input / output with external devices, such as USB (Universal Serial Bus), HDMI (High Definition Multimedia Interface), and audio / video terminals.

[0074] Specifically, the communication interface (120) can receive product information (101) from the user terminal (200).

[0075] The processor (130) is intended to control the overall configuration of the electronic device (100) and can control the operation of the electronic device (100) by executing at least one instruction stored in memory (110). It can be implemented as a single processor as well as as multiple processors. The processor (130) can be implemented in various forms, such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerated Processing Unit), DSP (Digital Signal Processor), NPU (Neural Processing Unit), or a hardware accelerator dedicated to machine learning.

[0076] Additionally, the processor (130) may be configured as a single-core processor or a multi-core processor, and in the case of a multi-core processor, each core may execute instructions independently or cooperatively to perform the method according to the embodiment of the present disclosure. For example, the first and second operations may be performed on a general-purpose CPU core, and the third operation may be performed on an NPU core specialized for AI computation.

[0077] In this way, the processor (130) is implemented in various structures and methods, thereby enabling stable and efficient control of the electronic device (100).

[0078] Referring to FIG. 2a, the processor (130) can control the information acquisition module (131), the suitability evaluation module (132), and the influencer selection module (133).

[0079] The information acquisition module (131) is configured to collect basic data for evaluating influencers.

[0080] Specifically, the information acquisition module (131) can perform the function of acquiring product information (product name, type, function) input from a user terminal, follower information (age, gender, region) provided from a social platform, engagement information (views, likes and comments history), past content script information and sales history information of an influencer.

[0081] Specifically, the information acquisition module (131) can acquire follower information (102) and engagement information from a social platform through API integration or log collection technology, etc.

[0082] Specifically, the information acquisition module (131) can acquire follower information (102) and engagement information through various methods such as API integration, log collection, crawling, and database access from a social platform.

[0083] The information acquisition module (131) can acquire sales history information through various integrations with commerce platforms, etc. For example, sales history information can be acquired through methods such as API integration, data feed collection, and linkage with data provision services, but is not limited thereto.

[0084] The suitability evaluation module (132) is configured to calculate indicators for evaluating from various angles whether an influencer is suitable for promoting a target product.

[0085] Specifically, the suitability evaluation module (132) inputs follower information and product information into the suitability calculation model (111) to obtain the suitability between the target product and the influencer.

[0086] In addition, by inputting engagement information into the activity persistence model (112) described later, an activity persistence can be obtained that indicates the degree to which the level of interest of followers is maintained for a certain period.

[0087] In addition, the influencer's past content script information is input into the duplication model (113) to obtain the duplication based on the similarity and repetition patterns between the content.

[0088] Furthermore, sales history information, or in the case of no sales history, follower information and engagement information can be input into the sales prediction model (114) to obtain the sales prediction indicator of the influencer.

[0089] In addition, by inputting the temporal growth rate of engagement information for past content into the diffusion speed model (115), a diffusion speed indicating the degree of diffusion of content can be obtained.

[0090] In this way, the suitability evaluation module (132) can perform the function of supporting decision-making in the influencer selection module (133) by comprehensively obtaining performance indicators such as suitability, activity duration, overlap, sales forecast, and diffusion speed.

[0091] The influencer selection module (133) is configured to select the optimal influencer by synthesizing the calculated performance indicators.

[0092] Specifically, the influencer selection module (133) can input indicators such as suitability, activity duration, overlap, sales forecast, and diffusion speed into the influencer selection model (116) to perform the function of selecting one or more top influencers among multiple influencers.

[0093] Referring to FIG. 2b, the memory (110) can store an activity persistence model (112), a redundancy model (113), a sales prediction model (114), a diffusion rate model (115), and a generative model (117) in addition to the above-described fitness calculation model (111) and influencer selection model (116).

[0094] The activity persistence model (112) is an artificial intelligence model that receives engagement information and obtains an activity persistence that indicates the degree of interest maintained by followers in time intervals over a certain period.

[0095] The activity persistence model (112) may correspond to a series of artificial intelligence models such as a recurrent neural network (RNN), a long short-term memory network (LSTM), and a time series regression model.

[0096] The activity persistence model (112) may be a model trained using supervised learning, time series prediction learning methods, etc., with follower engagement log data (number of views, number of likes, number of comments, etc.) and follower reaction trend data over time as training data.

[0097] The duplication model (113) is an artificial intelligence model that receives script information of multiple past contents uploaded by an influencer over a certain period and obtains duplication based on similarity between contents and upload frequency.

[0098] The redundancy model (113) may correspond to an embedding-based similarity analysis series of artificial intelligence models, such as document embedding-based similarity analysis models (e.g., BERT, RoBERTa) and multimodal embedding models (e.g., CLIP).

[0099] The overlap model (113) may be a model trained using supervised learning, unsupervised learning, clustering methods, etc., with training data such as script data of past content, posting frequency data, and similarity scores between content.

[0100] The sales prediction model (114) is a model for calculating the expected sales indicator of a target product by receiving sales history information or engagement information for each influencer.

[0101] The sales prediction model (114) may correspond to a regression prediction series artificial intelligence model such as a regression analysis model (e.g., XGBoost, LightGBM, etc.), a time series prediction model (e.g., Prophet), or a deep learning-based regression neural network model (DNN Regressor).

[0102] The sales prediction model (114) may be a model trained using supervised learning, regression analysis, time series prediction methods, etc., with the influencer sales history (purchase conversion rate, number of reviews, actual sales data), follower information, engagement logs, and past similar campaign performance data as training data.

[0103] The diffusion speed model (115) is a model for receiving past content engagement growth rate data and calculating the unit diffusion indicator and final diffusion speed indicator of the content.

[0104] The diffusion rate model (115) may correspond to a series of artificial intelligence models for time series prediction, such as a time series analysis model (ARIMA, Prophet), a recurrent neural network (RNN), a long short-term memory network (LSTM), and a transformer-based time series prediction model (Temporal Transformer).

[0105] The diffusion speed model (115) may be a model trained using supervised learning, time series prediction, regression analysis, etc., with the upload time per content, time interval Engagement log (increase in views, likes, and comments), and follower reaction trend data as training data.

[0106] The generative model (117) is a model that receives product information and influencer characteristic information and generates email content for product collaboration requests or scenarios for promoting target products.

[0107] The generative model (117) may correspond to a large language model (LLM) family artificial intelligence model, such as a Transformer-based natural language processing model (e.g., GPT family LLM, T5, BART).

[0108] The generative model (117) may be a model trained using supervised learning, fine-tuning, prompt-based learning methods, etc., with training data such as product description data, past sponsorship emails, marketing scenario data, content progression order (e.g., progression stage, schedule management information), content volume information (e.g., post length, video length), concept information (e.g., style, tone and manner), call-to-action (CTA) data (e.g., call-to-purchase phrase, hashtag, link placement method), past advertising performance reports, trend analysis data, etc.

[0109] Referring to FIG. 2b, the processor (130) can control a content creation module (134) in addition to the information acquisition module (131), suitability evaluation module (132), and influencer selection module (133) described above.

[0110] The content creation module (134) is configured to create email content for product collaboration requests or promotional scenarios for promoting target products.

[0111] Specifically, the content creation module (134) can input product information (101) and influencer characteristic information into the creation model (117) to create collaboration request email content (601) to be sent to the influencer, or create a promotion scenario (701) including information on the sequence of events for promoting the target product, information on the amount of content, concept information, and call-to-action information.

[0112] FIG. 3 is a flowchart illustrating a series of flows for an electronic device to select an influencer based on suitability according to one embodiment of the present disclosure.

[0113] Referring to FIG. 3, the electronic device (100) can receive product information (101) including various information such as the product name, product type (e.g., cosmetics, food, clothing, etc.), provider information (e.g., brand name), main ingredients, product function, price, specifications, precautions, expiration date, etc. of the target product through user input received from the user terminal (200) (S310).

[0114] For example, product information (101) can be input through an application, webpage, or linked management system (such as a GUI-based input system) of a user terminal (200), and can be received by an electronic device (100) through a communication interface (120) in a manner such as HTTPS-based communication, RESTful API calls, or transmission of a JSON-formatted message.

[0115] The electronic device (100) can obtain follower information including at least one of the age group, gender, and nationality of each of the followers of each of the multiple influencers active on at least one social platform (S320).

[0116] In one embodiment, the information acquisition module (131) can acquire follower information including demographic items such as age group, gender, and nationality for each influencer's followers by utilizing a creator statistics API or affiliate marketing platform (e.g., influencer marketing solution) that is publicly available on major social platforms such as Instagram and YouTube.

[0117] For example, the follower information of Influencer C shows that among the total followers, teenagers make up about 10%, those in their 20s about 70%, those in their 30s about 15%, and other age groups about 5%, while the gender ratio can be analyzed as 85% female and 15% male. Additionally, by nationality, it can be composed of 90% South Korean, 5% Japanese, and 5% from other countries.

[0118] The electronic device (100) can calculate a degree of suitability indicating the suitability of promotion for a target product for each of a plurality of influencers based on follower information (102) and product information (101) obtained for each influencer (S330).

[0119] The suitability evaluation module (132) inputs follower information (102) and product information (101) into a suitability calculation model (111) to calculate a suitability that indicates the suitability of each influencer's promotion of the target product.

[0120] In one embodiment, the fit calculation model (111) can convert the attribute distribution (e.g., age, gender, region) of follower information (102) into a vector form, convert the target attributes (e.g., product category, function, expected consumer base) of product information (101) into a vector form, and then calculate the fit based on the similarity and correction factor between the two vectors.

[0121] The electronic device (100) can select at least one influencer among a plurality of influencers based on the calculated suitability (S340).

[0122] The influencer selection module (133) inputs performance indicators such as activity duration, overlap, sales forecast, and diffusion speed, along with the suitability calculated by the suitability evaluation module (132), into the influencer selection model (116) to select at least one influencer among multiple influencers.

[0123] In one embodiment, the influencer selection model (116) can convert each performance indicator into a normalized vector form, calculate a total score by applying predefined weights, and then rank the influencers based on the total score.

[0124] In addition, the influencer selection model (116) can identify the optimal influencer by applying a learning-to-ranking algorithm (e.g., LambdaMART, RankNet) and a multi-criteria decision-making technique (e.g., AHP, TOPSIS, etc.) to simultaneously consider advertising efficiency and brand suitability.

[0125] The electronic device (100) can obtain influencer characteristic information for each of a plurality of influencers based on follower information and script information.

[0126] Influencer characteristic information refers to an indicator representing an influencer's activity tendencies and promotional style, and may include follower information (age group, gender, region), content topics and content expression methods, main keywords, writing style and tone and manner, frequency of past sponsorships, etc.

[0127] In one embodiment, the electronic device (100) can obtain influencer characteristic information by inputting follower information (102) and script information (103) into a generative model (117).

[0128] A generative model (117) can convert script information (103) into main topics and vocabulary patterns through natural language processing techniques (e.g., Transformer-based embeddings), convert follower information (102) into vector forms, and then combine the two pieces of information to produce influencer characteristic information.

[0129] Influencer characteristic information can be used as an indicator of each influencer's activity tendencies and promotional style.

[0130] The electronic device (100) can calculate the suitability for each of a plurality of influencers based on influencer characteristic information and product information (101).

[0131] This enables a precise assessment of promotional suitability that considers both the influencer's actual activity patterns and the product's target attributes, rather than relying solely on follower counts or single metrics. Consequently, companies and brands can efficiently identify the influencers best suited for their products, thereby maximizing marketing performance.

[0132] The electronic device (100) can obtain engagement information including the number of views and interest of followers on at least one past content provided by each of the plurality of influencers.

[0133] The information acquisition module (131) can acquire engagement information by inputting past content provided by each of the multiple influencers into the generative model (117).

[0134] In this case, engagement information is an indicator of the reactions and interactions shown by followers to past content, and may include items such as view count, like count, comment count, share count, save count, watch time, and click-through rate (CTR).

[0135] For example, the electronic device (100) can receive data such as 500,000 views, 12,000 likes, 900 comments, and an average viewing time of 3 minutes and 40 seconds for a cosmetic review video posted on YouTube by influencer A, and store it as engagement information.

[0136] The electronic device (100) can obtain engagement information including time-based and daily log data (e.g., hourly views, daily number of likes and comments, weekly share growth rate, etc.) from a web page or API of a social platform.

[0137] The conformity assessment module (132) can obtain an activity duration indicating the continuity of the followers' engagement information by time interval during a first period that is pre-set based on the engagement information.

[0138] Activity persistence is a metric that quantifies how follower engagement (views, likes, comments, shares, etc.) decreases or remains constant over time following content upload, indicating whether an influencer's content receives only fleeting interest or sustains attention over an extended period.

[0139] The suitability evaluation module (132) inputs engagement information (e.g., number of views, number of likes, number of comments, number of shares by time period) into the activity persistence model (112) to obtain an activity persistence that indicates the degree of engagement retention of each content.

[0140] The activity persistence model (112) may correspond to a model that receives engagement information as a time-series input and identifies changes in response intensity over time. Specifically, the engagement information may be normalized by time unit and configured as an input sequence, and the activity persistence model (112) may estimate the initial response size, the rate of decrease over time, the long-term retention rate, etc., based on this input using techniques such as a recurrent neural network (RNN), a long-term memory network (LSTM), or a time-series regression method, and may calculate the activity persistence by synthesizing these estimated values.

[0141] Unlike simple cumulative engagement figures, activity persistence allows for a quantitative assessment of how long followers' interest is maintained. Consequently, companies can distinguish between influencers with only temporary high engagement and those who consistently maintain interest over the long term, enabling them to more accurately identify influencers with expected long-term marketing effects.

[0142] The electronic device (100) can obtain script information (103) for multiple past content uploaded during a second period pre-set by multiple influencers.

[0143] The electronic device (100) can obtain similarity between the plurality of past contents based on script information (103).

[0144] The conformity evaluation module (132) inputs the script information (103) into the redundancy model (113) to obtain the similarity between past contents.

[0145] Similarity refers to an indicator that shows how similar the subject, keywords, writing style, and tone and manner are by comparing the script information of each of multiple past content items.

[0146] The electronic device (100) can identify a group of similar content whose similarity is greater than or equal to a preset threshold.

[0147] A similar content group refers to a set of content among multiple past contents in which the similarity of script information is identified as being above a preset threshold. By identifying a similar content group, the electronic device (100) can distinguish content that has been repeatedly produced with the same or similar subject or expression style.

[0148] The electronic device (100) can obtain a degree of redundancy indicating the degree of repetition of past content based on the frequency with which similar content groups are uploaded within a pre-set second period and the number of similar content included in the similar content groups.

[0149] Duplication refers to an indicator showing how repeatedly an influencer posts the same or similar content over a certain period.

[0150] The conformity evaluation module (132) inputs the script information (103) into the redundancy model (113) to obtain the redundancy between past contents.

[0151] The redundancy model (113) converts input script information into an embedding vector and calculates similarity indicators such as cosine similarity and Jaccard similarity between contents, and can identify contents whose similarity is greater than or equal to a preset threshold as similar content groups. Subsequently, the frequency of uploads of similar content groups within a preset second period and the number of similar contents included in the group can be calculated to calculate the redundancy indicating the degree of repetition of past contents.

[0152] Through this, the electronic device (100) can evaluate whether an influencer excessively and repeatedly uploads content of the same or similar subject within a certain period and can determine the diversity of the content. Therefore, the degree of duplication can be used as an important criterion for evaluating an influencer's differentiated promotional capabilities, rather than a simple view count-based metric.

[0153] The electronic device (100) can identify sales history information for each of a plurality of influencers based on at least one of a commerce platform and a social platform.

[0154] A commerce platform refers to an e-commerce service (e.g., online shopping malls, live commerce platforms, affiliate marketing systems, etc.) that allows influencers to sell products directly or generate sales through affiliate links.

[0155] Sales history information is performance data generated through product promotion or sales activities performed by an influencer in the past, and may include the number of sales per product, sales amount, purchase conversion rate, cart add rate, purchase reviews, etc.

[0156] The electronic device (100) can obtain sales forecast indicators based on sales history information when sales history information is identified.

[0157] Sales forecast indicators refer to metrics that quantify the expected sales volume or projected revenue amount when a relevant influencer conducts promotional activities for a specific product.

[0158] If the electronic device (100) does not identify sales history information, it can obtain a sales forecast indicator by referring to a pre-set forecast indicator based on the follower information (102) of multiple influencers.

[0159] The projected indicators correspond to baseline data that reflects the general purchasing propensity shown by each follower group for specific product categories, by mapping the demographic attributes of followers (e.g., age group, gender, nationality, region of residence) with the average purchase rate by product category.

[0160] Through this, the electronic device (100) can perform sales forecasting based on individual performance for influencers with past sales history data, and can also make reasonable sales forecasting for new influencers without sales history by utilizing follower characteristics and market average indicators.

[0161] The electronic device (100) can obtain an increase rate of Engagement information for each past content during a pre-set third period including the time of upload.

[0162] The conformity evaluation module (132) can obtain an increase rate by inputting engagement information, including the upload time, into the diffusion rate model (115).

[0163] The diffusion speed model (115) can obtain an increase rate by calculating the amount of change in engagement metrics (views, likes, comments, shares, etc.) collected at regular time intervals (e.g., 1 minute, 10 minutes, 1 hour) after content upload, based on a cumulative basis, a ratio basis, etc.

[0164] The conformity evaluation module (132) can obtain a unit diffusion indicator representing the diffusion speed of each past content based on the growth rate.

[0165] A unit diffusion metric is a quantitative indicator representing how quickly engagement increases during a certain period immediately after content is uploaded, and can be defined as the rate of engagement growth per unit of time (e.g., views / minute, comments / hour, shares / day, etc.).

[0166] The conformity evaluation module (132) can obtain a diffusion rate indicator based on a unit diffusion indicator.

[0167] The diffusion speed indicator is a metric calculated by aggregating and integrating the unit diffusion indicators of individual content for multiple past pieces of content provided by a specific influencer, and it represents the overall diffusion tendency of the influencer.

[0168] The diffusion speed model (115) can estimate the diffusion pattern for each influencer by aggregating the input unit diffusion indicators in a time series or by applying weights to average them. Specifically, the diffusion speed model (115) can fit a curve of the growth rate over time after content upload and calculate the diffusion speed indicator by combining the estimated coefficients (e.g., initial slope, half-life, diffusion coefficient, etc.).

[0169] The electronic device (100) can reflect not only the total amount of engagement but also the pattern of diffusion over time, and can effectively identify influencers who receive a rapid response, particularly within a short period of time. This enables the recommendation of influencers suitable for product promotion that requires a rapid viral effect.

[0170] The influencer selection module (133) can select an influencer suitable for promoting a target product based on at least one of the items of suitability, activity duration, overlap, sales prediction indicator and diffusion speed indicator.

[0171] In one embodiment, the influencer selection module (133) can select the influencer with the highest total score calculated by assigning weights to each item.

[0172] In another embodiment, the influencer selection module (133) may select an influencer if at least one of the scores calculated by weighting each item excluding duplicates exceeds a preset threshold.

[0173] In another embodiment, the influencer selection module (133) can select from the remaining influencers, excluding influencers whose duplication level is greater than or equal to a preset duplication level.

[0174] According to this configuration, content diversity and novelty can be secured by excluding influencers who excessively produce only repetitive and similar content. Therefore, it is effective to prioritize the selection of influencers who can enhance the promotional effect of the target product and maintain consumer interest for an extended period.

[0175] FIG. 4 is a diagram illustrating the operation of an electronic device providing a list of influencers with high suitability and a chatbot function according to one embodiment of the present disclosure.

[0176] Referring to FIG. 4, the electronic device (100) generates a list (400) containing a certain number of influencers with high suitability among a plurality of influencers, provides the generated list (400) to a user terminal (200), and can select at least one influencer from the list (400) according to user input received through the user terminal (200).

[0177] The list (400) may be configured to include sales performance information, follower information, engagement information, and scores for a certain number of influencers with high scores among multiple influencers, and for example, the sales performance, number of followers, engagement indicators, and score (e.g., 90 points) of 'Influencer A' may be displayed together.

[0178] Additionally, the electronic device (100) may include a chatbot function (410) to support interaction with the user, and the chatbot function (410) may provide detailed information about each influencer included in the list (400) in real time according to the user's query.

[0179] The chatbot function (410) may include a natural language processing-based question-and-answer system or a conversational artificial intelligence function, and may be linked with components included in the list (400) to support the user's information search.

[0180] In one embodiment, the user can input a natural language query such as “Tell me what products Influencer B promoted in the past” or “I am curious about Influencer C’s recent content activity,” and accordingly, the chatbot function (410) can output information corresponding to the query in a conversational form.

[0181] FIG. 5 is a diagram illustrating the operation of an electronic device providing a search function according to one embodiment of the present disclosure.

[0182] Referring to FIG. 5, the electronic device (100) may include a search function (501).

[0183] The search function (501) is a model for receiving a user's natural language query to search for suitable influencers and provide related information, and may include a query analysis model, an attribute extraction model, a condition-based filtering model, and a similarity-based sorting model.

[0184] In one embodiment, when a user inputs a query saying, “Introduce me to a beauty influencer that is popular among women in their teens and twenties in the United States,” the search function (501) can analyze the input sentence through a query analysis model to extract features such as ‘region: United States’, ‘age group: teens and twenties’, ‘gender: female’, and ‘category: beauty’, search for influencers that meet the conditions through a condition-based filtering model, and then sort the results through a similarity-based sorting model to provide the results to the user in the form of an influencer list.

[0185] FIG. 6 is a diagram illustrating the operation of an electronic device generating mail content and sending it to the mail address of a selected influencer according to one embodiment of the present disclosure.

[0186] Referring to FIG. 6, the electronic device (100) can generate email content (601) regarding product introduction and collaboration requests through a generative model (117) based on script information (103) and product information (101) of a selected influencer.

[0187] The electronic device (100) can send the generated email content to the email address of a selected influencer.

[0188] Additionally, although not shown in the drawing, the electronic device (100) may include a collection module, and the collection module is configured to obtain the influencer's email address.

[0189] The collection module can obtain the influencer's email address through methods such as web crawling-based information collection, natural language processing (NLP)-based contact extraction, and email pattern recognition using regular expressions. The electronic device (100) can send the generated email content (601) to the selected influencer's email address obtained through the collection module.

[0190] The electronic device (100) can generate common data based on product information (101) and individual data based on script information (103) of a selected influencer by using generative artificial intelligence techniques such as a large-scale language model (LLM) through a generative model (117).

[0191] For example, common data may include content that can be included in all emails regardless of a specific influencer, such as product information (101) (e.g., product features, brand image, efficacy) and collaboration proposals, and individual data may include information regarding the characteristics of a selected influencer (e.g., main content field, buzzwords, tone of voice) or pre-set keywords (e.g., travel, filming, lecture, etc.) based on the selected influencer's past script information (103).

[0192] In one embodiment, the electronic device (100) may input product information (101) (e.g., product features, brand image, efficacy) and collaboration proposal text (e.g., a proposal email for brand promotion, a marketing campaign proposal phrase, a sponsorship request letter, etc.) into a generative model (117) for the generation of common data. Based on the input product information (101) and collaboration proposal text, the generative model (117) may output common data including phrases explaining the main efficacy of the product and brand message, guidance phrases encouraging product use, and explanatory phrases related to the collaboration proposal.

[0193] Additionally, the electronic device (100) may input past script information (103) of a selected influencer and preset keywords (e.g., travel, filming, lecture, event participation) into a generative model (117) for the generation of individual data. Based on the input script information (103) and keywords, the generative model (117) may output individual data including preset keywords corresponding to the characteristics of the selected influencer (e.g., main content field, buzzword, tone of voice) or the current status information of the selected influencer.

[0194] To this end, the electronic device (100) can identify pre-set keywords (e.g., travel, exercise, filming, event participation, etc.) corresponding to current status information within the text based on the script information (103) of the selected influencer through the generative model (117).

[0195] Meanwhile, the beginning of the email content (601) may be a greeting area corresponding to the first paragraph of the email body. The electronic device (100) can generate email content (601) that attracts the attention of a selected influencer by including individual data generated through a generative model (117) in the beginning. For example, the electronic device (100) can generate a greeting phrase that reflects the influencer's characteristics (e.g., main content field, slang, tone of voice) based on the selected influencer's past script information (103). Additionally, the electronic device (100) can generate a phrase related to the influencer's recent activities by utilizing pre-set keywords (e.g., travel, exercise, filming, event participation, etc.).

[0196] In this case, the beginning of the email content (601) may include phrases that reflect the influencer's characteristics or pre-set keywords, such as “I was impressed by how you incorporated the inspiration you gained from your recent travels into your content” or “The sincere expressions you always show in the ○○ field stand out.” These phrases can have the effect of making the recipient feel positive, thereby improving the likelihood of acceptance of the collaboration proposal.

[0197] At this time, the generative model (117) can be trained to generate common data using text data such as collaboration request emails, product promotion emails, and general business communication emails as training data. Additionally, the generative model (117) can be trained to generate individual data using data such as email data that multiple influencers actually evaluated positively or felt favorably upon receiving, and data such as content style, expression method, and preferred writing style included in the past content of multiple influencers as training data.

[0198] Accordingly, the electronic device (100) can generate mail content (601) containing common data and individual data through a generative model (117), and even when multiple influencers are selected, it can generate and send mail content (601) containing common data and individual data generated for each selected influencer.

[0199] FIG. 7 is a diagram illustrating the operation of an electronic device generating a promotional scenario according to one embodiment of the present disclosure.

[0200] Referring to FIG. 7, the electronic device (100) can obtain promotional plan information including at least one of a promotional budget, a promotional concept, and content volume information for the promotion of a target product in a generative model (117) for generating a scenario.

[0201] Here, promotional plan information refers to predefined conditions or constraints that a selected influencer may refer to when promoting a target product. For example, the promotional budget refers to the amount of compensation to be paid to the influencer or the costs to be incurred for content creation and distribution, and the promotional concept may include the message to be emphasized during product promotion, brand image, and atmosphere tailored to the target audience. Additionally, content length information defines the length or scale of promotional content and may include, for instance, running time (e.g., 30 seconds, 1 minute, 5 minutes) for video content, the number of images (e.g., 1 image, 5 images) for image content, or the number of paragraphs or characters for text content.

[0202] The electronic device (100) can input promotional plan information, product information, and follower information to obtain a promotional scenario (701) for target content for a selected influencer to promote a target product.

[0203] Here, a promotional scenario refers to data defining the developmental structure and expression elements of content to be generated by an influencer, and may include at least one of sequence information, content length information, and concept information. Sequence information may include a developmental sequence such as introduction-body-conclusion during content creation, or a storyline such as product introduction-usage scene-effect emphasis. Content length information refers to length or volume restrictions per piece of content (e.g., video length, number of images, number of paragraphs), and concept information refers to the brand message, thematic tone, key keywords, and emotional appeal method to be emphasized during product promotion.

[0204] Through these promotional scenarios, influencers can promote products naturally and effectively, while users can secure differentiated content from each influencer while maintaining a consistent message.

[0205] The content creation module (134) can obtain a promotion scenario (701) by inputting promotion plan information, product information, and follower information into the generative model (117).

[0206] For this purpose, the generative model (117) can be trained using product promotion campaign data, past influencer marketing scenario data, brand-specific promotion manuals, advertising copy data and consumer response data as training data.

[0207] FIG. 8 is a diagram illustrating the operation of an electronic device according to one embodiment of the present disclosure to determine the contribution of an influencer to target content.

[0208] Referring to FIG. 8, when target content (801) for promoting a target product of a selected influencer is uploaded to a social platform, the electronic device (100) can determine the contribution of the target content (801) to the sales of the target product.

[0209] The electronic device (100) can determine the contribution based on the engagement information for the target content (801) and the sales information of the target product, which are verified in time intervals within a certain period including the time of uploading the target content (801).

[0210] To determine the contribution, the electronic device (100) can calculate the contribution based on engagement information regarding the target content (801) and sales information of the target product, which are verified in time intervals within a certain period including the time of uploading the target content (801). By comparing changes in engagement and sales based on the period before and after the upload of the target content (801), the causal relationship between the target content (801) and the increase in sales of the target product can be analyzed more accurately.

[0211] Additionally, the electronic device (100) may include a contribution judgment model, although not shown in the drawing. The contribution judgment model receives engagement information (e.g., number of views, number of likes, number of comments, etc.) and sales information for a plurality of preset intervals including the time of upload of the target content (801), and can identify the degree of influence that the engagement has had on the change in sales in each interval. In this case, a high weight may be assigned if the correlation is high in the interval immediately after upload, and a low weight may be assigned in the interval that is far away.

[0212] To this end, the contribution judgment model can be trained with training data that includes the past content upload history of multiple influencers by product category (e.g., fashion, beauty, home appliances), engagement information and sales information before and after uploading, product information (101) (e.g., price range, brand image, etc.), and follower information (102) (e.g., age, gender, regional distribution, etc.). In this case, the impact of engagement on sales by segment can be learned by including cases where content by different influencers is uploaded for the same product.

[0213] Specifically, the contribution judgment model utilizes machine learning-based models such as regression analysis, time series models (LSTM, Transformer, etc.), Random Forest, and Gradient Boosting to calculate the degree of influence of pre-set engagement information for multiple intervals on changes in sales.

[0214] Additionally, the electronic device (100) may include a contribution determination module, although not shown in the drawing. The contribution determination module inputs segment-specific engagement information and sales information into a contribution determination model to obtain the contribution of the target content (801) to the sales information of the target product.

[0215] In another embodiment, the electronic device (100) can generate thank-you email content for a selected influencer through a generative model (117) when the contribution of the target content (801) exceeds a preset threshold. For example, the thank-you email content may include thank-you keywords (e.g., gratitude for collaborative achievements, expression of intent for long-term collaboration), the contribution of the target content (801), and sales information. This has the effect of strengthening the relationship with the influencer and increasing the possibility of long-term collaboration.

[0216] Meanwhile, the various embodiments described above may be implemented by combining two or more embodiments, provided that they do not conflict or contradict each other.

[0217] Specifically, each of the multiple operations, steps, and configurations for implementing one embodiment may be embodied in another embodiment, or the operations and steps of another embodiment may be followed by the last operation or step of one embodiment, but are not limited thereto.

[0218] Meanwhile, computer instructions or computer programs for performing processing operations of each device, device, terminal, computer, processor, module, or engine, etc., according to the various embodiments of the present disclosure described above, may be stored on a non-transitory computer-readable medium. When such computer instructions or computer programs stored on the non-transitory computer-readable medium are executed by the processor of a specific device, they cause the specific device described above to perform processing operations according to the various embodiments described above.

[0219] A non-transient computer-readable medium refers to a medium that stores data semi-permanently and can be read by a device, unlike media that store data for a short period of time such as registers, caches, and memory. Specific examples of non-transient computer-readable media include CDs, DVDs, hard disks, Blu-ray discs, USBs, memory cards, and ROMs.

[0220] According to one embodiment, the method according to the various embodiments disclosed herein may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)), or distributed online (e.g., download or upload) through an application store (e.g., Play Store™) or directly between two user devices (e.g., smartphones). In the case of online distribution, at least a portion of the computer program product (e.g., downloadable app) may be temporarily stored or temporarily created on a device-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0221] Although preferred embodiments of the present disclosure have been illustrated and described above, the present disclosure is not limited to the specific embodiments described above. It is understood that various modifications can be made by those skilled in the art without departing from the essence of the present disclosure as claimed in the claims, and such modifications should not be understood individually from the technical spirit or perspective of the present disclosure. Explanation of the symbols

[0222] 100: Electronic device 101: Target Products 102: Follower Information 103: Script Information 104: Promotion Plan Information 110: Memory 111: Goodness-of-fit model 112: Activity Continuity Model 113: Redundancy Model 114: Sales Forecast Model 115: Diffusion rate model 116: Influencer Selection Model 117: Generative Model 120: Communication interface 130: Processor 131: Information Acquisition Module 132: Conformity Assessment Module 133: Influencer Selection Module 134: Content Creation Module 200: User terminal 400: List 410: Chatbot Features 501: Search function 601: Mail Content 701: Promotional Scenario 801: Target content

Claims

Claim 1 A method of operating an electronic device comprises: receiving product information including at least one of a product name, product type, and product function of a target product through user input received from a user terminal; obtaining follower information including at least one of an age group, gender, and nationality of followers for each of a plurality of influencers active on at least one social platform; and calculating a degree of suitability indicating the suitability for promotion of the target product for each of the plurality of influencers based on the follower information and product information obtained for each influencer. The method of operation of the electronic device includes the step of selecting at least one influencer among the plurality of influencers based on the suitability calculated above; and the method of operation of the electronic device includes the step of obtaining Engagement information including the number of views and interest of followers for at least one past content provided by each of the plurality of influencers; and the step of calculating the suitability includes the step of obtaining an activity persistence indicating the persistence of the followers' Engagement information by time interval during a preset first period based on the Engagement information; the step of obtaining an increase rate of the Engagement information for each of the past content during a preset third period including the upload time; the step of obtaining a unit diffusion index indicating the diffusion speed of each of the past content based on the increase rate; and the step of obtaining a diffusion speed index based on the unit diffusion index; and the method of operation of the electronic device includes the step of obtaining script information for a plurality of past content uploaded by the plurality of influencers during a preset second period; and the step of obtaining similarity between the plurality of past content based on the script information. and a step of identifying a group of similar content whose similarity is greater than or equal to a preset threshold;A method of operation of an electronic device comprising: a step of calculating suitability, wherein the step of calculating suitability for each of the plurality of influencers based on the follower information and the product information; and a step of obtaining a redundancy indicating the degree of repetition of the past content based on the frequency with which the similar content group was uploaded within the preset second period and the number of similar content included in the similar content group; and wherein the method of operation of the electronic device comprises a step of selecting at least one influencer among the plurality of influencers based on at least one item among the suitability, the activity duration, the redundancy, and the diffusion speed indicator. Claim 2 In claim 1, the method of operation of the electronic device comprises: a step of obtaining script information of at least one past content provided by each of the plurality of influencers to the social platform; and a step of obtaining influencer characteristic information for each of the plurality of influencers based on the follower information and the script information; and the step of calculating the suitability is to calculate the suitability for each of the plurality of influencers based on the influencer characteristic information and the product information. Claim 3 delete Claim 4 delete Claim 5 In claim 1, the method of operation of the electronic device comprises: a step of identifying sales history information for each of the plurality of influencers based on at least one of a commerce platform and a social platform; and a step of calculating the suitability, wherein the step of calculating the suitability for each of the plurality of influencers based on follower information and product information; and a step of obtaining a sales forecast indicator based on the sales history information when the sales history information is identified. Claim 6 A method of operation of an electronic device according to claim 5, wherein the step of calculating the suitability comprises: a step of calculating the suitability for each of the plurality of influencers based on the follower information and the product information; and, if the sales history information is not identified, a step of obtaining the sales forecast indicator by referring to a pre-set forecast indicator based on the follower information of the plurality of influencers. Claim 7 delete Claim 8 In paragraph 2, the method of operation of the electronic device further comprises: a step of generating email content regarding a product introduction and a collaboration request through a generative model based on the script information and product information of the selected influencer; and a step of sending the generated email content to the email address of the selected influencer. Claim 9 The method of operation of the electronic device according to claim 1 comprises: a step of obtaining promotional plan information including at least one of a promotional budget, a promotional concept, and content volume information for the promotion of the target product in a generative model for generating a scenario; and a step of obtaining a promotional scenario for target content for the promotion of the target product by the selected influencer by inputting the promotional plan information, the product information, and the follower information; wherein the promotional scenario includes at least one of progress order information, content volume information, and concept information. Claim 10 In claim 1, the method of operation of the electronic device comprises: a step of determining the contribution of the target content to the sales of the target product when the target content for promoting the target product of the selected influencer is uploaded to the social platform; wherein the step of determining the contribution to the sales of the target product is to determine the contribution based on Engagement information regarding the target content and sales information of the target product confirmed in time intervals within a certain period including the time of upload of the target content.

Images