KR20230029158A - method for automatically creating shopping mall banner sentences - Google Patents

Abstract

According to an embodiment of the present invention, provided is a method for automatically generating shopping mall banner sentences, which includes the steps of: collecting banner learning element information from a pre-built platform network; building a natural language learning model by learning combinational relationship information between banner sentence elements corresponding to a promotion effect, based on the banner learning element information; and applying target platform information to the learning model and outputting banner sentence information corresponding to the target platform information and having an optimized promotion effect.

Description

Method for automatically creating shopping mall banner sentences {method for automatically creating shopping mall banner sentences}

The present invention relates to a method for automatically producing a shopping mall banner text. More specifically, the present invention relates to a method for automatically generating a sentence for a shopping mall banner based on a natural language learning model.

In the process of operating online e-commerce (which can be called a shopping mall), promotion of the shopping mall is an indispensable activity in order to increase awareness of the shopping mall and increase sales. Such publicity activities of shopping malls are often called "promotions", which can be understood as "various publicity activities conducted with a specific theme".

A shopping mall operator uses various publicity means to increase the effect of promotion. Among them, a banner that shows the concept of promotion and related information concisely and with impact can be said to be the core of the publicity means. The performance of the promotion has become an important factor to the extent that it depends on the quality of the banner, but the task of creating a high-quality banner is the domain of experts who specialize in design. It is a reality that we are commissioned to produce.

Accordingly, automation methods for efficiently creating banner designs without relying on experts have been proposed.

Common banner automation methods are based on template-based methods. The template method is a method of providing a user with a banner layout combination according to a plurality of preset template types and providing an editing interface allowing the user to select and edit a template.

However, in the template-based method, the user must finally select a specific template and edit it again. In the case of a user without professional knowledge, the banner itself is not perfect because it selects any template or a template that does not consider customers and consumers. There is a problem with losing.

In addition, the banner automation methods proposed so far can only be configured based on a uniform template, and the text of the banner is forced as product information without considering the image characteristics of each product or person. Since it takes more time to reinforce through manual work, actual automation is not implemented.

In addition, due to the uniform template-based configuration, only banners with similar sentences are repeatedly generated as a whole, and characteristics required by each individual user cannot be reflected, and similarly, there is a problem involving manual work.

The present invention has been made to solve the above problems, in response to the classification of characteristics of each arrangement element such as a product object and a banner text in a banner image, inferring a combinational relationship between banner text elements that can maximize a promotion effect. A natural language learning model is built with information collected through the existing online shopping mall network system, and based on this natural language learning model, as automated banner sentences are implemented, professionalism can be improved and promotional effects can be enhanced. In natural language learning, An object of the present invention is to provide a method for automatically producing banner sentences for a shopping mall based on a natural language learning model, which enables personalization of appropriately automatically generated banners according to the reflection of detailed user tendencies.

A method according to an embodiment of the present invention for solving the above problems includes collecting banner learning element information from a pre-built platform network; Based on the banner learning element information, building a natural language learning model by learning combinational relationship information between banner sentence elements corresponding to a promotion effect; and applying target platform information to the learning model and outputting banner text information corresponding to the target platform information and having an optimized promotion effect.

In addition, the method according to the embodiment of the present invention for solving the above problems may be implemented as a computer readable recording medium and a computer program for executing the method on a computer.

According to an embodiment of the present invention, a natural language learning model of a banner sentence element associated with a product in a banner image is built with information collected through an existing online shopping mall network system, and based on this natural language learning model, an automated banner sentence is created. can be implemented

Accordingly, according to an embodiment of the present invention, it is possible to improve the professionalism and promotion effect in the banner design automation process, and to personalize appropriate automatically generated banner sentences according to the reflection of detailed user tendencies in natural language learning. It is possible to provide a method of automatically producing a shopping mall banner sentence based on a natural language learning model.

1 is a conceptual diagram schematically showing the entire system of the present invention.
2 is a block diagram showing an apparatus for automatically generating a banner according to an embodiment of the present invention in more detail.
3 to 5 are diagrams for explaining in detail a banner design service interface according to the operation of the apparatus for automatically generating a banner according to an embodiment of the present invention.

The following merely illustrates the principles of the present invention. Therefore, those skilled in the art can invent various devices that embody the principles of the present invention and fall within the concept and scope of the present invention, even though not explicitly described or shown herein. In addition, it is to be understood that all conditional terms and embodiments listed herein are, in principle, expressly intended only for the purpose of making the concept of the present invention understood, and not limited to such specifically listed embodiments and conditions. It should be.

Further, it should be understood that all detailed descriptions reciting specific embodiments, as well as principles, aspects and embodiments of the present invention, are intended to encompass structural and functional equivalents of these matters. In addition, it should be understood that such equivalents include not only currently known equivalents but also equivalents developed in the future, that is, all devices invented to perform the same function regardless of structure.

Thus, for example, the block diagrams herein are to be understood as representing conceptual views of exemplary circuits embodying the principles of the present invention. Similarly, all flowcharts, state transition diagrams, pseudo code, etc., are meant to be tangibly represented on computer readable media and represent various processes performed by a computer or processor, whether or not the computer or processor is explicitly depicted. It should be.

The functions of various elements shown in the drawings including functional blocks represented by processors or similar concepts may be provided using dedicated hardware as well as hardware capable of executing software in conjunction with appropriate software. When provided by a processor, the functionality may be provided by a single dedicated processor, a single shared processor, or a plurality of separate processors, some of which may be shared.

In addition, the explicit use of terms presented as processor, control, or similar concepts should not be construed as exclusively citing hardware capable of executing software, but without limitation, digital signal processor (DSP) hardware, ROM for storing software (ROM), random access memory (RAM) and non-volatile memory. Other hardware for the governor's use may also be included.

In the claims of this specification, components expressed as means for performing the functions described in the detailed description include, for example, a combination of circuit elements performing the functions or all types of software including firmware/microcode, etc. It is intended to include any method that performs the function of performing the function, combined with suitable circuitry for executing the software to perform the function. Since the invention defined by these claims combines the functions provided by the various enumerated means and is combined in the manner required by the claims, any means capable of providing such functions is equivalent to that discerned from this specification. should be understood as

The above objects, features and advantages will become more apparent through the following detailed description in conjunction with the accompanying drawings, and accordingly, those skilled in the art to which the present invention belongs can easily implement the technical idea of the present invention. There will be. In addition, in describing the present invention, if it is determined that a detailed description of a known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a conceptual diagram schematically illustrating an entire system according to an embodiment of the present invention.

The entire system according to an embodiment of the present invention may include an automatic banner generation device 100, a user terminal 200, a natural language learning model 300, a platform management server 400, and a third party platform 500.

In particular, the automatic banner generation device 100 includes each user terminal 200, a third party platform 500, and a platform management server ( 400), can be connected to the natural language learning model 300 through a wired/wireless network, and can perform mutual communication.

Here, each network is a Local Area Network (LAN), a Wide Area Network (WAN), a Value Added Network (VAN), a Personal Area Network (PAN), a mobile communication network ( It can be implemented in all types of wired/wireless networks, such as a mobile radiocommunication network) or a satellite communication network.

In addition, the user terminal 200 is any one of a computer, a mobile phone, a smart phone, a smart pad, a laptop computer, a personal digital assistant (PDA), and a portable media player (PMP). It may be at least one multi-device of an individual device or a common device such as a kiosk or a stationary display device installed in a specific place, and provides user information to the automatic banner generation device 100, and the automatic banner creation device 100 It is possible to receive and output the processed information from

In such a system configuration, the banner automatic generation device 100 collects banner learning element information from the third party platform 500 configured with a pre-built platform network, and responds to a promotion effect based on the banner learning element information. A banner sentence (phrase) in which the promotional effect corresponding to the target platform information is optimized by building a natural language learning model by learning combinational relationship information between banner sentence (phrase) elements that information can be printed.

And, based on the platform information collected from the platform management server 400 and the user information collected from the user terminal 200, the banner automatically generating device 100 personalizes and optimizes the banner text information for each user and platform. A personalized and optimized automatic banner design may be configured and provided to the user terminal 200 or the platform management server 400 .

Here, the third party platform 500, as an existing online shopping mall, holds a lot of various banner information, can provide pre-established banner design information, as well as sales ranking and sales volume information, so learning Promotion effect information for may also be collected at the same time. In addition, since the third party platform 500 has a design tendency for each platform, platform tendency information may also be collected according to a preset standard table.

Accordingly, the automatic banner generation device 100 may identify banner sentence element classification information corresponding to the banner characteristic information from banner information collected from the third party platform 500, and word combination information between the banner sentence element classification information is a natural language. It can be used as an input to a learning model. In addition, the platform propensity information described above may also be used as an input, and promotion effect information based on sales ranking or sales volume information may be used for output verification purposes.

Here, the banner sentence element classification information may include at least one of a trend keyword, a product category keyword, an abbreviation keyword, a neologism keyword, an attribute keyword, a tag keyword, and a product name text identified from the banner sentence text.

And, the automatic banner generation device 100 builds the natural language learning model 300 from the collected banner learning element information. A known Convolutional Neural Network (CNN) model may be used as the learning model, and a learning natural language learning model that takes the above-described relationship information between the banner sentence element classification information and platform propensity information as inputs and promotes effect information as an output. this can be built.

Accordingly, the natural language learning model may be constructed as a model that outputs, based on learning, how each banner sentence element classification is relatively combined to form a banner sentence to optimize a promotion effect. Examples of these results will be described later with reference to FIGS. 4 and 5 .

According to this process, the automatic banner generation device 100, based on the pre-built natural language learning model 300, when platform information from the platform management server 400 and user information of the user terminal 200 are input, It is possible to provide an automatic banner design service that obtains corresponding optimized banner text information and provides automatic banner design information based on the banner text information to the platform management server 400 or the user terminal 200 .

2 is a block diagram illustrating an apparatus for automatically generating a banner according to an embodiment of the present invention in more detail.

Referring to FIG. 2 , the banner automatic generation device 100 according to an embodiment of the present invention includes a data collection unit 110, a learning model building unit 120, a platform information management unit 130, and a user information management unit 140. , an automatic banner information processing unit 150 and a feedback information processing unit 160.

The data collection unit 110 collects banner learning element information from a pre-built platform network.

The pre-built platform network may be an example of an online shopping mall network operated on a platform 500 of another company.

The banner learning element may include banner sentence element classification information corresponding to the banner sentence element characteristic information, and the banner sentence element classification information may include a trend keyword, a product category keyword, an abbreviation keyword, and a new word identified from the banner sentence text. It may include at least one of keywords, attribute keywords, tag keywords, and product name text.

Also, the banner learning element may include platform tendency information. For example, various styles such as a modern style, a chic style, a monotone style, a pajama style, and a dandy style may be allocated as each platform tendency information.

Also, the banner learning element may include combination relation information between banner sentence element classification information. For example, the combination relation information may be calculated as vector information representing adjacency of a second banner sentence element corresponding to a first banner sentence element. In addition, the relative combination relationship information may further include selection priority information, removal priority information, and the like between the first banner sentence element and the second banner sentence element considering the length of the overall banner text.

Further, the learning model construction unit 120 builds the natural language learning model 300 by learning combination relation information between banner sentence elements corresponding to a promotion effect based on the banner learning element information. As the natural language learning model, a keyword classification dictionary built based on the NER (Named Entity Recognition) method, which is known to be effective in noun recognition, and a NLP (Natural Language Processing) method learning model for natural language processing may be used, respectively. The correct output value for performing backpropagation of the learning model through processing of the error (LOSS) function is the detailed product information and promotion information (sales, sales ranking, search ranking, popularity, etc.) included in the banner learning element described above. can be used

More specifically, for example, the natural language learning model is a reinforcement learning model capable of reinforcing word combinations of banner sentences for product detailed information with high promotional effect, such as trend keywords, product category keywords, and abbreviated keywords extracted from product detailed information. The learning model uses a combination of at least one of a new word keyword, attribute keyword, tag keyword, and product name text to add a specific keyword that is inferred to have a high promotional effect or remove a specific keyword that is inferred to have a low promotional effect. It can be formed according to the construction.

In addition, the automatic banner information processing unit 150 applies user information and target platform information to the learning model, and outputs banner sentence information in which the promotion effect is optimized corresponding to the target platform information. The automatic banner information processing unit 150 may configure a banner design based on the banner text information and provide it to the user terminal 200 or the platform management server 400 .

Here, the target platform information may include target product detailed information to be provided to customers on the target platform.

The automatic banner information processing unit 150 extracts the banner learning element information from the target product detailed information, applies it to the learning model, and combines the banner text corresponding to the product detailed information having the highest promotion effect from the output of the learning model. information can be obtained.

In addition, the automatic banner information processing unit 150 may provide banner sentence information configured based on the banner sentence combination information to the platform management server 400 .

In addition, the feedback information processing unit 160 may receive feedback information from the platform management server 400 or the user terminal 200 to process renewal learning of the natural language learning model.

3 to 5 are diagrams for explaining in detail a banner design service interface according to the operation of the apparatus for automatically generating a banner according to an embodiment of the present invention.

Referring to FIG. 3 , target platform information and user information for automatic banner design according to an embodiment of the present invention may be input through the user terminal 200 and provided to the automatic banner creation device 100 . The target platform information can be provided to the platform information management unit 130, and the user information is the user information, such as title, font, relative strength of automatic location (upper, middle, lower), font size, responsive application, font sentence , learning model propensity, fixed position, etc. can be selected and input.

In addition, the user can designate in advance the headline area 1 where the banner sentence is located compared to the banner area 2, and the automatic banner information processing unit 150 automatically displays the headline area 1 and the headline area 1 according to the user information input. Combination information of banner sentence elements maximizing a promotion effect within a limited range of elements may be output.

4 and 5 illustrate banner sentences that are optimized according to product detailed information according to the optimized result. It is preferable to separate and exclude industry jargon from product detailed information as a whole, and season, attribute, color It may be desirable that combinations of words such as etc. are simply arranged and placed. In addition, even if the user does not have such professional knowledge, by using the automated banner design according to natural language learning, the user can always receive a non-uniform banner design with an optimized promotion effect, which is important for platform service operation. Convenience can be maximized.

The method according to the present invention described above may be produced as a program to be executed on a computer and stored in a computer-readable recording medium. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, and magnetic tape. , floppy disks, and optical data storage devices.

The computer-readable recording medium is distributed to computer systems connected through a network, so that computer-readable codes can be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the method can be easily inferred by programmers in the technical field to which the present invention belongs.

In addition, although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the specific embodiments described above, and the technical field to which the present invention belongs without departing from the gist of the present invention claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or perspective of the present invention.

Claims (4)

In the method of automatically producing a shopping mall banner sentence,
Collecting banner learning element information from a pre-built platform network;
Based on the banner learning element information, building a natural language learning model by learning combinational relationship information between banner sentence elements corresponding to a promotion effect; and
Applying target platform information to the learning model and outputting banner sentence information having an optimized promotion effect corresponding to the target platform information.
How to automatically create shopping mall banner sentences. According to claim 1,
The banner learning element information,
The banner learning element information includes banner sentence element classification information;
The banner sentence element classification information includes at least one of a trend keyword, a product category keyword, an abbreviation keyword, a new word keyword, an attribute keyword, a tag keyword, and a product name text identified from the banner sentence text.
How to automatically create shopping mall banner sentences. According to claim 1,
The target platform information includes target product detailed information to be provided to customers on the target platform;
The automatic banner processing unit extracts the banner learning element information from the target product detailed information, applies it to the learning model, and obtains combination relation information between the banner sentence elements having the highest promotion effect from the output of the learning model. further comprising
How to automatically create shopping mall banner sentences. According to claim 1,
Further comprising providing the banner sentence information to a platform management server, receiving feedback information from the platform management server, and processing renewal learning of the natural language learning model
How to automatically create shopping mall banner sentences.

Images