KR102183143B1 - Method and apparatus for determining recommended color of the target product based on big data - Google Patents

Abstract

Disclosed is a method for determining a recommended color of a target product based on big data to provide a systematic and reliable recommendation result. According to the present invention, the method comprises the following steps of, by a computing device: acquiring metadata of a target product; based on sale result data of a product group having the same metadata as that of the target product for a predetermined period, determining a first cluster placed in a color space based on distribution of color vectors of products having a higher sale result than a reference value within the product group; analyzing images posted in a predetermined social network service platform for a predetermined period to extract the images of the products having the same metadata as that of the target product from the posted images and extracting colors appeared in the images of the products to determine a second cluster placed in the color space; overlapping the first and second clusters to determine a candidate color cluster; and determining a recommended color of the target product based on distribution of the candidate color cluster.

Description

Method and device for determining the recommended color of the target product based on big data {METHOD AND APPARATUS FOR DETERMINING RECOMMENDED COLOR OF THE TARGET PRODUCT BASED ON BIG DATA}

The following description relates to a method and apparatus for determining a recommended color of a target product at the stage of designing a target product. More specifically, the method and apparatus for determining the recommended color of the target product based on the sales performance information of the product related to the target product, analysis information of the images posted on the online social network service platform, and trend color information issued by a professional organization Is initiated.

As the performance of products has recently been leveled upward, manufacturing companies are putting a very large weight on design to increase product competitiveness. Consumers prefer products with high aesthetics, and accordingly, the design of the product has a great influence on the competitiveness of the product. In product design, not only the appearance but also the color of the product is also a very important factor.

Therefore, manufacturers spend enormous budgets on color-determining products. However, there is no systematic technology or methodology to determine the appropriate color for a product. Currently, the process of determining the color of a product is completely dependent on the personal sense of the expert.

Therefore, when experts have different opinions on the color of a product, the company has difficulty in determining the color of the product. In addition, if a company releases a product designed in an inappropriate color, there is a possibility that the consumer will not respond well, resulting in a huge loss.

According to at least one embodiment, a method and apparatus for determining a recommended color of a target product based on big data analysis are disclosed.

According to an aspect, a method of determining a recommended color of a target product based on big data is disclosed. The disclosed method includes the steps of: (a) obtaining, by a computing device, metadata of a target product; (b) The computing device, based on the sales performance data of a product group having the same metadata as the target product for a predetermined period of time, the color space (color determining a first cluster located in space); (c) The computing device extracts images of products having the same metadata as the target product from the posted images by analyzing the images posted on the social network service platform specified in advance for a predetermined period, and the images of the products Extracting colors appearing in the fields and determining a second cluster located in the color space; (d) determining a candidate color cluster by overlapping the first cluster and the second cluster; And (e) determining a recommended color of the target product based on the distribution of the candidate color cluster.

The metadata of the target product may include information on a manufacturing brand of the target product, a category of the target product, a user gender of the target product, and a user age of the target product.

The step (b) may include: (b-1) obtaining sales data of the first product included in the product group having the same meta data as the target product; (b-2) determining a first weight of each of the products based on sales data of the first product; (b-3) determining a second weight of the specific color in the first product according to a ratio of the specific color in each of the products; (b-4) determining a distribution of a first cluster element formed at a position where the specific color vector is indicated in the color space based on the first weight and the second weight; And (b-5) determining the distribution of a plurality of cluster elements by repeating steps (b-1) to (b-4) for a plurality of products belonging to the product group, and overlapping the plurality of cluster elements 1 cluster is determined, wherein the distribution of the elements of the first cluster follows a multivariate Gaussian normal distribution, and the scale of the multivariate Gaussian distribution is determined according to the first weight, and the multivariate The variance of the Gaussian distribution may be determined according to the second weight.

The step (c) includes at least one of (c-1) a profile of a publisher who posted the posted images, tag information assigned to the posted images, a posting date of the posted images, and a number of views of the posted images. Filtering the posted images based on one; (c-2) identifying an image region of a product related to the target product using an artificial neural network from at least one of the images extracted by filtering; (c-3) extracting information on a representative color appearing in the image area of the product using an artificial neural network; (c-4) determining second cluster elements based on a vector distribution corresponding to a representative color in the color space; And (c-5) determining the second cluster by overlapping second cluster elements obtained from a plurality of extracted images.

The step (e) may include: (e-1) determining a distribution density of the candidate color cluster based on the distribution density of the first cluster and the distribution density of the second cluster; And (e-2) determining a center coordinate of the candidate color cluster based on the distribution density of the candidate color cluster, and determining a recommended color of the target product based on the center coordinate.

The step (e) includes: (e-3) obtaining information on at least one trend color designated by the professional agency from a report issued by the professional agency; And (e-4) correcting the recommended color of the target product based on the center coordinate of the candidate color cluster and vectors corresponding to the trend color.

According to at least one embodiment, the computing device may output a systematic and highly reliable recommendation result by determining a recommended color of a target product based on big data analysis. According to at least one embodiment, the computing device forms a first cluster based on sales information of a product group having the same meta data as the target product, and determines a recommended color of the target product based on the first cluster. Recommended colors can be determined to increase sales. According to at least one embodiment, the computing device may analyze images disclosed on a social network service platform to form a second cluster, thereby determining a color suitable for consumers' taste as a recommended color of a target product. According to at least one embodiment, the computing device corrects a recommended color of a target product based on a report issued by a specialized institution, so that information on a color that will be fashionable in the future may be reflected in the recommended color of the target product.

Fig. 1 is a block diagram showing a computing device that performs a method of determining a recommended color of a target product according to an exemplary embodiment.
Fig. 2 is a flow chart showing a method of determining a recommended color of a target product according to an exemplary embodiment.
3 is a flowchart illustrating a process in which step S120 of FIG. 2 is performed.
4 is a conceptual diagram illustrating sales data acquired by a computing device and metadata of a target product by way of example.
5 is a conceptual diagram illustrating a first cluster determined by a computing device by way of example.
6 is a conceptual diagram showing a change in a Gaussian distribution according to a change in a variance value.
7 is a conceptual diagram showing a scale change of a Gaussian distribution.
8 is a flowchart illustrating a process in which step S130 of FIG. 2 is performed.
9 is a conceptual diagram illustrating a second cluster determined by the computing device by way of example.
10 is a flowchart illustrating a process in which step S150 of FIG. 2 is performed.
11 is a conceptual diagram illustrating a vector of a recommended color determined based on a distribution of a candidate color cluster and a distribution of a candidate color cluster.
12 is a flowchart illustrating another example of a process in which step S150 of FIG. 2 is performed.
13 is a conceptual diagram illustrating that a computing device corrects a recommended color of a target product based on a vector of a trend color and a center coordinate of a candidate color cluster.

Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only, and may be changed in various forms and implemented. Accordingly, the embodiments are not limited to a specific disclosure form, and the scope of the present specification includes changes, equivalents, or substitutes included in the technical idea.

Although terms such as first or second may be used to describe various components, these terms should be interpreted only for the purpose of distinguishing one component from other components. For example, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component.

When a component is referred to as being "connected" to another component, it is to be understood that it may be directly connected or connected to the other component, but other components may exist in the middle.

Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present specification, terms such as "comprise" or "have" are intended to designate that the described features, numbers, steps, actions, components, parts, or combinations thereof exist, but one or more other features, numbers, and steps It is to be understood that it does not preclude the possibility of addition or presence of, operations, components, parts, or combinations thereof.

Unless otherwise defined, all terms, including technical or scientific terms, used herein have the same meaning as commonly understood by one of ordinary skill in the relevant technical field. Terms as defined in a commonly used dictionary should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and should not be interpreted as an ideal or excessively formal meaning unless explicitly defined in this specification. Does not.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. In the description with reference to the accompanying drawings, the same reference numerals are assigned to the same components regardless of the reference numerals, and redundant descriptions thereof will be omitted.

Fig. 1 is a block diagram showing a computing device 100 that performs a method of determining a recommended color for a target product according to an exemplary embodiment.

Referring to FIG. 1, the computing device 100 may include a communication unit 110 and a processor 120. The communication unit 110 may communicate with an Internet network or another device through a network. The processor 120 controls the communication unit 110 and may process information obtained from the communication unit 110. The processor 120 may perform an operation necessary to determine the recommended color of the target product.

Computing device 100 is a device that may include typical computer hardware (eg, computer processors, memory, storage, input and output devices, and other components of conventional computing devices; electronic communication devices such as routers, switches, etc.; networks; Combination of electronic information storage systems such as network-attached storage (NAS) and storage area network (SAN)) and computer software (i.e., instructions that cause a computing device to function in a specific way) It may be to achieve the desired system performance by using.

In addition, the above-described network includes a wired network, a wireless network, and the like as a network connecting the components. The network may be a closed network such as a local area network (LAN), a wide area network (WAN), or an open network such as the Internet. The Internet is the TCP/IP protocol and several services that exist in the upper layer, namely HTTP (HyperText Transfer Protocol), Telnet, FTP (File Transfer Protocol), DNS (Domain Name System), SMTP (Simple Mail Transfer Protocol), SNMP ( It refers to a global open computer network structure that provides Simple Network Management Protocol), Network File Service (NFS), and Network Information Service (NIS).

Fig. 2 is a flow chart showing a method of determining a recommended color of a target product according to an exemplary embodiment.

Referring to FIG. 2, in step S110, the computing device 100 may acquire metadata of a target product. The target product may be a product to be designed. The computing device 100 may determine a recommended color of a target product and provide the determined information to the user. Here, the user may be a person who intends to design, manufacture, or sell a target product. The computing device 100 may acquire metadata of a target product based on user input data.

The metadata of the target product may include information on the manufacturing brand of the target product, the category of the target product, the gender of the user of the target product, and the age of the user of the target product. However, the above-described information is merely exemplary, and embodiments are not limited thereto. For example, the metadata of the target product may further include information on the price of the target product and the distribution channel of the target product.

In step S120, the computing device 100 may determine a first cluster located in a color space based on sales performance data of a product group having the same metadata as the target product. The first cluster may include a plurality of first cluster elements. The color space may be a space concept in which a color indicator is displayed in n-dimensional (mainly three-dimensional). The color space may be any one of an RGB color space, a CMYK color space, an HSV color space, and a CIE color space, but embodiments are not limited thereto. Hereinafter, an exemplary embodiment will be described using an RGB color space as an example. However, the embodiments are not limited to the following description, and it is obvious to a typical practice that the embodiments of the present invention may be applied to a color space other than an RGB color space.

3 is a flowchart illustrating a process in which step S120 of FIG. 2 is performed.

Referring to FIG. 3, in step S121, the computing device 100 may acquire sales data of a first product belonging to a product family having the same meta data as the target product. The sales data of the first product may include a sales amount of the first product, information on a design color of the first product, and the like.

4 is a conceptual diagram illustrating sales data acquired by the computing device 100 and metadata of a target product as an example.

Referring to FIG. 4, metadata of a target product acquired by the computing device 100 may include information on a category of a target product, a user's gender, a user's age, and a brand. The computing device 100 may extract a product group having the same meta data as the target product. Since the computing device 100 refers to product sales data of the same category as the target product, it may be easier to extract information on a color suitable for application to the target product. In addition, the computing device 100 may refer to sales data of a product having the same user gender, age, and brand as the target product. Accordingly, the brand's own sales color and the user's tendency to be sold may be reflected in determining the recommended color of the target product.

The sales information DB 30 may include columns C2 to C5 including information on metadata of the sold product. The computing device 100 may refer to columns C2 to C5 of the sales information DB 30 to extract a product group having the same meta data as the target product. The computing device 100 may check the color of each product belonging to the product family by referring to column C6. The computing device 100 may check the proportion of a specific color in each product by referring to column C7. The computing device 100 may check the sales volume of each product by referring to column C8.

Referring back to FIG. 3, in step S122, the computing device 100 may determine a first weight of each product by checking the sales amount of each product belonging to the product family. For example, as the sales volume of the product increases, the computing device 100 may assign a higher first weight to the product. The first weight may be related to a distribution density scale of cluster elements to be described later.

In step S123, the computing device 100 may determine a second weight of a specific color in the corresponding product according to a ratio of the specific color in each of the products belonging to the product family. For example, the computing device 100 may give the second weight higher as the weight occupied by a specific color in the corresponding product increases. The second weight may be related to a variance parameter that determines the distribution shape of the cluster element to be described later.

In operation S124, the computing device 100 may determine the distribution of the first cluster element based on the first weight and the second weight. The computing device 100 may repeat steps S122 to S124 for products having the same meta data as the target product. For example, the computing device 100 may determine four first cluster elements with reference to sales information of basketball shoes A, basketball shoes B, basketball shoes C, and basketball shoes D having the same metadata as the target product shown in FIG. 4. The distribution of the four first cluster elements may be determined according to the main colors of basketball shoes A, basketball shoes B, basketball shoes C, and basketball shoes D, the color ratio of the main shoes, and sales volume.

In operation S125, the computing device 100 may determine a first cluster by overlapping a plurality of first cluster elements.

5 is a conceptual diagram illustrating a first cluster 40 determined by the computing device 100 by way of example.

In Fig. 5, the first cluster 40 is shown on the RGB color space. However, embodiments are not limited thereto, and the first cluster 40 may be displayed in a CMYK color space, an HSV color space, or a CIE color space.

The first cluster 40 may be determined by overlapping a plurality of first cluster elements 41, 42, 43, 44. The distribution of the first cluster elements 41, 42, 43, 44 may be determined based on sales data of products belonging to the same product group as the target product.

For example, the distribution of the first cluster element 41 may be determined by sales data of basketball shoes A in FIG. 4. The center of the first cluster element 41 may correspond to a coordinate indicated by a vector corresponding to the primary color of basketball shoe A (R243:G91:B58). The distribution of the first cluster element 41 may be determined according to the proportion of the main color in basketball shoes A and the sales volume of basketball shoes A.

For example, the distribution of each of the first cluster elements 41, 42, 43, 44 may follow a multivariate Gaussian normal distribution. For example, the distribution of the first cluster elements 41, 42, 43, 44 in the RGB space may follow a 3D Gaussian distribution.

6 is a conceptual diagram showing a change in a Gaussian distribution according to a change in a variance value. In FIG. 6, for convenience, Gaussian distributions for one-dimensional variables are shown.

Referring to FIG. 6, the smaller the variance of the Gaussian distribution, the higher the density of the center distribution. The computing device 100 may determine a variance value that determines the distribution shape of the first cluster elements 41, 42, 43, 44 based on the above-described second weight. For example, the computing device 100 may set a variance value of the first cluster element to be smaller as the second weight increases. That is, the computing device 100 may reduce the variance value of the first cluster element as the ratio occupied by a specific color in the product increases, so that the distribution of the first cluster element is concentrated in the center.

Accordingly, the computing device 100 may make the distribution of the first cluster 41 obtained from the sales information of the basketball shoe A concentrated in the center. As the distribution of the first cluster 41 is concentrated in the center, the effective volume of the first cluster 41 may decrease. Therefore, in FIG. 5, the effective volume of the first cluster 41 corresponding to the basketball shoe A having the highest main color ratio may be displayed as the smallest. The computing device 100 may concentrate cluster elements around a specific color as a proportion of a specific color in a sale product is large, so that a color having a large proportion of the product is reflected in a recommended color of a target product with a high weight.

7 is a conceptual diagram showing a scale change of a Gaussian distribution.

Referring to FIG. 7, the size of the distribution density value can be changed by adjusting the scale of two Gaussian graphs 62 and 64 having the same variance. The computing device 100 may determine a scale value for determining the distribution shape of the first cluster elements 41, 42, 43, 44 based on the above-described first weight. For example, the computing device 100 may set the scale value of the first cluster element as the first weight increases. That is, as the sales volume of the product increases, the computing device 100 may set the scale of the first cluster element corresponding to the product to be larger.

Accordingly, the computing device 100 may set the distribution scale of the first cluster 42 obtained from the sales information of the basketball shoe B to be larger than the distribution scale of the first clusters 41, 43, 44 obtained from other products. have. Accordingly, in FIG. 5, the distribution density of the second cluster 42 corresponding to the basketball shoe B with the highest sales volume may be displayed as the highest. The computing device 100 may increase the distribution scale of the cluster element as the sales volume of the product increases, so that the color of the product that has been sold a lot is reflected in the recommended color of the target product with a high weight.

Referring back to FIG. 2, images posted on a predetermined social network service platform for a predetermined period of time in step S130 may be analyzed. As an example, the predetermined period may be set within one year from the expected sales time of the target product, but the embodiment is not limited thereto, and it is obvious that a person skilled in the art can easily change the predetermined period as necessary. The social network service platform may include a platform for users to post images, such as Facebook and Instagram.

The computing device 100 may analyze images disclosed on a social network platform and extract images of products having the same meta data as the target product from the images. The computing device 100 may determine a second cluster by identifying a display area of a product having the same metadata as a target product from the images using an artificial neural network, and extracting color information of the corresponding product.

8 is a flowchart illustrating a process in which step S130 of FIG. 2 is performed.

Referring to FIG. 8, in step S131, the computing device 100 may filter images posted on a social network service platform. For example, the computing device 100 may extract images published within the predetermined period based on the posting date of the images. The computing device 100 may extract images that are expected to be highly relevant to the target product from tag information attached to the posted image. For example, when the target product is a basketball shoe, the computing device 100 may extract images tagged with a tag such as #basketball, #basketball shoe, #exercise, #dunk, #competition, and the like.

The computing device 100 may extract images expected to be highly relevant to a target product based on a profile of a publisher who posted the images. For example, when the target product is basketball shoes, the computing device 100 may extract images posted by a publisher who has an age and gender who mainly uses basketball shoes, or a publisher who is identified as having high interest in basketball. The computing device 100 may extract images based on the number of times the posted image is viewed. For example, the computing device 100 may extract an image in which the number of views of the posted image is greater than the reference number.

In step S132, the computing device 100 may identify an image area of a product related to the target product from each of the extracted images. For example, the computing device 100 may identify the image area of the product using a Grabcut algorithm or a background removal algorithm. As another example, the computing device 100 may identify an image area of a product using an artificial neural network such as a convolutional neural network (CNN).

In step S133, the computing device 100 may extract color information of an image area of a product using an artificial neural network. For example, the computing device 100 may extract color information of an image area of a product using a CNN or Grad-weighted Class Activation Mapping (GAP) combined with CNN and Global Average Pooling (GAP).

In step S134, the computing device 100 may determine the second cluster element based on color information of the image area of the product. The distribution of the second cluster elements may follow a multivariate Gaussian distribution. The computing device 100 may set a larger distribution scale of the second cluster element as the number of times the image corresponding to the corresponding second cluster element is viewed increases. The computing device 100 may set the variance of the distribution of the second cluster element to be smaller as the ratio of the representative color of the product in the image corresponding to the corresponding second cluster element increases.

The computing device 100 may determine a plurality of cluster elements by extracting a plurality of product images related to a target product from the social network platform.

In step S135, the computing device 100 may form a second cluster by overlapping the second cluster elements.

9 is a conceptual diagram illustrating a second cluster 50 determined by the computing device 100 by way of example.

Referring to FIG. 9, the second cluster 50 may include a plurality of second cluster elements. In FIG. 9, for convenience, the number of second cluster elements included in the second cluster 50 is shown as seven, but the embodiment is not limited thereto. As the number of social platform posted images extracted and analyzed by the computing device 100 increases, the number of second cluster elements may increase.

The computing device 100 may determine the distribution of each of the second cluster elements based on the number of views of the posted image, the ratio of the representative color in the product in the posted image, and the like.

Referring back to FIG. 2, in step S140, the computing device 100 may determine a candidate color cluster by overlapping the first cluster and the second cluster. In operation S150, the computing device 100 may determine a recommended color of the target product based on the distribution of the candidate color cluster.

10 is a flowchart illustrating a process in which step S150 of FIG. 2 is performed. 11 is a conceptual diagram illustrating a vector of a recommended color determined based on a distribution of a candidate color cluster and a distribution of a candidate color cluster by way of example.

10 and 11, in step S151, the computing device 100 may determine a distribution density of a candidate color cluster. The computing device 100 may determine the distribution density of the candidate color cluster based on the distribution density of the first cluster 40 and the distribution density of the second cluster 50.

For example, the computing device 100 may determine the distribution density of the candidate color cluster by using Equation 1.

는 색 공간 좌표 (r, g, b)에서 후보 색 클러스터의 분포 밀도를 나타내고,

는 제1 클러스터 요소의 인덱스이며

는 제1 클러스터에 포함된 제1 클러스터 요소들의 인덱스 집합을 나타내고,

는 색 공간의 좌표 (r, g, b)에서

번째 제1 클러스터 요소의 분포 밀도를 나타낸다. 또한,

은 제2 클러스터 요소의 인덱스이며

은 제2 클러스터에 포함된 제2 클러스터 요소들의 인덱스 집합을 나타내고,

는 색 공간의 좌표 (r, g, b)에서

번째 제2 클러스터 요소의 분포 밀도를 나타낸다.In Equation 1

Denotes the distribution density of candidate color clusters in color space coordinates (r, g, b),

Is the index of the first cluster element

Represents an index set of first cluster elements included in the first cluster,

Is at coordinates (r, g, b) in color space

Represents the distribution density of the first cluster element. In addition,

Is the index of the second cluster element

Represents an index set of second cluster elements included in the second cluster,

Is at coordinates (r, g, b) in color space

Represents the distribution density of the second second cluster element.

Referring to Equation 1, the distribution density of the candidate color cluster at the coordinates of the color space (r, g, b) is the sum of the distribution density of the first cluster elements and the distribution density of the second cluster elements at the coordinates of the corresponding color space. It can be determined based on the sum. However, the embodiment is not limited to Equation 1. For example, a weight for the sum of distribution densities of the first cluster elements and a weight for the sum of distribution densities of the second cluster elements may be different from each other.

In operation S152, the computing device 100 may determine the center coordinates of the candidate color cluster based on the distribution density of the candidate color cluster. The computing device 100 may determine the center coordinates of the candidate color cluster using a principle similar to the principle of calculating the center of mass. As an example, the computing device may determine the center coordinates of the candidate color cluster using Equation 2.

는 색 공간의 좌표 (r, g, b)에서 후보 색 클러스터의 분포 밀도를 나타내고,

은 후보 색 클러스터의 중심 좌표의 위치를 나타내는 벡터이다.In Equation 2

Denotes the distribution density of candidate color clusters at coordinates (r, g, b) in the color space,

Is a vector indicating the location of the center coordinates of the candidate color cluster.

를 계산할 수 있다. 컴퓨팅 장치(100)는

에 대응하는 색을 타깃 제품의 추천 색으로 결정할 수 있다. 하지만, 실시예가 이에 제한되는 것은 아니다. 예를 들어, 컴퓨팅 장치(100)는 전문 기관의 리포트 정보에 기초하여 타깃 제품의 추천 색을 보정할 수 있다.The computing device 100 uses Equation 2 to determine the location of the center coordinates of the candidate color cluster.

Can be calculated. The computing device 100

The color corresponding to the can be determined as the recommended color of the target product. However, the embodiment is not limited thereto. For example, the computing device 100 may correct a recommended color of a target product based on report information of a specialized agency.

This is the flow chart that was issued. 13 is a conceptual diagram illustrating that the computing device 100 corrects a recommended color of a target product based on a vector of a trend color and a center coordinate of a candidate color cluster. In the description of the embodiments of FIGS. 12 and 13, content overlapping with FIGS. 10 and 11 will be omitted.

12 and 13, in step S153, the computing device 100 may obtain information on at least one trend color from a report issued by a specialized agency. The computing device 100 may obtain information on at least one trend color by crawling a report issued by a specialized agency. 13 illustrates a case where the computing device 100 obtains information on a plurality of trend colors, the embodiment is not limited thereto. For example, the computing device 100 may obtain information on one trend color.

)를 결정할 수 있다. 컴퓨팅 장치(100)는 후보 색 클러스터의 중심 좌표에 대응하는 벡터(

)와 벡터(

) 사이의 거리, 벡터(

)과 제1 트렌드 색에 대응하는 벡터(TR1) 사이의 거리, 벡터(

)과 제2 트렌드 색에 대응하는 벡터(TR2) 사이의 거리 및 벡터(

)과 제1 트렌드 색에 대응하는 벡터(TR2) 사이의 거리의 합이 최소가 되도록 벡터(

)을 결정할 수 있다. 컴퓨팅 장치(100)는 벡터(

)으로부터 타깃 제품의 추천 색을 결정할 수 있다. 하지만, 실시예가 이에 제한되는 것은 아니다. 컴퓨팅 장치(100)는 상술한 거리들의 합을 계산하는 과정에서 각 거리의 가중치를 다르게 설정할 수도 있다. 컴퓨팅 장치(100)가 트렌드 색에 대한 정보를 추가적으로 참조하여 타깃 제품의 추천 색을 보정함으로써 전문기관의 리포트 결과가 타깃 제품의 추천 색 결정에 함께 반영될 수 있다.In step S154, the computing device 100 may correct the recommended color of the target product based on the vectors TR1, TR2, and TR3 corresponding to the center coordinate of the candidate color cluster and the trend color. The computing device 100 is a vector corresponding to the recommended color of the corrected target product (

) Can be determined. The computing device 100 includes a vector corresponding to the center coordinates of the candidate color cluster (

) And vector (

) Distance between, vector(

) And the vector (TR1) corresponding to the first trend color, the vector (

) And the vector (TR2) corresponding to the second trend color (

) So that the sum of the distances between the vector (TR2) corresponding to the first trend color is minimum.

) Can be determined. Computing device 100 is a vector (

), the recommended color of the target product can be determined. However, the embodiment is not limited thereto. The computing device 100 may set different weights for each distance in the process of calculating the sum of the aforementioned distances. The computing device 100 may additionally refer to information on the trend color to correct the recommended color of the target product, so that the report result of the specialized agency may be reflected in the determination of the recommended color of the target product.

In the above, a method and an apparatus for determining a recommended color of a target product according to exemplary embodiments have been described with reference to FIGS. 1 to 13. According to at least one embodiment, the computing device 100 may output a systematic and highly reliable recommendation result by determining a recommended color of a target product based on big data analysis. According to at least one embodiment, the computing device 100 forms a first cluster based on sales information of a product group having the same meta data as the target product, and determines a recommended color of the target product based on the first cluster. Recommended colors can be determined to increase sales of the target product. According to at least one embodiment, the computing device 100 may analyze images disclosed on a social network service platform to form a second cluster, thereby determining a color suitable for consumers' taste as a recommended color of a target product. According to at least one embodiment, the computing device 100 corrects a recommended color of a target product based on a report issued by a specialized institution, so that information on a color that will be fashionable in the future may be reflected in the recommended color of the target product. .

The embodiments described above may be implemented as a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, the devices, methods, and components described in the embodiments include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). array), programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions, such as one or more general purpose computers or special purpose computers. The processing device may execute an operating system (OS) and one or more software applications executed on the operating system. In addition, the processing device may access, store, manipulate, process, and generate data in response to the execution of software. For the convenience of understanding, although it is sometimes described that one processing device is used, one of ordinary skill in the art, the processing device is a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it may include. For example, the processing device may include a plurality of processors or one processor and one controller. In addition, other processing configurations are possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of these, configuring the processing unit to behave as desired or processed independently or collectively. You can command the device. Software and/or data may be interpreted by a processing device or to provide instructions or data to a processing device, of any type of machine, component, physical device, virtual equipment, computer storage medium or device. , Or may be permanently or temporarily embodyed in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like alone or in combination. The program instructions recorded on the computer-readable medium may be specially designed and configured for the embodiment, or may be known and usable to those skilled in the computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -A hardware device specially configured to store and execute program instructions such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of the program instructions include not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

As described above, although the embodiments have been described by the limited drawings, a person of ordinary skill in the art can apply various technical modifications and variations based on the above. For example, the described techniques are performed in a different order from the described method, and/or components such as a system, structure, device, circuit, etc. described are combined or combined in a form different from the described method, or other components Alternatively, even if substituted or substituted by an equivalent, an appropriate result can be achieved.

Claims (6)

In the method of determining the recommended color of the target product based on big data,
(a) obtaining, by a computing device, meta data of a target product;
(b) The computing device, based on the sales performance data of a product group having the same manufacturing brand and category as the target product for a predetermined period, based on the distribution of color vectors of products whose sales performance is more than a reference value in the product group. determining a first cluster located in the (color space);
(c) The computing device extracts images of products that are highly relevant to the target product from the posted images by analyzing the images posted on the social network service platform designated in advance for a predetermined period, and extracts images of the products. Extracting the displayed colors and determining a second cluster located in the color space;
(d) determining a candidate color cluster by overlapping the first cluster and the second cluster; And
(e) determining a recommended color of the target product based on the distribution of the candidate color cluster,
The step (b),
(b-1) obtaining sales data of a first product having the same manufacturing brand and category as the target product; (b-2) determining a first weight of each of the products based on sales data of the first product; (b-3) determining a second weight of the specific color in the first product according to a ratio of the specific color in each of the products; (b-4) determining a distribution of a first cluster element formed at a position where the specific color vector is indicated in the color space based on the first weight and the second weight; And (b-5) determining the distribution of a plurality of cluster elements by repeating steps (b-1) to (b-4) for a plurality of products belonging to the product group, and overlapping the plurality of cluster elements 1 determining a cluster, wherein the distribution of the elements of the first cluster follows a multivariate Gaussian normal distribution, and the scale of the multivariate Gaussian distribution is determined according to the first weight, and the multivariate The variance of the Gaussian distribution is determined according to the second weight,
The step (c) includes (c-1) a profile of the publisher who posted the posted images, tag information assigned to the posted images, and the target product based on the posting date of the posted images. Extracting images, and extracting images having a viewing count higher than a reference count among images associated with the target product; (c-2) identifying an image region of a product related to the target product using an artificial neural network from at least one of the extracted images; (c-3) extracting information on a representative color appearing in the image area of the product using an artificial neural network; (c-4) determining second cluster elements based on a vector distribution corresponding to a representative color in the color space; And (c-5) determining the second cluster by overlapping second cluster elements obtained from a plurality of extracted images,
The step (e) may include: (e-1) determining a distribution density of the candidate color cluster based on the distribution density of the first cluster and the distribution density of the second cluster; And (e-2) determining a center coordinate of the candidate color cluster based on a distribution density of the candidate color cluster, and determining a recommended color of a target product based on the center coordinate, wherein the candidate color cluster The center coordinate of is obtained by integrating a product of the density of each region in the candidate color cluster and the color space position vector for each region. The method of claim 1,
The metadata of the target product includes information on a manufacturing brand of the target product, a category of the target product, a user gender of the target product, and a user age of the target product. delete delete delete The method of claim 1,
The step (e),
(e-3) obtaining information on at least one trend color designated by the professional agency from the report issued by the professional agency;
(e-4) a method of determining a recommended color of a target product, comprising correcting a recommended color of the target product based on the center coordinates of the candidate color cluster and vectors corresponding to the trend color.

Images