KR20220079249A - Regular delivery method and system that reflect the result of forecasting consumer demand - Google Patents

Abstract

The present invention predicts consumer demand and enables regular delivery corresponding thereto, and the collection unit a collecting step of collecting information of a plurality of buyers who purchase products from a plurality of external servers including at least one product sales platform and users using a plurality of external servers; A database construction step including a database generated by accumulating the information of the buyer and the user collected from an external server; and an analysis step of performing AI analysis based on the database by the AI unit, wherein the analysis step analyzes information of buyers and users, calculates information that meets a predetermined demand standard, and responds to the published information A regular delivery method is provided that allows the product to be delivered on a regular basis.

Description

A regular delivery method and system that predicts consumer demand and reflects it

The present invention predicts consumer demand and enables regular delivery corresponding thereto, and relates to a regular delivery method reflecting consumer demand prediction.

In general, regular delivery can be made based on consumer information. For example, based on the information permitted by the consumer or the information surveyed by the consumer, the corresponding product will be delivered regularly. In this case, the needs of potential consumers who are interested in consumer preferences or trends and information on regular delivery may be different from each other, and it is impossible to respond immediately to changing consumer preferences, resulting in lower satisfaction with regular delivery. Conventionally, methods for performing analysis through statistical data related to sales items have been used, but it is difficult to comprehensively analyze with existing data because the disclosure cycle of statistical data related to sales items is long. In particular, because of this, it is impossible to use it as a forecast within a relatively short period such as within a week.

Therefore, it is required to be able to secure and regularly provide products through the results of predicting consumer demand by collecting data that can be analyzed in a shorter time period for data collection.

Republic of Korea Patent Publication No. 2020-0055353 (May 20, 2020)

An embodiment of the present invention aims to quantify the demand for a specific product in which the reaction and preference of consumers are reflected in real time and apply it to a regular delivery product.

An embodiment of the present invention aims to quantify consumer demand to provide a product at a lower cost from a place of origin.

The present invention predicts consumer demand and enables regular delivery corresponding thereto, and a plurality of buyers and a plurality of buyers who purchase products from a plurality of external servers including at least one product sales platform by a collection unit a collection step of collecting user information using an external server; A database construction step including a database generated by accumulating the information of the buyer and the user collected from an external server; and an analysis step of performing AI analysis based on the database by the AI unit, wherein the analysis step analyzes information of buyers and users, calculates information that meets a predetermined demand standard, and responds to the published information A regular delivery method is provided that allows the product to be delivered on a regular basis.

In addition, the user information may reflect the consumer demand prediction, including buzz data in which the exposure amount is increased by more than a predetermined reference value from the external server.

In addition, the buyer's information, static data including the buyer's profile; and dynamic data including purchase history, consumption content, access frequency, and product review in the product sales platform by the purchaser.

In addition, product reviews can be quantified numerically by being collected in the database unit and divided into keywords meaning positive or negative by the machine-learning AI unit, and adding or deducting points to each keyword.

In addition, the AI unit may apply a random forest technique to the information or database collected from the user to extract key points, and apply a principal component analysis (PCA) technique to the key points to filter it.

In addition, the database unit may further include geographic information.

In addition, the predetermined demand criterion may include a preference for a product, a product type, and a regular delivery cycle of the product.

A database created by collecting information of a plurality of buyers who purchase products from a plurality of external servers including at least one product sales platform and users using a plurality of external servers, and accumulating the collected information of buyers and users A database unit comprising a; AI unit that performs AI analysis based on the database included in the database unit; and

Including a distribution management unit that receives delivery information from the AI unit and delivers to the buyer, the AI unit analyzes buyer and user information, calculates information that meets predetermined demand standards, and A regular delivery system is provided, allowing the product to be delivered to the distribution management unit.

In addition, the user information may reflect the consumer demand prediction, including buzz data in which the exposure amount is increased by more than a predetermined reference value from the external server.

In addition, the buyer's information, static data including the buyer's profile; and dynamic data including purchase history, consumption content, access frequency, and product review in the product sales platform by the purchaser.

In addition, product reviews can be quantified numerically by being collected in the database unit and divided into keywords meaning positive or negative by the machine-learning AI unit, and adding or deducting points to each keyword.

In addition, the AI unit may apply a random forest technique to the information or database collected from the user to extract key points, and apply a principal component analysis (PCA) technique to the key points to filter it.

In addition, the database unit may further include geographic information.

In addition, the predetermined demand criterion may include a preference for a product, a product type, and a regular delivery cycle of the product.

According to an embodiment of the present invention, there is provided a method of collecting and quantifying buzz data generated from an accessible server in order to quantify the demand for a specific product in which the consumer's reaction and preference are reflected in real time and apply it to regular delivery products. do.

According to an embodiment of the present invention, there is provided a method for comprehensively analyzing product price fluctuations and demand through AI analysis in order to quantify consumer demand and provide products at a lower cost from the origin.

1 is a process diagram showing a process in which regular delivery is performed according to an embodiment of the present invention;
Figure 2 is a flow chart in which electric delivery is made according to an embodiment of the present invention.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. However, this is merely an example, and the present invention is not limited thereto.

In the description of the present invention, if it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted. And, the terms to be described later are terms defined in consideration of functions in the present invention, which may vary according to intentions or customs of users and operators. Therefore, the definition should be made based on the content throughout this specification.

The technical spirit of the present invention is determined by the claims, and the following examples are only one means for efficiently explaining the technical spirit of the present invention to those of ordinary skill in the art to which the present invention belongs.

The regular delivery method, which is an embodiment of the present invention, which will be described below, sells a specific product, and may be meat and processed meat products, for example. Through this example, the analysis process of the AI unit, which is one of the characteristics of the present invention, can be described. In the case of meat, which is the product, it is not possible to increase the supply by artificially producing the product, and the quantity demanded is a product that increases or decreases based on a certain quantity. It can be analyzed to explain how to determine the time of purchase of meat, the type of product regularly delivered to the buyer, the amount of product, and the delivery time.

1 is a process diagram illustrating a process in which regular delivery is performed according to an embodiment of the present invention, and FIG. 2 is a flowchart illustrating a process in which regular delivery is performed according to an embodiment of the present invention.

First, referring to FIG. 1 , a user can access an arbitrary server (external server) to produce and record data, and build a database by collecting the produced and recorded data. The database constructed here is analyzed by the AI unit, the purchase time of the product is determined according to the analysis result, the purchase is attempted from the purchase point from the product holder, and the purchased product is delivered from the distribution management unit to the buyer who is a part of the user. will be shipped

This is to sequentially perform the first data collection step (S10), the second data collection step (S20), the database building step (S30), the AI analysis step (S40), and the delivery step (S50) as shown in FIG. Let me explain in detail.

Here, in addition to the data transmitted from the database, the AI unit can collect data from arbitrary servers A, B, and C servers. Here, data collected from the external server may be first data, and data collected from server A, server B, and server C may be second data.

Furthermore, the purchase time determined by the AI unit may be determined by comprehensively analyzing the first data and the second data. Specifically, the first data may be data by a user. Here, the user may be a purchaser or a user accessing an external server including a plurality of servers. That is, the user can collect the information recorded and produced in the external server. That is, user information including buyer-related big data among users can be collected in the external server. The buyer refers to a potential purchase target, a user who has already made a purchase, and may be defined with the same meaning as a consumer in the contents described through an embodiment of the present invention. In addition, the big data related to the buyer means data including information collected from the buyer at least below.

Here, the external server including a plurality of servers may include at least one platform that sells a product to be delivered to the purchaser to the purchaser.

More specifically, information recorded and generated from a user includes buzz data. Buzz data refers to the contents mentioned by a user who has accessed the external server in relation to a specific topic, which is a record (including text, voice, and video) left by means of a recordable server such as a post or comment. means, and AI analysis can be made on the amount of buzz.

Of course, it is not collected all the buzz volume included in the external server, but through a filtering process, the records having a low relevance or irrelevance to the specific topic are excluded and collected. This process can be performed by the intervention of the machine-learned AI unit. Here, the external server includes one or more platforms among SNS, online communities, blogs, video media, and crowdfunding sites in addition to the above-mentioned platforms. In addition, the specific subject may be meat and processed meat products.

On the other hand, server A, server B, and server C may be the Livestock Distribution Comprehensive Information Center, Livestock Products Quality Evaluation Institute, and Agriculture, Forestry and Fisheries Food Education and Culture Information Center, respectively. It is possible to receive partial meat price information from the Livestock Distribution Information Center, public data API related to living body and carcass from the Livestock Product Quality Assessment Service, and public data API for wholesale Korean beef and calf from the Agriculture, Forestry, Fisheries Food Education and Culture Information Service. Through this, web crawling data can be collected into a database, and this is called second data.

That is, in the case of the first data, it means user information and information recorded or generated by the user in an external server, and the second data may be web crawled information through a public API. In the case of the first data, since it is collected and analyzed in real time, it is possible to immediately collect and analyze the preferences and demands of buyers, including changes in trends and product sales targets. This is data that forms public opinion through the vast amount of buzz data, but can be excluded from the standardized aggregation method.

In addition, it is possible to respond quickly to events that occur atypically. For example, when meat consumption decreases due to disease of livestock, it is an unpredictable fluctuation, so it is reliable because it is public data that relies on the web crawl cycle. Therefore, it is possible to respond immediately by collecting and analyzing buzz data in real time.

This may allow the consumer's demand forecasting rate to be further improved. Due to atypical issues, it is difficult to reflect the impact of issues occurring within a shorter period through a system that predicts and discloses changes in price or market conditions at weekly or monthly intervals, and this causes late responses. Therefore, it is possible to more accurately predict demand by providing the forecast value as an immediate response through the AI analysis stage.

In the case of data recorded as text among buzz data, the machine-learned AI unit learns positive and negative keywords in units of words, analyzes the text, and evaluation of positive or negative in units of sentences or paragraphs can be made. For example, it is possible to quantify and evaluate the reviews by deducting points as much as the number of negative words are inserted in the text, which is a review (product review) left on a meat sales platform, and adding points as much as the number of inserting positive words. Therefore, the evaluation of the product purchased by the buyer can be applied and reflected in the process of determining subsequent regular delivery.

In addition, it is possible to collect static data including a profile of a buyer accessing a meat sales platform or crowdfunding site included in an external server, and dynamic data including the purchaser's purchase history, consumption content, access frequency and product reviews. Through this, purchase history and commonalities between users can be extracted, and the purchase conversion rate, click-through rate, content consumption time, etc. used to measure the effect of advertisement exposure (the degree of sales through advertisements such as banners) can be utilized.

In addition, in the digital advertising market, changes in user interest, usage time, or behavioral patterns for devices are similar to the changes in the part of the purchase of Korean beef (meat) of the users who accessed the external server, and changes in the behavioral patterns according to the buyer's device. can Specifically, by analyzing major keywords from news articles such as food purchases, Korean beef, and economic conditions that are exposed to users and accessed by users from external servers, it is possible to suggest subscriptions for customized regular delivery according to their level of interest. This may advertise products that are judged to have high satisfaction from information exposed to users through natural language processing algorithms.

Meanwhile, the AI unit may extract a feature point by applying a random forest technique to information or a database collected from a user, and apply a principal component analysis (PCA) technique to the feature point to filter it. That is, as one method of machine learning, it is possible to output a class or average predicted value from a plurality of decision trees constructed in the training process. At this time, the average predicted value is referred to as the feature point, and only the principal component can be extracted by applying the PCA technique to the surrounding values including the feature point. This may be a technique for forming a predetermined demand standard. Here, the predetermined demand criterion may include a preference for a product, a product type, and a regular delivery cycle of the product.

As described above, the regular delivery system implementing the regular delivery method collects information of a plurality of buyers and users using the plurality of external servers who purchase products from a plurality of external servers including at least one product sales platform, and , and a database unit including a database generated by accumulating the collected information of the buyer and the information of the user.

And, it includes an AI unit for performing AI analysis based on the database included in the database unit. And, it may include a distribution management unit that receives the delivery information from the AI unit to deliver to the buyer.

Here, the AI unit may analyze the information of the purchaser and the user, calculate information that meets a predetermined demand standard, and deliver the product purchased from the product holder to the distribution management unit.

Of course, the AI department can predict the price trend of products that can be purchased from server A, server B, and server C through product holding locations. Of course, as described above, this may not depend on the public gator API, but may be the result of analysis together with the buzz data of the external server. Through this, the purchase time and price are determined based on the product owner and the demand of the buyer, so that the product can be purchased and moved to the distribution management department.

Here, the database unit may collect the first data and geographic information, and may provide a delivery area to the AI unit by matching the first data and geographic information, and the AI unit may deliver delivery information for each delivery area to the distribution management unit. .

Although representative embodiments of the present invention have been described in detail above, those of ordinary skill in the art to which the present invention pertains will understand that various modifications are possible within the limits without departing from the scope of the present invention with respect to the above-described embodiments. . Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims described below as well as the claims and equivalents.

S10: first data collection step
S20: second data collection step
S30: Database building stage
S40: AI analysis stage
S50: Shipping Phase

Claims (14)

a collecting step of collecting information of a plurality of buyers who purchase products from a plurality of external servers including at least one product sales platform and users using the plurality of external servers by a collecting unit;
a database construction step including a database generated by accumulating the information of the purchaser and the information of the user collected from the external server; and
An analysis step of performing AI analysis based on the database by the AI unit,
The analysis step is
A regular delivery method that analyzes the information of the purchaser and the user, calculates information that meets a predetermined demand standard, and delivers the product regularly to correspond to the displayed information.
The method according to claim 1,
The user information is
A regular delivery method, including buzz data in which exposure is increased above a predetermined reference value in the external server, and big data related to the purchaser in the external server.
The method according to claim 1,
The buyer's information,
static data including the buyer's profile; and
The purchaser's purchase history in the product sales platform, consumption content, and dynamic data including the frequency of access and product reviews; including; regular delivery method.
4. The method according to claim 3,
The product review is
collected in the database unit,
A regular delivery method that is divided into keywords meaning positive or negative by the machine-learned AI unit and quantified numerically by giving additional points or deductions to each keyword.
The method according to claim 1,
The AI unit,
A regular delivery method for extracting key points by applying a random forest technique to the information collected from the user or the database, and filtering by applying a principal component analysis (PCA) technique to the key points.
The method according to claim 1,
The database unit further comprises geographic information, a regular delivery method.
The method according to claim 1,
The predetermined demand standard is,
A regular delivery method, including a preference for the product, a type of the product, and a regular delivery cycle of the product.
Collecting information of a plurality of buyers who purchase products from a plurality of external servers including at least one product sales platform and users using the plurality of external servers, and accumulating the information of the buyers and the users a database unit including the generated database;
AI unit for performing AI analysis based on the database included in the database unit; and
A distribution management unit that receives the delivery information from the AI unit and delivers it to the buyer;
The AI unit,
A regular delivery system that analyzes the information of the buyer and the user, calculates information that meets a predetermined demand standard, and delivers the product purchased from a product holder to the distribution management unit.
10. The method of claim 9,
The user information is
A regular delivery system that reflects consumer demand forecasting, including buzz data in which the exposure amount is increased above the predetermined reference value from the external server.
10. The method of claim 9,
The buyer's information,
static data including the buyer's profile; and
The purchaser's purchase history in the product sales platform, consumption content, and dynamic data including the frequency of access and product reviews; including; regular delivery system.
11. The method of claim 10,
The product review is
collected in the database unit,
A regular delivery system that is divided into keywords meaning positive or negative by the machine-learned AI unit and quantified numerically by giving additional points or deductions to each keyword.
9. The method of claim 8,
The AI unit,
A regular delivery system that extracts key points by applying a random forest technique to the information collected from the user or the database, and filters by applying a principal component analysis (PCA) technique to the key points.
9. The method of claim 8,
The database unit further comprises geographic information, a regular delivery system.
9. The method of claim 8,
The predetermined demand standard is,
A regular delivery system, including a preference for the product, a type of the product, and a regular delivery period of the product.

Images