KR20190066966A - System and method for recommending goods - Google Patents

Abstract

A server for recommending goods related to one embodiment of the present invention comprises: a communication unit; a review database storing review information on a plurality of goods; and a control unit electrically connected to the communication unit and review database. One piece of review information corresponding to one goods includes at least one piece of review content and at least one piece of rating information. The control unit recommends goods within a critical rank when listing the goods in order of ranking according to evaluation values calculated using the freshness, frequency, and rating of review information corresponding to each of the goods, wherein the goods within the critical rank includes a keyword received through the communication unit among the goods as a tag.

Description

{SYSTEM AND METHOD FOR RECOMMENDING GOODS}

The present disclosure relates to a product recommendation server and a product recommendation method.

Recently, various products and services are being sold by electronic commerce. In this electronic commerce method, consumers acquire information about goods and services to be purchased with reference to the evaluation of existing buyers about goods and services.

However, as the type of product and the information to be referred to each product increase, consumers may have difficulty in selecting a product that meets their needs.

In order to solve these problems, a product recommendation system has been introduced to provide consumers with a list of recommended products, thereby helping consumers to easily find products with a high possibility of purchasing.

The present disclosure aims to solve the above-mentioned problems and other problems. Another object of the present invention is to provide a commodity recommendation server and a commodity recommendation method that can recommend a commodity that meets a user's request.

To achieve the above and other objects, a product recommendation server according to an embodiment includes a communication unit, a review database for storing review information on a plurality of products, and a control unit electrically connected to the communication unit and the review database, Wherein one review information corresponding to a product includes at least one review content and at least one rating information and wherein the control unit is a product including a keyword received through a communication unit of the plurality of products as a tag, And recommends a product within a critical rank when listing a plurality of products in order of ranking according to the evaluation value calculated using the latestness, frequency, and rating of review information.

The controller extracts a word having a frequency of at least a threshold value from at least one review content of one review information as a tag, calculates a weight for the tag according to the degree of use of the tag in the review information for a plurality of products, The tag value corresponding to the tag can be calculated by multiplying the frequency value by the weight.

The control unit may be configured to determine at least one of the following: an equal number of at least one review content of one review information, a time at which each of the at least one review content of one review information is created, Using one piece of rating information, the RFS value of one good can be calculated.

The control unit searches for a product having a tag corresponding to a keyword among a plurality of products, calculates an evaluation value using an RFS value of the retrieved product and a tag value of the tag corresponding to the retrieved product keyword, , And recommends products corresponding to the upper ranking.

,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 빈도,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 개수,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 모든 단어들의 개수일 수 있다.The control unit calculates the frequency of each of the words in at least one review content of one review information using the following equation,

,

The frequency of a particular word in at least one review content of one review information,

The number of specific words in at least one review content of one review information,

May be the number of all words in at least one review content of one review information.

는 태그에 대한 가중치,

는 복수의 상품에 대한 리뷰 정보들의 전체 개수,

는 복수의 상품에 대한 리뷰 정보들 중 특정 단어를 포함하는 리뷰 정보의 개수일 수 있다.The control unit calculates a weight for the tag using the following equation, ,

Is the weight for the tag,

The total number of review information on a plurality of products,

May be the number of review information including a specific word of review information for a plurality of products.

, WF는 복수의 상품에 대한 리뷰 정보들 각각의 컨텐츠의 개수의 분포 중 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠의 개수의 등위에 따라 결정된 가중치, WR은 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠 각각이 작성된 시점에 따라 결정된 가중치, 및 S는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠 각각에 대응하는 평점일 수 있다.The control unit calculates an RFS value of one product using the following equation,

, WF is a weight determined according to the degree of the number of at least one review contents of one review information among distributions of the contents of respective contents of review information for a plurality of goods, WR is a weight determined based on at least one review contents A weight determined according to the time at which each is created, and S may be a rating corresponding to each of at least one review content of one review information.

The control unit may calculate the evaluation value by multiplying the square root of the RFS value of the searched product by the tag value of the tag corresponding to the keyword of the searched product.

A method of recommending a product according to one embodiment includes reading review information for a plurality of products from a review database, wherein one review information corresponding to one product includes at least one review content and at least one rating information, , Receiving a keyword through a communication unit, searching for a product including a keyword as a tag among a plurality of products, calculating an evaluation value calculated using the latestness, frequency, and rating of review information corresponding to each searched product And determining a product within a critical rank as a recommended product when listing the products searched in the ranking order according to the ranking.

The step of determining the item as a recommended product may include extracting, as a tag, a word whose frequency is equal to or higher than a threshold value in at least one review content of one review information, Calculating a weight value, and calculating a tag value corresponding to the tag by multiplying a frequency value of the tag by a weight value.

The step of determining the product as a recommendation product includes the steps of: determining an order of the number of at least one review content of one review information among the distribution of the number of contents of the review information of the plurality of products, at least one review content of one review information And using the at least one rating information to calculate an RFS value of one product.

Calculating an evaluation value using an RFS value of the searched product and a tag value of a tag corresponding to the searched product keyword, determining a ranking of the plurality of products according to the evaluation value, recommending products corresponding to the upper ranking Step < / RTI >

,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 빈도,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 개수,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 모든 단어들의 개수일 수 있다.The step of extracting, as a tag, a word whose frequency is equal to or higher than a threshold value in at least one review content of one review information is calculated by calculating the frequency of each word in at least one review content of one review information using the following equation , ≪ / RTI >

,

The frequency of a particular word in at least one review content of one review information,

The number of specific words in at least one review content of one review information,

May be the number of all words in at least one review content of one review information.

,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 빈도,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 특정 단어의 개수,

는 하나의 리뷰 정보의 적어도 하나의 리뷰 컨텐츠에서의 모든 단어들의 개수일 수 있다.The step of calculating the weight for the tag according to the degree of use of the tag in the review information on the plurality of products may include calculating the frequency of each of the words in at least one review content of one review information Gt;

,

The frequency of a particular word in at least one review content of one review information,

The number of specific words in at least one review content of one review information,

May be the number of all words in at least one review content of one review information.

Effects of the product recommendation server and the product recommendation method according to the present disclosure will be described as follows.

According to at least one of the embodiments of the present disclosure, there is an advantage in that it is possible to recommend a product meeting the user's request.

According to at least one of the embodiments of the present disclosure, there is an advantage in that the sales amount of the goods can be increased.

1 is a block diagram for explaining a product recommendation system according to an embodiment.
2 is a block diagram illustrating a product recommendation server according to an embodiment.
3 is a flowchart showing a product recommendation method according to an embodiment.
4 is a graph illustrating performance evaluation of a product recommendation server and a product recommendation method according to an embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification and claims, when a section is referred to as "including " an element, it is understood that it does not exclude other elements, but may include other elements, unless specifically stated otherwise.

Hereinafter, the product recommendation server and the product recommendation method according to the embodiment will be described in detail with reference to necessary drawings.

In the present disclosure, a case where a product is recommended by a product recommendation system will be described as an example. However, the present invention is not limited thereto, and thus the product recommendation server and the product recommendation method according to the embodiments can be applied to the case of recommending intangible contents and services instead of products.

1 is a block diagram for explaining a product recommendation system according to an embodiment. As shown, the product recommendation system includes a server 100 and user terminals 200 that are capable of communicating with each other by a communication network 300.

The server 100 may provide a recommendation service for determining rankings of products according to keywords input from each of the user terminals 200 and providing information on the ranking to the user terminals 200. [

The user terminals 200 can access the server 100 and receive a recommendation service for the goods. The user terminal 200 may be a variety of commonly known communication terminals such as a smart phone, a PC, and a notebook computer.

The user may be connected to the server 100 through the communication network 300 with the user terminal 200 of the user, wherein the communication network 300 is a general Internet line and includes a wired communication network 300 and a wireless communication network 300 ). ≪ / RTI > Further, when the product recommendation method according to one embodiment is provided in a limited range of a specific company or the like, the network may be a limited area or a virtual private network or an intra network accessible only to an authorized person.

In addition, the server 100 may provide a recommendation service to the user terminals 200 in cooperation with a general search service server. For example, the user may connect to a search service server he normally uses to input a keyword, Can be requested. In this case, the search service server requests the commodity recommendation according to the keyword of the user to the server 100, and the server 100 may provide the commodity recommendation result to the search service server.

Next, with reference to Fig. 2, the product recommendation server 100 according to the embodiment will be described.

2 is a block diagram illustrating a product recommendation server 100 according to an embodiment.

The product recommendation server 100 may include a review database 110, a controller 120, a review analysis database 130, and a communication unit 140.

The review database 100 may store information about goods. The information on the products may include review information of each product. Here, one review information for one product may include at least one of content and rating inputted by the buyers for the product.

The control unit 120 may include a tag extracting unit 122, a review analyzing unit 124, and a ranking determining unit 126.

The tag extracting unit 122 can read the review information on the products to be recommended from the review database 110. [

The tag extracting unit 122 extracts tags from each review information read from the review database 110. For example, the tag extracting unit 122 can extract a word frequently appearing in the contents of all reviews of a product as a tag having explanatory power for the corresponding product.

Specifically, when the contents of all review information corresponding to a specific product are referred to as one document (), the word frequency value TF for a specific word () in the document () is determined by the following equation (1).

This means that the frequency of a specific word extracted from the document is divided by the sum of frequencies of all the words extracted from the document.

The tag extracting unit 122 extracts words based on general nouns and proper nouns from the review information, and extracts words having the upper cumulative sum of the word frequency values tf equal to or higher than the threshold value among the extracted words, It can be extracted as an element.

Further, the tag extracting unit 122 may calculate a weight for each tag. For example, since words frequently appearing in all products are not suitable for describing a particular product, weights can be calculated for the extracted tags to exclude these words.

The weight (IDF) can be calculated using the following equation (2).

Where | {| } Is the number of documents containing a particular word (), and || is the total number of documents.

The tag extracting unit 122 may calculate the tag value TagValue by multiplying the word frequency value TF of each tag by the weight value IDF as shown in the following Equation 3. [

Here is an element of the tag (). The tag value (TagValue) is a value dependent on a combination of a tag and a product. That is, the tag value (TagValue) may be different depending on the product even if the same tag is used. The higher the tag value (TagValue), the better the description of the product.

Next, the review analyzing unit 124 can also read review information on the products to be recommended from the review database 110. [ The review analyzer 124 can analyze the review information by dividing the review information into three indicators (RFS: Recency, Frequency, Score).

(1) Recency

The review analyzing unit 124 can assign a higher weight to the review information that has been recently generated. For example, the review analyzer 124 may assign weighted recency (WR) to each review information as shown in Table 1 below.

(2) Frequency

The review analyzer 124 may assign a higher weight to the product considering the distribution of the number of reviews of all the products and the number of reviews of the product. For example, when the review analyzer 124 sorts the products in the order of the number of reviews of each product, the review analyzer 124 calculates a weight (for each product) of each product according to the degree to which the number of reviews of the product belongs WF: Weighted Frequency).

For example, the review analyzer 124 may determine that the number of reviews of product 1 is 3, the number of reviews of product 2 is 4, the number of reviews of product 3 is 10, the number of reviews of product 4 is 8 , The weight WF of the commodity 1 may be 1.0, the weight WF of the commodity 2 may be 1.1, the weight WF of the commodity 4 may be 1.2, and the weight WF of the commodity 3 may be 1.3.

(3) Score

The review analyzer 124 can read the rating included in the review information.

Then, the review analyzer 124 may calculate an RFS value for one product having n reviews, as shown in Equation (4).

The RFS value has a unique value for each product, and the higher the value, the higher the popularity of the product.

The review analysis database 130 may store the tag information and the RFS value corresponding to each product. The tag information includes at least one tag corresponding to one product and a tag value of the tag.

Next, the ranking determining unit 126 may search for a product including the keyword inputted from the review analysis database 130 as a tag, and determine the ranking of the searched products. The ranking determining unit 126 may output information on products within a predetermined rank (for example, within the fifth rank) to the communication unit 140 according to ranking of evaluation values of the products.

The ranking determining unit 126 may calculate an evaluation value TRFS for one product using the following equation (5).

Here, TagValue is a tag value calculated for the tag of one commodity including the input keyword as a tag, and RFS is the RFS value of the commodity.

When a plurality of keywords are input, the ranking determining unit 126 may calculate a total evaluation value (Total TRFS) for reflecting all the inputted keywords using Equation (6) below.

Here, keyword is a set of words entered by the user.

The ranking determining unit 126 may output information on the commodities of the critical rank or higher to the communication unit 140 according to the ranking according to the total evaluation value (Total TRFS).

For example, when the product recommendation server 100 receives a plurality of keywords, the total evaluation value may be calculated and information on the top five products having the largest value may be output to the communication unit 140.

The communication unit 140 can communicate between the product recommendation server 100 and the communication network 300, between the product recommendation server 100 and another search server, or between the product recommendation server 100 and the user terminal 200 Lt; RTI ID = 0.0 > and / or < / RTI >

Next, referring to Fig. 3, a description will be given of a product recommendation method.

3 is a flowchart showing a product recommendation method according to an embodiment.

First, the tag extracting unit 122 extracts tags from each review information read from the review database 110 (S100). The tag extracting unit 122 may calculate a weight for each tag. The tag extracting unit 122 may calculate the tag value (TagValue) by multiplying the word frequency value TF of each tag by the weight value IDF using Equation (3) above.

Next, the review analyzing unit 124 divides the review information into three indicators (RFS) (S110). The review analyzer 124 may calculate RFS values for each product using Equation (4).

The review analysis database 130 can store the tag information extracted by the tag extraction unit 122 and the RFS value calculated by the review analysis unit 124 in association with each product.

Next, the ranking determining unit 126 searches for a product including the keyword input from the review analysis database 130 as a tag, and determines the ranking of the searched products (S140).

The ranking determining unit 126 provides information about the products having a ranking of the critical rank or higher as a recommended product list according to the ranking of the evaluation values of each of the products (S150).

According to the embodiment described above, the product recommendation server 100 and the product recommendation method extract tags of goods, calculate tag values and RFS values, and provide information on products matching the inputted keywords, The advantage is that you can recommend products that meet your needs.

The performance of the product recommendation server 100 and the product recommendation method can be evaluated by the following equation (7).

Here, [] may be a TagValue of the third product.

Equation (7) indicates a DCG (Discounted Cumulative Gain) scale, and the lower the value, the smaller the value becomes by the log. Therefore, the DCG value increases as the highly correlated product appears in the higher rank. In other words, the higher the relevance of the third product to the search results, the better the ranking should be.

In addition, the performance of the product recommendation server 100 and the product recommendation method can be evaluated by the variance of the TRFS value. The smaller the TRFS variance value, the more stable the ranking by the specific tag can be prevented. When a plurality of keywords are input from a user, the TRFS variance should be small in order to reflect the influence of the keywords evenly.

For performance evaluation, the DCG value and the TRFS variance value at the time of calculating the TRFS value by other mathematical expressions (the following equations (8) to (10)) besides the equation (5) are evaluated.

For performance evaluation, 18,525 reviews of 658 business / economic books were used. 6,028 tags were extracted from 658 books and 200 tags were sampled for the experiment. For each tag, the top five books with the TRFS value according to the functions according to Equations (5) and (8) to (10) were configured as recommendation items. The average of the DCG of the recommendation result obtained by applying 200 tags to the four functions is as shown in Fig. 4 (a), and the average of the TRFS dispersion is as shown in Fig. 4 (b).

4 is a graph illustrating performance evaluation of a product recommendation server and a product recommendation method according to an embodiment.

As shown in FIG. 4 (a), it can be seen that the DCG of the function according to Equation (5) is the largest at 10.845. Also, as shown in FIG. 4B, the TRFS dispersion of the function according to Equation (5) is the smallest at 46.84. Therefore, when using the function according to Equation (5), it can be seen that the performance of the product recommendation method is better than other functions.

The embodiment of the present invention is not limited to the above-described apparatus and / or method, but may be applied to a program recorded on a recording medium or a recording medium on which the program is recorded to realize a function corresponding to the configuration of the embodiment of the present invention And the present invention can be easily implemented by those skilled in the art from the description of the embodiments described above.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, It belongs to the scope of right.

Claims (14)

Communication unit,
A review database that stores review information for a plurality of products, and
A control unit electrically connected to the communication unit and the review database,
Including the
One review information corresponding to one product includes at least one review content and at least one rating information,
Wherein the control unit is a product that includes a keyword received through the communication unit among the plurality of products as tags, and the control unit is configured to determine, based on the evaluation value calculated using the latestness and frequency of review information corresponding to each of the plurality of products, Recommending a commodity within a critical rank when listing the plurality of commodities in order of ranking,
Product recommendation server. The method according to claim 1,
Wherein the control unit extracts a word having a frequency equal to or higher than a threshold value from at least one review content of the review information as a tag and extracts a weight for the tag based on the degree of use of the tag in the review information for the plurality of products And a tag value corresponding to the tag is calculated by multiplying the frequency value of the tag by the weight,
Product recommendation server. 3. The method of claim 2,
Wherein the control unit is further configured to determine whether the number of reviews of at least one review content of the one review information among the distribution of the number of contents of the review information of the plurality of products, And using the at least one rating information to calculate an RFS value of the one item,
Product recommendation server. The method of claim 3,
The control unit searches for a product having a tag corresponding to the keyword among the plurality of products and calculates an evaluation value using an RFS value of the searched product and a tag value of a tag corresponding to the keyword of the searched product Determining a ranking of the plurality of products in accordance with the evaluation value, recommending products corresponding to an upper ranking,
Product recommendation server. 3. The method of claim 2,
Wherein the controller calculates the frequency of each of the words in at least one review content of the one review information using the following equation,

,

A frequency of a specific word in at least one review content of the one review information,

The number of the specific words in at least one review content of the one review information,

Which is the number of all words in at least one review content of the one review information,
Product recommendation server. 6. The method of claim 5,
The control unit calculates a weight for the tag using the following equation,

,

A weight for the tag,

The total number of review information on the plurality of products,

Which is the number of review information including the specific word among the review information on the plurality of goods,
Product recommendation server. The method of claim 3,
Wherein the controller calculates an RFS value of the one item using the following equation,

,
Wherein the WF is a weight determined according to an equal number of at least one review content of the one review information among distributions of contents of respective review information on the plurality of products, A weight determined according to a time when each of the review contents is created, and S is a rating corresponding to each of at least one review content of the one review information,
Product recommendation server. 5. The method of claim 4,
Wherein the control unit calculates the evaluation value by multiplying the square root of the RFS value of the searched product by the tag value of the tag corresponding to the keyword of the searched product,
Product recommendation server. Reading review information for a plurality of products from a review database, wherein one review information corresponding to one product includes at least one review content and at least one rating information,
Receiving a keyword through a communication unit,
Searching for a commodity including the keyword as the tag among the plurality of commodities and searching for the retrieved commodity in the order of ranking according to the evaluation value calculated using the latestness, frequency, and rating of the review information corresponding to each of the retrieved commodities, A step of determining a product within a critical rank as a recommended product when listing the items
The product recommendation method. 10. The method of claim 9,
Wherein the step of determining as the recommended product comprises:
Extracting, as a tag, a word whose frequency is equal to or higher than a threshold value in at least one review content of the review information;
Calculating a weight for the tag according to the degree of use of the tag in review information on the plurality of products, and
And multiplying the frequency value of the tag by the weight to calculate a tag value corresponding to the tag.
How to recommend products. 11. The method of claim 10,
Wherein the step of determining as the recommended product comprises:
An equal number of at least one review content of the one review information among distributions of the contents of the respective review information about the plurality of products, a time at which each of the review contents of the one review information is created, Further comprising using the at least one rating information to calculate an RFS value for the one item.
How to recommend products. 12. The method of claim 11,
Calculating an evaluation value using an RFS value of the searched product and a tag value of a tag corresponding to the keyword of the searched product; and
Determining a ranking of the plurality of products in accordance with the evaluation value, and recommending products corresponding to the upper ranking,
How to recommend products. 12. The method of claim 11,
The step of extracting, as a tag, a word whose frequency is equal to or higher than a threshold value in at least one review content of the review information,
Calculating the frequency of each of the words in at least one review content of the one review information using the following equation,

,

A frequency of a specific word in at least one review content of the one review information,

The number of the specific words in at least one review content of the one review information,

Which is the number of all words in at least one review content of the one review information,
How to recommend products. 12. The method of claim 11,
The step of calculating a weight for the tag according to the degree of use of the tag in the review information for the plurality of products,
Calculating the frequency of each of the words in at least one review content of the one review information using the following equation,

,

A frequency of a specific word in at least one review content of the one review information,

The number of the specific words in at least one review content of the one review information,

Which is the number of all words in at least one review content of the one review information,
How to recommend products.

Images