JP6816899B2 - Optimized transaction recommendation system by analyzing purchase or reservation emails - Google Patents

Description

The present invention relates to an optimized transaction recommendation system, and more particularly to a transaction recommendation system optimized by analyzing purchase emails or reservation emails of goods and services.

In recent years, the number of online stores and online services has been increasing rapidly, and the share of online purchases and reservations of goods and services is increasing. According to Ministry of Economy, Trade and Industry statistics, it is estimated that by 2020, 20% of all household consumption will come from online shopping.

An e-commerce operator such as an online store presents similar products and services to users and purchases them by analyzing the purchase history of products and services sold at the online stores operated by the company. We make recommendations such as supporting. By making recommendations, users can easily find products and services that meet their tastes and needs from a large number of products and services.

However, users such as online stores often use not only one online store but also a plurality of online stores and online services. In such a situation, it is not enough to analyze only the purchase history of online stores operated by one operator in order to fully understand the purchase tendency of users and make more appropriate recommendations. There is a need for a transaction recommendation system that can analyze purchase history across the operators of the company.

As a system for making recommendations in an online store, Patent Document 1 is an information providing server that provides a user with recommendation information that the user may be interested in, and is a second with respect to information related to the first history information for the first site. The history information of the first site and the third history information for the second site are acquired, and the evaluation value for the related information is calculated based on the history information of the first site and the second site. It discloses that related information is transmitted to terminals based on the history of terminals that span different sites such as sites.

Further, Patent Document 2 is a system in which a user associates a product from a plurality of store sites that sell the product based on the user's behavior on the store site and recommends the related product to the user. When moving to a store site, identify the product sold at the source store site that caused the move to another store site, and use the specified product as the recommended product to the recommended product at the destination store site. It is disclosed that the link of is presented.

Further, Patent Document 3 is an information providing method for providing recommended information about a product recommended to a user who has accessed the EC site based on the purchase history information of the product purchased by a plurality of users using the EC site. The post-purchase product group Ys, which is a set of post-purchase products Y purchased after the purchase of product X by another user who has a history of purchasing X, is extracted from the purchase history information, and the product X is selected from the post-purchase product group. It discloses that the post-purchase product Y, which has the most correlation with the above, is selected as the recommended product.

Japanese Unexamined Patent Publication No. 2013-250753 Japanese Unexamined Patent Publication No. 2014-48998 JP-A-2018-169987

Patent Document 1 makes it possible to give the user relevant information that becomes noticeable for taking further action based on the user's action history between sites of different genres such as a music store site and a video store site. It is something to do.

Further, in Patent Document 2, when a user moves between a plurality of store sites, the user himself / herself recommends products of different store sites by presenting a link to a recommended product at the destination store site. This is to prevent the store from leaving the store site and improve the incentive to participate in the recommendation system in which each store site recommends recommended products to each other.

Further, Patent Document 3 meets the needs of the user by selecting the post-purchase product Y, which has the greatest correlation with the product X among the products Y purchased after the other user has purchased the product X, as the recommended product. It is intended to eliminate the provision of information on products that do not, and to provide the user with more qualified recommendation information including products that meet the potential needs of the user.

However, although Patent Documents 1 to 3 attempt to present recommendations based on behavior history between different genres, different store sites, and different users, the analysis is limited to one operator. However, it does not disclose a specific solution on the assumption that the recommendations will be optimized by broadly cross-sectionally analyzing the different operators to which a large number of store sites belong.

Therefore, the present invention provides a transaction recommendation system that solves the above problems and optimizes recommendations by broadly cross-sectionally analyzing different operators to which a large number of store sites belong.

In order to solve the above problems, in the present invention, transaction data acquired from a plurality of external services is analyzed, recommendations related to transactions are generated and displayed on the user terminal, and recommendations are generated in consideration of feedback from the user terminal. Provides a trading recommendation system. In the present invention, in particular, as transaction data, messages such as e-mails exchanged between users and a plurality of operators operating a plurality of on-sign stores and online services are analyzed, and recommendations regarding transactions are generated. It is characterized by that.

The transaction recommendation system according to the present invention is stored in a transaction data acquisition unit that acquires transaction data from a plurality of external services, a transaction data storage unit that stores transaction data acquired by the transaction data acquisition unit, and a transaction data storage unit. It is characterized by including a recommendation generation unit that analyzes transaction data, generates recommendations and displays them on the user terminal, and a feedback processing unit that receives feedback from the user terminal.

In the transaction recommendation system according to the present invention, the transaction data acquisition unit filters messages exchanged in an external service, and at least acquires a message transmitted at the time of purchase or reservation of a product or service as transaction data. To do.

In the transaction recommendation system according to the present invention, the transaction data acquisition unit is further characterized in that a message transmitted when canceling a purchase or reservation of a product or service or withdrawing from a service is acquired as transaction data.

In the transaction recommendation system according to the present invention, the message is an e-mail message, a message stored in the cloud, or a message exchanged by SNS.

In the transaction recommendation system according to the present invention, the recommendation generation unit analyzes transaction data and generates recommendations regarding the purchase or reservation of goods or services.

In the transaction recommendation system according to the present invention, the feedback processing unit is characterized in that a questionnaire is automatically transmitted to a user terminal at the time of purchasing a product or service or after a predetermined period of time has elapsed from the time of purchase.

The transaction recommendation system according to the present invention is characterized in that the feedback processing unit acquires a response to a questionnaire from a user terminal and transmits it to a recommendation generation unit.

In the transaction recommendation system according to the present invention, the recommendation generation unit is characterized in that the response to the questionnaire obtained from the feedback processing unit is added to the transaction data, analyzed, and the recommendation is generated.

In the transaction recommendation system according to the present invention, the questionnaire is characterized by including a user's evaluation or comment regarding a product or service.

According to the present invention, there is provided a transaction recommendation system that analyzes transaction data acquired from a plurality of external services, generates recommendations related to transactions, displays them on a user terminal, and generates recommendations by adding feedback from the user terminals. By doing so, it is possible to provide the user with the optimum recommendation regarding the transaction, improve the efficiency of the recommendation, and improve the satisfaction of the user.

Further, according to the present invention in another aspect, it is more optimal by analyzing not only the transaction data related to the purchase or reservation but also the transaction data regarding the cancellation or withdrawal of the purchase or reservation and the transaction data including the feedback from the user. It is possible to provide optimized recommendations.
Other objects, features and advantages of the present invention will become apparent from the following description of the embodiments of the present invention with respect to the accompanying drawings.

FIG. 1 is a conceptual diagram showing the entire transaction recommendation system according to the present invention. FIG. 2 is a diagram showing processing of a transaction data acquisition unit of the transaction recommendation system according to the present invention. FIG. 3 is a diagram showing an example of data stored in the transaction data storage unit of the transaction recommendation system according to the present invention. FIG. 4 is a diagram showing processing of a recommendation generation unit of the transaction recommendation system according to the present invention. FIG. 5 is a diagram showing processing of the feedback processing unit of the transaction recommendation system according to the present invention.

FIG. 1 is a conceptual diagram showing the entire recommendation system according to the present invention.
The transaction recommendation system 1 according to the present invention includes a transaction data acquisition unit 10 that acquires transaction data from a plurality of external services 2, a transaction data storage unit 20 that stores transaction data acquired by the transaction data acquisition unit 10, and transaction data storage. It includes a recommendation generation unit 30 that analyzes transaction data stored in the unit 20 to generate recommendations and displays them on the user terminal 50, and a feedback processing unit 40 that receives feedback from the user terminal 50.

Here, the external service 2 refers to various services capable of sending and receiving messages, such as an e-mail service, a cloud service, an SNS service, and a calendar service. The external service 2 may include not only a plurality of external services already deployed in the market, but also a service capable of transmitting and receiving various messages to be added in the future.

These external services 2 are used between an online store, a user who uses the online store, an operator of an electronic commerce such as an online store, and the like. For example, the exchange of e-mails regarding the purchase and reservation of products and the exchange of messages such as SNS are performed via the external service 2.

The transaction data acquisition unit 10 filters messages such as e-mails exchanged in the external service 2, and acquires at least a message transmitted at the time of purchase or reservation of a product or service as transaction data. The e-mails and SNS messages exchanged in the external service 2 include those that are not related to transactions such as purchases and reservations of products and services, and the transaction data acquisition unit 10 filters the products and services from among them. Extract items related to transactions such as purchases and reservations.

The transaction data acquisition unit 10 sends transaction data as an e-mail or SNS message regarding the purchase or reservation of a product or service, as well as a message sent when canceling the purchase or reservation of a product or service or withdrawing from the service. Get as.

Here, the message is an e-mail message, a message stored in the cloud, or a message exchanged by SNS. For example, the e-mail messages exchanged in the plurality of e-mail services 3 are aggregated on the cloud service 4, and the aggregated e-mail messages are acquired by the transaction data acquisition unit 10. When aggregating e-mail messages and the like on the cloud service 4, the consent of the user is obtained in advance. The e-mail message may be acquired directly from the e-mail service 3 by the transaction data acquisition unit 10.

The transaction data storage unit 20 stores the transaction data acquired by the transaction data acquisition unit 10. The acquired transaction data is stored in the form of a database or table. For example, the sender, receiver, transmission date and time, reception date and time, title, text, etc. of the e-mail are stored in the transaction data storage unit 20. The transaction data acquisition unit 10 and the transaction data storage unit 20 may be constructed on the cloud. Further, the transaction data acquisition unit 10 and the transaction data storage unit 20 may be combined to form one server.

The recommendation generation unit 30 analyzes the transaction data and generates recommendations regarding the purchase or reservation of goods or services. Here, the recommendation refers to what is proposed to the user regarding the purchase or reservation of goods or services.

Recommendations include, for example, recommending a product that a user is likely to purchase next or a service that is likely to be booked next, or recommending a product that a user should purchase next or a service that should be booked next. This includes predicting the inventory of consumables in the home or business office owned by the user, and recommending the product when it is predicted that the product is in short supply.

The feedback processing unit 40 automatically sends a questionnaire to the user terminal at the time of purchasing the product or service or after a predetermined period of time has elapsed from the time of purchase.

The feedback processing unit 40 acquires a response to the questionnaire from the user terminal and transmits it to the recommendation generation unit.

The recommendation generation unit 30 adds the response to the questionnaire obtained from the feedback processing unit 40 to the transaction data, analyzes it, and generates a recommendation.

Here, the questionnaire includes a user's evaluation or comment regarding the product or service.

The user terminal 50 is, for example, preferably a smartphone, but may be any terminal used by the user, such as a tablet, laptop or stand-alone computer.

FIG. 2 is a diagram showing processing of the transaction data acquisition unit 10 of the transaction recommendation system 1 according to the present invention.
The transaction data acquisition unit 10 acquires a message regarding a transaction of a product or service from the external service 2 as transaction data. The transaction data acquired by the transaction data acquisition unit 10 includes, for example, an online store product purchase email, a product or service reservation email, and the like. The transaction data includes, for example, information such as purchasing product P at online store A and making a reservation for service Q at online service B, and these information includes e-mails obtained from external service 2 and the like. Extracted from the content of the message.

Here, the online store purchase email is, for example, an email notifying the user who purchased the product that the purchase procedure has been completed when the purchase of the product is completed in the online store. The reservation email is, for example, an email notifying the user who has reserved the product or service that the reservation procedure has been completed when the reservation for the product or service is made at the online store. The purchase email or reservation email is, for example, a message such as an e-mail or SNS that is automatically delivered from the online store after purchasing the product at the online store.

The transaction data acquisition unit 10 filters messages such as e-mails exchanged in the external service 2. The transaction data acquisition unit 10 extracts only those related to transactions such as purchase and reservation of products from the filtering data. The transaction data acquisition unit 10 stores a message related to a transaction of a product or service extracted by filtering in the transaction data storage unit 20 as transaction data.

FIG. 3 is a diagram showing an example of data stored in the transaction data storage unit 20 of the transaction recommendation system 1 according to the present invention.
In the example of FIG. 3, the items are "ID", "purchase / reservation date", "product name / service name", "location", "start", "end", "withdrawal", "category", and "store". Have. A serial number of data is stored in the item "ID". In the item "purchase / reservation date and time", the purchase date and time of the product or the reservation date and time of the service is stored. In the item "product name / service name", the product name purchased by the user or the service name reserved by the user is stored. The item "location" stores a location related to the goods or services. For example, when booking an aircraft ticket, the airport name may be stored as the location. The start date and time of the service is stored in the item "start". The end date and time of the service is stored in the item "end". For example, in the case of booking an aircraft ticket, the departure date and time of the flight may be stored in the item "start" and the arrival date and time of the flight may be stored in the item "end". In the item "withdrawal", for example, a flag "1" may be stored when the service is withdrawn. The item "category" stores a category of goods or services. For example, categories such as "merchandise", "flight", and "video subscription" are stored. The name of the online store is stored in the item "store". Note that the item names and item types in FIG. 3 are examples, and are not limited thereto.

The data stored in the transaction data storage unit 20 is acquired from a plurality of external services 2 by the transaction data acquisition unit 10 and filtered, that is, for example, purchase mail, reservation mail, cancellation mail, withdrawal mail. Etc., SNS messages, or other cloud-aggregated messages.

FIG. 4 is a diagram showing the processing of the recommendation generation unit 30 of the transaction recommendation system 1 according to the present invention.
The recommendation generation unit 30 analyzes the transaction data stored in the transaction data storage unit 20. The recommendation generation unit 30 preferably sends the analysis result as a recommendation to the user terminal 50 and stores it in the transaction data storage unit 20. The analysis result is stored in the transaction data storage unit 20 as past transaction data. The past transaction data accumulated in the transaction data storage unit 20 can also be used as learning data for analysis in the recommendation generation unit 30.

The recommendation generation unit 30 extracts a product or service with a high probability that the user will purchase or make a reservation next by multiplying a plurality of conditions. In a mixture of various products and services, the motives and conditions for recommendations also differ. However, according to the recommendation generation unit 30 of the present invention, in a mixture of various products and services, products or services that are likely to be purchased or reserved by the user next are extracted by a common mechanism or platform. To do. The recommendation generation unit 30 statistically detects and analyzes what the user tends to purchase next, preferably by multiplying the online store, product name, product category, service name, service category, date and time, location, and the like. ..

Multiple conditions used for crossing include, for example, purchase date and time, reservation date and time, product name, service name, product category, service category, product production area and production area, service provision location, service start date and time, and service. It may be a plurality of arbitrary attributes related to the product or service, such as the end date and time and withdrawal from the service.

In order to collectively evaluate products or services that should be recommended in a mixture of products and services of various genres, the degree of recommendation is scored or recommended using a common evaluation value common to various products and services. Can be ranked. If the common evaluation value is high, the degree of recommendation is high, and the recommendation generation unit 30 presents it to the user terminal 50 as a recommendation target.

In addition, the recommendation generation unit 30 can recommend a product that the user is likely to purchase next or a service that is likely to be reserved next. The product or service that the user is likely to purchase next or the service that is likely to be booked next is, for example, for each of the products or services stored in the user's past purchase history, that product or service is next. It is extracted by calculating the probability of purchase or reservation.

The recommendation generation unit 30 extracts frequently purchased products or frequently used services from the user's past purchase history or reservation history stored in the transaction data storage unit 20. When using the common evaluation value, points are added to the common evaluation value of products that are frequently purchased or services that are frequently used. The common evaluation value of each product or service calculated by the recommendation generation unit 30 is stored in the transaction data storage unit 20.

The recommendation generation unit 30 makes the past purchase history or reservation history refer not only to the past purchase history or reservation history of the user but also to the past purchase history or reservation history of other users similar to the user. May be good. That is, a frequently purchased product or a frequently used service is extracted from the past purchase history or reservation history of another user similar to that user. When the common evaluation value is used, points are added to the common evaluation value of other users who are similar to the user and who purchase frequently or frequently use services.

In addition, the recommendation generation unit 30 can recommend the product to be purchased next or the service to be reserved next by the user. For example, if a user purchases a toy that does not come with batteries and the batteries run on four AAA batteries, we recommend a pack of four or more AAA batteries as the next product to purchase. To do. In another example, if the user has reserved an aircraft ticket but has not yet reserved an inn, it recommends an inn reservation according to the destination of the aircraft.

In this way, the degree of recommendation is high for products that are used simultaneously or continuously in time when using a certain product, or services that are used simultaneously or continuously in time when using a certain service. Become. When the common evaluation value is used, the common evaluation of products that are used simultaneously or continuously in time when using a certain product, or services that are used simultaneously or continuously in time when using a certain service. Add points to the value.

In addition, the recommendation generation unit 30 can make recommendations that reflect the user's tastes. For example, it is possible to analyze the purchase history or reservation history of the user's product, extract the genre of the product or service with high purchase frequency or reservation frequency, and reflect it in the recommendation. Further, for example, in the case of content such as music, moving images, books, etc., a person such as an artist, author, or author of the user's preference can be extracted and reflected in the recommendation. In addition, it is possible to send a questionnaire to the user, analyze the response result of the questionnaire, and recommend products and services that suit the user's taste.

In this way, the degree of recommendation is high for products in genres that are frequently purchased or services in genres that are frequently reserved. In addition, the degree of recommendation is high for products and services related to a person such as a user's favorite artist, author, or author. In addition, the degree of recommendation is high for products or services that are highly evaluated in the responses to questionnaires from users. When using the common evaluation value, points are added to the common evaluation value of products of the genre with high purchase frequency or services of the genre with high reservation frequency. In addition, points are added to the common evaluation value of products and services related to the user's favorite person. In addition, points are added to the common evaluation value of products or services that are highly evaluated in the responses to questionnaires from users.

In addition, the recommendation generation unit 30 can reflect the purchase of the product, the cancellation of the reservation of the service, or the withdrawal from the service in the generation of the recommendation. For example, a product or service that the user has canceled or withdrawn is considered to be less interested or necessary by the user, so the priority should be lowered when listing it as a recommendation.

That is, the degree of recommendation is low for products or services that the user has canceled or withdrawn. When the common evaluation value is used, points are deducted from the common evaluation value of the product or service that the user has canceled or withdrawn.

In addition, the recommendation generation unit 30 can predict the inventory of consumables in the home or business establishment owned by the user, and can recommend the product when it is predicted that the consumables are insufficient. That is, when a user repeatedly purchases a certain product (mainly consumables), the consumption pace of the product (mainly consumables) is predicted, and the current consumption pace is calculated from the previous purchase time of the product and the predicted consumption pace. Predict the inventory of the product in the home or business, and recommend the product when it is predicted that the product is insufficient. For example, if a user repeatedly purchases a dozen 2 liter PET bottles of water, the user's water consumption pace is predicted, and the current consumption pace is calculated from the previous purchase time of the water PET bottle and the predicted consumption pace. Predict the inventory of PET bottles of water, and recommend purchasing PET bottles of water when it is predicted that the inventory is insufficient.

In this way, the degree of recommendation is high for products that are predicted to be in short supply in the home or business. When using the common evaluation value, points are added to the common evaluation value of products that are predicted to be out of stock in the home or business.

In addition, the recommendation generation unit 30 predicts the usage pace of the service when the user repeatedly uses the service, and the desire or need of the user to use the service again from the previous usage time of the service. If it is predicted that the number of users will increase and the need for using the service will increase, we will recommend the service. For example, if a user repeatedly uses outpatient clinics, dental treatments, gyms, sports facilities, massage shops, beauty salons, etc. for health and beauty maintenance, a certain period of time has passed since the last use of the service. If so, anticipate that the need for use of the service is increasing and recommend the service.

In particular, the next usage date is predicted from the usage pace of the service, and it is judged that the need for using the service increases as the predicted usage date approaches without using the service next time. In addition, it is judged that the need for using the service increases as the predicted usage date passes and the period elapses without using the service next time. For example, if a user who uses a beauty salon once every two months does not make a reservation for a beauty salon for three months, he predicts that the need for using the beauty salon is increasing and makes a reservation for the beauty salon. I recommend it.

In this way, when the needs for using the service are high, the degree of recommendation for the service is high. When the common evaluation value is used, points are added to the common evaluation value of the service if the needs for use of the service are high.

In addition, the recommendation generation unit 30 can recommend the reservation of the service according to the itinerary. That is, when the user is in the middle of a trip or a business trip, the user's next action can be predicted and the reservation of the service required for that action can be recommended. For example, if the user makes a reservation for the Shinkansen and can determine from the departure time that the user is on the Shinkansen, when the arrival time of the Shinkansen approaches, a reservation for taxi dispatch to the arrival station To recommend. In this case, it is possible to make recommendations not only for taxi dispatch reservations, but also for possible transportation means such as reservations for limited express tickets on conventional lines, reservations for bus tickets, reservations for rental cars and rental bicycles.

In this way, a service that predicts the user's next behavior and is predicted to be a service necessary for that behavior has a high degree of recommendation. When the common evaluation value is used, the next action of the user is predicted, and points are added to the common evaluation value of the service predicted to be the service required for the action.

As described above, some examples of the generation of recommendations have been given, but the recommendations are not limited to these, and recommendations can be generated by combining the examples of the generation of any one or more recommendations as described above. When the common evaluation value is used, the common evaluation value is calculated by summing the points added or deducted in the method for generating one or more recommendations as described above.

The recommendation generation unit 30 displays the recommendation on the user terminal 50. For example, the recommendation generation unit 30 displays product names and images of a plurality of products that the user is likely to purchase next, product descriptions, and a plurality of products that the user is likely to make a reservation for next on the display screen of the user terminal 50. Display the service name, image, service description, etc. of the service in a list or thumbnail format. Goods or services displayed as recommendations may be labeled or marked according to their category or may be color-coded for display. In addition, the products or services displayed as recommendations may be sorted and displayed in an arbitrary order such as a recommended order or a category.

Further, the recommendation generation unit 30 may display the reason for the recommendation by characters, icons, color coding, or the like when displaying the recommendation on the user terminal 50. For example, products or services that have been recommended due to high purchase frequency or reservation frequency are displayed as "usual", and for products or services that are predicted to be out of stock at home or in the office. If the needs for use are high, a service or a service that predicts the next action of the user and is predicted to be a service necessary for that action may be displayed as "may be necessary". This makes it possible to easily recognize the reason for the recommendation in a mixture of various products and services. In addition, depending on the user's selection or the like, only those to which a specific reason for recommendation is given may be extracted and displayed, or may be sorted and displayed according to the reason.

FIG. 5 is a diagram showing processing of the feedback processing unit 40 of the transaction recommendation system 1 according to the present invention.
The feedback processing unit 40 sends a questionnaire to the user terminal 50 and obtains the answer to the questionnaire from the user terminal 50. The feedback processing unit 40 transmits the answer to the questionnaire acquired from the user terminal 50 to the recommendation generation unit 30.

The feedback processing unit 40 automatically sends a questionnaire to the user terminal 50. The questionnaire is conducted to obtain the user's evaluation of the product or service after purchasing the product or booking the service, and to analyze the user's interests and preferences. The feedback processing unit 40 transmits the answer to the questionnaire to the recommendation generation unit 30 so that the answer to the questionnaire can be added when the recommendation generation unit 30 generates a new recommendation. The recommendation generation unit 30 generates a new recommendation using the analysis result of the past questionnaire. The questionnaire is preferably sent to the user terminal 50 by transmitting an electronically displayable input form to the user terminal 50. In another example, the questionnaire is sent to the user terminal 50 by describing the contents of the questionnaire in the body of the message of the e-mail or the SNS service, and the reply from the user terminal 50 is also returned in the form of the message of the e-mail or the SNS service. It may be done.

The feedback processing unit 40 automatically sends a questionnaire to the user terminal 50 at the time of purchasing the product or service or after a predetermined period of time has elapsed from the time of purchase. The predetermined period is preferably a period in which it is predicted that the content of the product or service is confirmed and the product is actually used or the service is actually used, for example, one week or 10 days. In addition, the predetermined period may be automatically changed according to the content of the product or service.

The contents of the questionnaire include options that directly lead to the prediction of the next purchasing behavior, such as "I want to buy again", "I want to use it again", "I do not want to buy it again", "I do not want to use it again", etc. It may be asked to select an option indicating an evaluation of a product such as "good", "normal", or "bad". In addition, a free response column may be provided for entering an evaluation or comment on the product or service.

In addition, the content of the questionnaire may be automatically changed according to the content of the product or service. For example, in the case of a questionnaire for travel reservations, the option "I want to buy again" may be changed to the option "I want to go again". In that case, although the expressions "want to buy again" and "want to go again" are different, it can be said that the evaluations are substantially the same from the viewpoint of predicting the next purchasing behavior. Therefore, even if the expressions are different, the answer results may be encoded or scored so that the options that should be evaluated substantially the same have the same evaluation. This makes it possible to evaluate products and services of different genres with the same index.

The feedback processing unit 40 gives points to the user when the user receives the answer to the questionnaire in order to enhance the merit of the user answering the questionnaire.

The transaction recommendation system 1 according to the present invention can be used as a system provided to an online store operator, a research operator, an advertising agency, and a point system provider. In addition, a part of the transaction recommendation system according to the present invention can be used as an anonymized data linkage engine or recommendation engine.

The goods or services covered by the present invention are not limited to those exemplified above, and can be used for goods or services of any genre and form. In addition to product sales, delivery, flights, trains, buses, ships, rental cars, taxi dispatch, accommodation, restaurant reservations, delivery, beauty / esthetics, hospital / treatment reservations, concerts, live performances, watching sports, exhibitions It can be applied to the recommendation of all kinds of products or services such as event tickets for concerts and festivals.
Although the above description has been made for Examples, it will be apparent to those skilled in the art that the present invention is not limited thereto and various modifications and modifications can be made within the scope of the principles of the present invention and the appended claims.

1 Transaction recommendation system 2 External service 10 Transaction data acquisition unit 20 Transaction data storage unit 30 Recommendation generation unit 40 Feedback processing unit 50 User terminal

Claims (9)

It is a trading recommendation system,
A transaction data acquisition department that acquires transaction data from multiple external services,
A transaction data storage unit that stores transaction data acquired by the transaction data acquisition unit, and a transaction data storage unit.
A recommendation generation unit that analyzes transaction data stored in the transaction data storage unit, generates recommendations, and displays them on a user terminal.
A feedback processing unit that receives feedback from the user terminal,
With
The feedback processing unit automatically sends a questionnaire to the user terminal at the time of purchasing the product or service or after a predetermined period of time has elapsed from the time of purchase.
The feedback processing unit automatically changes the predetermined period according to the content of the product or service, and sends the questionnaire .
The recommendation generation unit scores the degree of recommendation using a common evaluation value common to different products or different services, and when using a certain product, the product or a service that is used simultaneously or continuously in time. The transaction recommendation system, characterized in that points are added to the common evaluation value of a service that is used simultaneously or continuously in time when using the service . The transaction data acquisition unit is characterized in that it filters messages exchanged in the external service and acquires at least a message transmitted at the time of purchase or reservation of a product or service as transaction data. Described transaction recommendation system. The transaction according to claim 2, wherein the transaction data acquisition unit further acquires a message transmitted when canceling the purchase or reservation of the product or service or withdrawing from the service as transaction data. Recommendation system. The transaction recommendation system according to claim 2 or 3, wherein the message is an e-mail message, a message stored in the cloud, or a message exchanged by SNS. The transaction recommendation system according to any one of claims 1 to 4, wherein the recommendation generation unit analyzes the transaction data and generates a recommendation regarding the purchase or reservation of a product or service. The feedback processing unit is characterized in that the content of the product or service is confirmed, the period required for the product to be actually used or the service to be actually used is predicted, and the predetermined period is determined. The transaction recommendation system according to any one of claims 1 to 5. The transaction recommendation system according to any one of claims 1 to 6, wherein the feedback processing unit acquires a response to the questionnaire from the user terminal and transmits the response to the recommendation generation unit. The transaction recommendation system according to claim 7, wherein the recommendation generation unit analyzes an answer to the questionnaire obtained from the feedback processing unit in addition to the transaction data, and generates the recommendation. The transaction recommendation system according to any one of claims 1 to 8, wherein the questionnaire includes a user's evaluation or comment regarding a product or service.