Recommendation control device

Abstract

This recommendation control device (10) comprises: an index derivation unit (11) that derives an engagement index indicating an intimacy degree between a certain service and a customer; and a control unit (12) that, on the basis of the derived engagement index, switches a recommendation method pertaining to a customer service to one recommendation method among a plurality of recommendation methods having mutually different purposes.

Description

Technical Field

[0001] The present disclosure relates to a recommendation control device that controls a recommendation system for providing customers with information on various services (broadly including product and content sales, service provision, etc.; details will be described later) as recommendation information.

Background Art

[0002] Conventionally, recommendation systems that provide customers with information on various services as recommendation information have been known and are used, for example, in e-commerce services, video distribution services, etc. In conventional recommendation systems, it has been mainstream to provide information that customers are likely to like according to the hobbies and preferences of individual customers. However, on the other hand, if one is too restricted by the hobbies and preferences of customers, there is a disadvantage that the diversity of the provided information is lacking, and opportunities for customers to make new discoveries and new realizations are reduced.

[0003] Therefore, in a recommendation system, while considering the hobbies and preferences of each customer, realizing an encounter with new and good products, content, etc. for the customer, that is, realizing serendipity, is considered to be one of the important elements for improving customer satisfaction. For example, Patent Document 1 below discloses a method for realizing serendipity.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] To achieve serendipity, one might consider switching to recommendation methods that expand the coverage (breadth of genres) of the products and content offered. However, simply switching to a recommendation method that expands coverage for all customers may result in a lower degree of suitability to individual customers' tastes and preferences, potentially leading to customers churn from their current service (e.g., subscriptions) (i.e., discontinuing use). Therefore, it is crucial to switch to a recommendation method appropriate for each customer.

[0006] This disclosure aims to address the above-mentioned challenges and enable the switching to a more appropriate recommendation method for each customer. [Means for solving the problem]

[0007] The recommendation control device according to this disclosure includes an indicator derivation unit that derives an engagement indicator representing the degree of intimacy between a service and a customer, and a control unit that, based on the engagement indicator derived by the indicator derivation unit, switches the recommendation method for the service for the customer to one of a plurality of recommendation methods, each with a different purpose. "Customer preference information" broadly refers to information including the customer's hobbies, tastes, preferences, etc. [Effects of the Invention]

[0008] According to this disclosure, it is possible to switch to an appropriate recommendation method for each customer. [Brief explanation of the drawing]

[0009] [Figure 1] This is a configuration diagram of the recommendation control device and related devices in the first to third embodiments. [Figure 2] This is a flowchart showing the processing performed by the indicator derivation unit. [Figure 3] (a) is a diagram illustrating the processing of the learning phase, and (b) is a diagram illustrating the processing of the estimation phase. [Figure 4]This is a flowchart showing the process executed by the recommendation control device of the first embodiment. [Figure 5] This figure shows the recommendation results in the first and second embodiments. [Figure 6] This is a flowchart showing the processing performed by the recommendation control device of the second embodiment. [Figure 7] This is a flowchart showing the process executed by the recommendation control device of the third embodiment. [Figure 8] This figure shows the recommendation results in the third embodiment. [Figure 9] This figure shows an example of the hardware configuration of a recommendation control device. [Modes for carrying out the invention]

[0010] Hereinafter, various embodiments of the recommendation control device according to this disclosure will be described in order with reference to the drawings. In the following, as the first embodiment, an embodiment in which the recommendation method is switched based on a change in an engagement index S representing the intimacy between a service and a customer will be described. As the second embodiment, an embodiment in which the recommendation method is switched based on the absolute value of the engagement index S in addition to the change in the engagement index S will be described. As the third embodiment, an embodiment in which a common recommendation method is set for multiple customers existing in the metaverse space will be described. In this disclosure, "service" is a broad concept that includes the sale of various goods and content, the provision of various services, etc. Below, examples of content recommendation in a video distribution service and examples of recommendation in a virtual space (e.g., the metaverse space) will be described, but in addition to these, it also includes various services provided by telecommunications carriers (i.e., companies that provide telecommunications services such as fixed telephones and mobile phones, so-called communication carriers). Furthermore, in this disclosure, "customer" means any person who is currently receiving or may receive services from now until in the future, and broadly includes (1) people who are currently receiving services, and (2) people who are not currently receiving services but may receive services in the future. If the scope of service provision extends to foreign countries, customers may also include all people (all of humanity).

[0011] [First Embodiment] Figure 1 shows the configuration of the recommendation control device 10 and its related devices in the first embodiment. As shown in Figure 1, the recommendation control device 10 includes an indicator derivation unit 11 that derives the aforementioned engagement indicator S in cooperation with a customer data management server 20 that stores and manages attribute data and behavioral data (hereinafter collectively referred to as "attribute and behavioral data") about various customers, a control unit 12 that controls the switching of recommendation methods described later, and a recommendation execution unit 13 that executes recommendations for various user terminals 40. Furthermore, the terminal 40 has the basic function of displaying the metaverse space on the display of the terminal 40 and displaying the avatar of the customer (user) operating the terminal 40 in the metaverse space in cooperation with a virtual space control server 30 that controls the virtual space (here, the metaverse space) as a whole.

[0012] The indicator derivation unit 11 described above has the function of generating a learning model M for estimating Next Purchase Intention (hereinafter referred to as "NPI") using machine learning, and deriving an engagement indicator S for the customer using the learning model M when actually performing customer recommendations. Here, NPI is represented by a binary value ("1" or "0"), where "1" means the customer wants to purchase / use the product / service again, and "0" means the customer does not want to purchase / use the product / service again. The engagement indicator S is an indicator that represents the degree of intimacy between a service and a customer, and takes the form of a continuous value within the range of 0 to 1, with a higher value (absolute value) indicating a higher degree of intimacy.

[0013] More specifically, as shown in Figures 2 and 3(a), the index derivation unit 11 uses data on NPI obtained from each of several customers of the service through questionnaires, etc. (the aforementioned values ​​of "1" or "0") as the target variable, and the attribute and behavioral data of each of the several customers as the explanatory variables, to generate a learning model M for estimating NPI using machine learning (learning phase F1 in Figure 2). This learning phase targets customers from whom NPI data has been obtained through questionnaires, etc. The generated learning model M is stored by the index derivation unit 11.

[0014] In practice, when performing recommendations for all customers (customers) who are the target of the recommendation, the metric derivation unit 11 queries the customer data management server 20 to obtain customer attribute and behavioral data from the customer data management server 20, and as shown in Figures 2 and 3(b), inputs the obtained customer attribute and behavioral data into the learning model M, thereby estimating (deriving) the metrics output from the learning model M as engagement metrics S for the customer (estimation phase F2 in Figure 2).

[0015] Returning to Figure 1, the functions of the control unit 12 are outlined. The control unit 12 has the function of switching the recommendation method for customer services to one of several recommendation methods, each with a different purpose, based on the engagement index derived by the index derivation unit 11. Specifically, the control unit 12 switches to one of several recommendation methods based on the value of the engagement index S derived by the index derivation unit 11 (for example, at least one of the change in value and the absolute value). Examples of "multiple recommendation methods" include a first recommendation method that makes recommendations for the purpose of expanding coverage and a second recommendation method that makes recommendations based on customer preference information. Of these, the first recommendation method can be, for example, a recommendation method using an existing natural language processing method called "Word2vec," and the second recommendation method can be, for example, a recommendation method based on hobby preference information using collaborative filtering. Furthermore, when evaluating the degree of recommendation coverage between recommendation methods using Word2vec and recommendation methods using collaborative filtering, using the ratio "number of items recommended by each method / number of items used for training," it is known that the Word2vec-based recommendation method yields a higher score than the collaborative filtering-based recommendation method, indicating that the Word2vec-based recommendation method can expand coverage.

[0016] The recommendation execution unit 13 shown in FIG. 1 has a function of selecting customer-oriented recommendation information (information such as services, products, etc.) using the recommendation method switched (set) by the control unit 12 and providing the selected recommendation information to the customer's terminal 40. For example, when the customer's terminal 40 cooperates with the virtual space control server 30 and the customer's avatar activities in the metaverse space, the recommendation execution unit 13 has a function of displaying the customer-oriented recommendation information in the metaverse space in association with the customer's avatar (for example, near the avatar) by cooperating with the virtual space control server 30.

[0017] Hereinafter, the process executed by the recommendation control device 10 of the first embodiment will be described according to the flowchart of FIG. 4. As a premise, it is assumed that at the start of the process, the index derivation unit 11 has completed the execution of the learning phase F1 in FIG. 2 and stored the generated learning model M.

[0018] First, the control unit 12 initializes the recommendation method to the first recommendation method (step S1 in FIG. 4).

[0019] The index derivation unit 11 obtains the customer's attribute and behavior data from the customer data management server 20 by querying the customer data management server 20 (step S2). Then, the index derivation unit 11 inputs the obtained customer's attribute and behavior data into the learning model M, and estimates (derives) the index output from the learning model M as the engagement index S for the customer (step S3). The obtained engagement index S is sent to the control unit 12.

[0020] The control unit 12 determines whether the engagement metric S obtained this time is the same as (maintained) or on an upward trend (step S4). Note that for the first time, since there is no previous engagement metric S, it is sufficient to determine that it is "the same (maintained)". If it is determined that the engagement metric S is maintained or on an upward trend (YES in step S4), it is unlikely that the customer will churn from the service, so the control unit 12 sets the recommendation method to the first recommendation method (a method for expanding coverage and realizing serendipity) (step S5).

[0021] On the other hand, if the engagement metric S is determined to be on a downward trend (NO in step S4), it is assumed that the customer will churn from the service, so the control unit 12 sets the recommendation method to the second recommendation method (a method that makes recommendations based on customer preference information) (step S6).

[0022] The information of the recommendation method set in step S5 or S6 above is transmitted from the control unit 12 to the recommendation execution unit 13, and the recommendation execution unit 13 executes recommendations according to the set recommendation method (step S7). For example, when the customer's terminal 40 is linked with the virtual space control server 30 and the customer's avatar is active in the metaverse space, the recommendation execution unit 13, in cooperation with the virtual space control server 30, displays recommendation information for the customer in the metaverse space in association with the customer's avatar. For example, Figure 5 shows an example where the customers are users A to C, and a recommendation method is set independently for each of users A to C, and the recommendation information for each user selected by the set recommendation method is displayed near each user's avatar in the metaverse space. Here, for example, for a user whose recommendation method has been switched to the first recommendation method, it is expected that recommendation information of a new genre will be displayed in light of the user's preference information. For example, a user with indoor hobbies and interests such as listening to music or playing games might be shown a list of recommendations in newer outdoor genres such as hiking and camping, along with a title like "For you, with high engagement metrics (closeness to the service)." On the other hand, for users whose recommendation method has been switched to the second recommendation method, it is expected that recommendations in genres and categories that match the user's preferences will be displayed. For example, a user with the aforementioned indoor hobbies and interests such as listening to music or playing games might be shown a list of recommendations in indoor genres that match those hobbies and interests, along with a title like "We will provide information that seems to match your hobbies and interests."

[0023] Returning to the process in Figure 4, steps S2 to S7 are repeated at predetermined intervals. That is, after step S7 is executed, the system waits until the predetermined period has elapsed, then returns to step S2 to obtain the latest attribute and behavioral data about the customer (step S2), and then continues to execute steps S3 and beyond.

[0024] According to the first embodiment described above, the recommendation method can be appropriately switched between the first recommendation method and the second recommendation method in response to changes in the engagement metric S derived for each customer. This makes it possible to prevent customers from abandoning the service while expanding the recommendation coverage and realizing serendipity.

[0025] Furthermore, as mentioned above, the indicator derivation unit generates a learning model for estimating NPI using machine learning, and by inputting customer attribute and behavioral data into the generated learning model, it derives the values ​​output from the learning model as engagement indicators for customers. In this way, it is possible to derive appropriate engagement indicators using a learning model.

[0026] Furthermore, if multiple customers exist in the virtual space, the control unit 12 can independently switch recommendation methods for each customer based on the engagement indicator S for each customer, thereby switching to a more appropriate recommendation method for each customer without being influenced by the attributes and behavioral data of other customers.

[0027] [Second Embodiment] Next, as a second embodiment, we will describe an embodiment in which the recommendation method is switched based on the absolute value of the engagement metric S, in addition to changes in the engagement metric S.

[0028] The configuration of the recommendation control device 10 and peripheral devices in the second embodiment is the same as the configuration in Figure 1 described above, so a redundant explanation will be omitted here.

[0029] The following describes the processes performed in the recommendation control device 10 of the second embodiment, following the flowchart in Figure 6.

[0030] Steps S1 (initial setup), S2 (acquisition of attribute and behavioral data), and S3 (derivation of engagement index S) in Figure 6 are the same as the processing in the first embodiment described above (Figure 4). In the second embodiment, in addition to the change in the derived engagement index S, the recommendation method is set (switched) as follows based on its absolute value. That is, the control unit 12 determines whether the engagement index S is 0.9 or higher (step S4A), and if the engagement index S is 0.9 or higher, it is unlikely that the customer will leave the service, so the unit proceeds to step S5 and sets the recommendation method to the first recommendation method (a method for expanding coverage and realizing serendipity).

[0031] On the other hand, if the engagement metric S is less than 0.9, the control unit 12 determines whether the engagement metric S is less than 0.1 (step S4B). If the engagement metric S is less than 0.1, it is assumed that the customer will abandon the service, so the system proceeds to step S6 and sets the recommendation method to the second recommendation method (a method that makes recommendations based on customer preference information).

[0032] Furthermore, if the engagement metric S is 0.1 or higher, the control unit 12 determines, as in the first embodiment, whether the engagement metric S obtained this time is the same as (maintained) or showing an increasing trend (step S4C). Note that for the first time, since there is no previous engagement metric S, it is sufficient to determine that it is "the same (maintained)". If it is determined that the engagement metric S is maintained or showing an increasing trend (YES in step S4C), it is unlikely that the customer will churn from the service, so the control unit 12 sets the recommendation method to the first recommendation method (step S5).

[0033] On the other hand, if the engagement metric S is determined to be on a downward trend (NO in step S4C), it is assumed that the customer will churn away from the service, so the control unit 12 sets the recommendation method to the second recommendation method (step S6).

[0034] Steps S7 and S8 are the same as in the first embodiment. In step S7, the recommendation execution unit 13 performs a recommendation according to the set recommendation method (step S7). As a result, for example as shown in Figure 5, if the customers are users A to C, a recommendation method is set independently for each of users A to C, and the recommendation information for each user selected by the set recommendation method is displayed near each user's avatar in the metaverse space. After the execution of step S7, the system waits until a predetermined period has elapsed and returns to step S2, so that the processing of steps S2 to S7 is repeated at predetermined intervals.

[0035] According to the second embodiment described above, in addition to changes in the engagement metric S derived for each customer, the recommendation method can be appropriately switched between the first recommendation method and the second recommendation method based on the absolute value of the engagement metric S. Specifically, if the engagement metric S is 0.9 or higher, it is unlikely that the customer will churn from the service, so the recommendation method can be immediately set to the first recommendation method regardless of changes in the engagement metric S (increasing / decreasing trend). Also, if the engagement metric S is less than 0.1, it is expected that the customer will churn from the service, so the recommendation method can be immediately set to the second recommendation method regardless of changes in the engagement metric S (increasing / decreasing trend). This makes it possible to prevent customers from churn from the service while expanding the coverage of recommendations and realizing serendipity.

[0036] [Third Embodiment] Next, as a third embodiment, we will describe an embodiment in which a common recommendation method is set for multiple customers (users A to C) existing in the metaverse space shown in Figure 8.

[0037] The configuration of the recommendation control device 10 and peripheral devices in the third embodiment is the same as the configuration in Figure 1 described above, so a redundant explanation will be omitted here.

[0038] The following describes the processes performed in the recommendation control device 10 of the third embodiment, following the flowchart in Figure 7.

[0039] Step S1 (initial setup) in Figure 7 is the same as the process in the first embodiment described above (Figure 4). In the next step S2A, the metric derivation unit 11 queries the customer data management server 20 to obtain attribute and behavioral data for multiple customers from the customer data management server 20 (step S2A). The metric derivation unit 11 then inputs the obtained attribute and behavioral data for each of the multiple customers into the learning model M, and estimates (derives) the metrics output from the learning model M as the metrics for each of the multiple customers (step S3A). The obtained metrics for each customer are sent to the control unit 12.

[0040] The control unit 12 determines the value of the indicator with the lowest absolute value among the multiple customer indicators derived in step S3A (i.e., the minimum value) and sets that as the engagement indicator S for the multiple customers (step S3B).

[0041] The processing in steps S4A to S8 thereafter is the same as in the second embodiment described above. That is, based on the changes and absolute values ​​of the engagement metric S set in step S3B, the recommendation method common to customers (users A to C) is set (switched) as follows.

[0042] In other words, the control unit 12 determines that if the engagement index S is 0.9 or higher, it is unlikely that the customer will abandon the service, and therefore sets the recommendation method to the first recommendation method (a method for expanding coverage and realizing serendipity) (steps S4A, S5).

[0043] Furthermore, if the engagement index S is less than 0.1, the control unit 12 assumes that the customer will abandon the service, and therefore sets the recommendation method to the second recommendation method (a method that makes recommendations based on customer preference information) (steps S4B, S6).

[0044] Furthermore, if the engagement index S is 0.1 or greater and less than 0.9, the control unit 12 sets the recommendation method to either the first recommendation method or the second recommendation method in accordance with the change in the engagement index S, similar to the first embodiment (steps S4C, S5, S6). In other words, if the engagement index S is maintained or increasing, it is unlikely that customers will leave the service, so the control unit 12 sets the recommendation method to the first recommendation method (step S5). If the engagement index S is decreasing, it is expected that customers will leave the service, so the recommendation method is set to the second recommendation method (step S6).

[0045] Then, in step S7, the recommendation execution unit 13 executes the recommendation according to the set recommendation method (step S7). As a result, a common recommendation method is set for customers (users A to C), and then, for example as shown in Figure 5, recommendation information for each user selected by the common recommendation method is displayed near each user's avatar in the metaverse space. After the execution of step S7, the system waits for a predetermined period of time to elapse and then returns to step S2A, so that the processing of steps S2A to S7 is repeated at predetermined intervals.

[0046] According to the third embodiment described above, a common recommendation method is set for multiple customers, and recommendation information selected for each user by the common recommendation method is provided to each user. In this case, the recommendation information for each user is not necessarily the same, but if the attributes and behavioral data of users A to C existing in the metaverse space are similar (for example, if they have common hobbies or similar preferences), it is quite possible that the recommendation information for each user will also be similar, and it is possible to provide highly useful recommendation information that is likely to resonate with (impress) multiple customers.

[0047] [Regarding other forms of modification] In the first to third embodiments described above, examples were shown in which customer engagement metrics were derived using NPI, but the use of NPI is not mandatory. Besides NPI, for example, the Net Promoter Score (hereinafter referred to as "NPS"), known as a metric for measuring customer loyalty, can be used. NPS® is a metric that measures the degree of service recommendation using integers from 0 to 10 (11 levels). For example, [0-6] can be replaced with detractors and [7-10] with promoters, resulting in a binary value ("0" or "1"), and it can be used as the target variable in a binary classification model, similar to the NPI described above. Reference site: https: / / www.nttcoms.com / service / nps / summary / #index02

[0048] Furthermore, in the first to third embodiments described above, we showed an example where the recommendation method is set to the first recommendation method when the engagement metric S is maintained (the same as the previous time), but it is not mandatory to set it to the first recommendation method. For example, if preventing customer churn is a priority, the recommendation method may be set to the second recommendation method when the engagement metric S is maintained.

[0049] Furthermore, in the first to third embodiments described above, an example was shown in step S1 of Figure 4 where the recommendation method is initially set to the first recommendation method. However, it is not mandatory to initially set it to the first recommendation method. For example, if preventing customer churn is a priority, the recommendation method may be initially set to the second recommendation method.

[0050] Furthermore, while the first embodiment described above shows an example of setting (switching) the recommendation method based on the "change" in the engagement metric S, and the second and third embodiments show examples of setting (switching) the recommendation method based on the "change and absolute value" of the engagement metric S, the recommendation method may also be set (switched) based only on the "absolute value" of the engagement metric S.

[0051] Furthermore, in the third embodiment described above, an example was shown in which the recommendation method is set (switched) based on the "change and absolute value" of the engagement indicator S, similar to the second embodiment. However, the recommendation method may also be set (switched) based on the "change" of the engagement indicator S, similar to the first embodiment, or it may be set (switched) based on the "absolute value" of the engagement indicator S.

[0052] Furthermore, in the third embodiment described above, the value of the indicator with the lowest absolute value among the indicators for each of the multiple customers (minimum value) was used as the engagement indicator S for the multiple customers, and an example was shown in which the recommendation method was set (switched) based on the change and absolute value of the engagement indicator S. However, instead of the minimum value, the arithmetic mean, median value, etc. of the indicators for each of the multiple customers may be used as the engagement indicator S for the multiple customers.

[0053] Furthermore, in the first to third embodiments described above, examples were shown of switching between a first recommendation method aimed at expanding coverage and a second recommendation method based on customer preference information. However, there may be three or more recommendation methods as candidates for switching.

[0054] As mentioned above, an example of "services" in this disclosure includes various services provided by telecommunications carriers. Assuming that data equivalent to the next purchase intention (NPI) for the entire brand of a particular telecommunications carrier has been obtained, the method described above can be used to obtain the NPI trend score for the entire brand of that telecommunications carrier as an engagement metric, and the appropriate recommendation method can be switched based on the obtained engagement metric. Specifically, the method of this disclosure can be applied to recommendation services that recommend information about various services provided by the telecommunications carrier (for example, telecommunications line provision services (both or both fixed-line and mobile phone services), and services that can be paid for comprehensively along with the usage fees for those services).

[0055] The gist of this disclosure is found in the following [1] to [6]. [1] An indicator derivation unit that derives an engagement indicator that represents the degree of closeness between a service and a customer, A control unit that, based on the engagement metrics derived by the metric derivation unit, switches the recommendation method for the customer regarding the service to one of a plurality of recommendation methods with different purposes, A recommendation control device equipped with [a specific feature]. [2] The recommendation control device according to [1], wherein the plurality of recommendation methods include a first recommendation method that makes recommendations for the purpose of expanding coverage and a second recommendation method that makes recommendations based on the customer's preference information. [3] The control unit switches the recommendation method based on the value of the engagement indicator, the recommendation control device according to [1] or [2]. [4] The recommendation control device according to [3], wherein the control unit switches the recommendation method based on at least one of the change and absolute value of the engagement indicator. [5] The index derivation unit is, Using data on the intention to make a next purchase obtained from each of several customers of the aforementioned service as the dependent variable, and using the attribute and behavioral data of each of the aforementioned customers as the independent variables, a learning model for estimating the intention to make a next purchase is generated by machine learning. A recommendation control device according to any one of [1] to [4], wherein the customer's attribute and behavioral data are input into the generated learning model, and the engagement indicator output from the learning model is derived as the engagement indicator for the customer. [6] If there are multiple customers in the virtual space, The control unit switches the recommendation method to a recommendation method common to the multiple customers based on the lowest engagement metric among the engagement metrics of the multiple customers, according to any one of [1] to [5]. [7] If there are multiple customers in the virtual space, The control unit switches the recommendation method independently for each customer based on the engagement indicator for each customer derived by the indicator derivation unit, according to any one of [1] to [5].

[0056] [Explanation of terms, explanation of hardware configuration (Figure 9), etc.] The block diagrams used in the description of the above embodiments show functional units. These functional blocks (components) are realized by any combination of at least one of hardware and software. Furthermore, the method of realizing each functional block is not particularly limited. That is, each functional block may be realized using one device that is physically or logically coupled, or it may be realized using two or more physically or logically separated devices that are directly or indirectly connected (for example, using wired or wireless connections). A functional block may also be realized by combining the above one device or the above multiple devices with software.

[0057] Functions include, but are not limited to, judgment, decision, judgment, calculation, calculation, processing, derivation, investigation, exploration, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, assumption, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating (mapping), and assigning. For example, a functional block (configuration part) that enables transmission is called a transmitting unit or transmitter. As mentioned above, the method of implementation is not particularly limited.

[0058] For example, a recommendation control device in one embodiment of the present disclosure may function as a computer that performs the processing of the present disclosure. Figure 9 is a diagram showing an example of the hardware configuration of a recommendation control device 10 according to one embodiment of the present disclosure. The recommendation control device 10 described above may be physically configured as a computer device including a processor 1001, memory 1002, storage 1003, communication device 1004, input device 1005, output device 1006, bus 1007, etc.

[0059] In the following explanation, the term "device" can be replaced with "circuit," "device," "unit," etc. The hardware configuration of the recommendation control device 10 may include one or more of the devices shown in the figure, or it may be configured to omit some of the devices.

[0060] Each function in the recommendation control device 10 is realized by loading predetermined software (programs) onto hardware such as the processor 1001 and memory 1002, which causes the processor 1001 to perform calculations, control communication by the communication device 1004, and control at least one of the reading and writing of data in the memory 1002 and storage 1003.

[0061] The processor 1001 controls the entire computer, for example, by running an operating system. The processor 1001 may consist of a central processing unit (CPU) that includes interfaces with peripheral devices, control units, arithmetic units, registers, and so on.

[0062] Furthermore, the processor 1001 reads programs (program code), software modules, data, etc., from at least one of the storage 1003 and the communication device 1004 into the memory 1002 and executes various processes accordingly. The program used is one that causes the computer to execute at least a part of the operations described in the above embodiment. Although it has been explained that the various processes are executed by one processor 1001, they may be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented by one or more chips. The program may also be transmitted from a network via a telecommunications line.

[0063] Memory 1002 is a computer-readable recording medium and may consist of at least one of the following: ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), RAM (Random Access Memory), etc. Memory 1002 may also be called a register, cache, main memory, etc. Memory 1002 can store executable programs (program code), software modules, etc., for carrying out a wireless communication method according to one embodiment of the present disclosure.

[0064] Storage 1003 is a computer-readable recording medium and may consist of at least one of the following: an optical disc such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disc, a digital multipurpose disc, a Blu-ray® disc), a smart card, flash memory (e.g., a card, a stick, a key drive), a floppy® disk, a magnetic strip, etc. Storage 1003 may also be called an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, server, or other suitable medium including at least one of memory 1002 and storage 1003.

[0065] The communication device 1004 is hardware (transceiver / receiver device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as a network device, network controller, network card, communication module, etc. The communication device 1004 may be configured to include, for example, a high-frequency switch, duplexer, filter, frequency synthesizer, etc., in order to implement at least one of frequency division duplex (FDD) and time division duplex (TDD).

[0066] The input device 1005 is an input device that accepts input from an external source (e.g., a keyboard, mouse, microphone, switch, button, sensor, etc.). The output device 1006 is an output device that outputs to an external source (e.g., a display, speaker, LED lamp, etc.). The input device 1005 and the output device 1006 may be configured as an integrated unit (e.g., a touch panel).

[0067] Furthermore, each device, such as the processor 1001 and memory 1002, is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or different buses may be configured for each device.

[0068] Furthermore, the recommendation control device 10 may be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array), and some or all of each functional block may be realized by such hardware. For example, the processor 1001 may be implemented using at least one of these hardware components.

[0069] Information notification is not limited to the embodiments described herein and may be carried out by other means. For example, information notification may be carried out by physical layer signaling (e.g., DCI (Downlink Control Information), UCI (Uplink Control Information)), upper layer signaling (e.g., RRC (Radio Resource Control) signaling, MAC (Medium Access Control) signaling, broadcast information (MIB (Master Information Block), SIB (System Information Block))), other signals, or combinations thereof. RRC signaling may also be called RRC messages, and may be, for example, RRC Connection Setup messages, RRC Connection Reconfiguration messages, etc.

[0070] Each aspect / embodiment described in this disclosure includes LTE (Long Term Evolution), LTE-A (LTE-Advanced), SUPER 3G, IMT-Advanced, 4G (4th generation mobile communication system), 5G (5th generation mobile communication system), 6th generation mobile communication system (6G), xth generation mobile communication system (xG) (xG (where x is, for example, an integer or decimal)), FRA (Future Radio Access), NR (new Radio), New radio access (NX), Future generation radio access (FX), W-CDMA (registered trademark), GSM (registered trademark), CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi (registered trademark)), IEEE 802.16 (WiMAX (registered trademark)), and IEEE This may apply to at least one system utilizing 802.20, UWB (Ultra-WideBand), Bluetooth®, or other appropriate systems, and to next-generation systems extended, modified, created, or defined based thereon. It may also apply to a combination of multiple systems (for example, a combination of at least one of LTE and LTE-A with 5G).

[0071] The processing procedures, sequences, flowcharts, etc., of each aspect / embodiment described herein may be reordered, provided they are consistent with each other. For example, the methods described herein present various step elements in an exemplary order and are not limited to that specific order.

[0072] Input and output information may be stored in a specific location (e.g., memory) or managed using a management table. Input and output information may be overwritten, updated, or appended to. Output information may be deleted. Input information may be transmitted to other devices.

[0073] The determination may be made by a value represented by 1 bit (0 or 1), by a boolean value (true or false), or by a numerical comparison (for example, a comparison with a predetermined value).

[0074] Each aspect / embodiment described herein may be used individually, in combination, or switched between as needed during implementation. Furthermore, notification of specific information (e.g., notification that "X is") is not limited to explicit notification, but may also be implicit (e.g., by not providing such notification).

[0075] Although the present disclosure has been described in detail above, it will be clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the intent and scope of the present disclosure as defined by the claims. Therefore, the descriptions in the present disclosure are illustrative and not intended to be restrictive in any way.

[0076] Software should be broadly interpreted to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, execution threads, procedures, functions, and so on, whether they are called software, firmware, middleware, microcode, hardware description languages, or by any other name.

[0077] Furthermore, software, instructions, information, etc., may be transmitted and received via a transmission medium. For example, if software is transmitted from a website, server, or other remote source using at least one of wired technology (such as coaxial cable, fiber optic cable, twisted pair, or digital subscriber line (DSL)) and wireless technology (such as infrared or microwave), then at least one of these wired and wireless technologies is included in the definition of a transmission medium.

[0078] The information, signals, etc. described in this disclosure may be represented using any of the various different techniques. For example, the data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0079] In addition, terms used in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings. For example, at least one of the channel and symbol may be a signal (signaling). Also, a signal may be a message. Furthermore, a component carrier (CC) may be called a carrier frequency, cell, frequency carrier, etc.

[0080] The terms “system” and “network” as used in this disclosure are interchangeable.

[0081] Furthermore, the information, parameters, etc., described in this disclosure may be expressed using absolute values, relative values ​​from a given value, or other corresponding information. For example, wireless resources may be indicated by an index.

[0082] The names used for the parameters described above are not restrictive in any way. Furthermore, the formulas and other expressions using these parameters may differ from those expressly disclosed in this disclosure. Various channels (e.g., PUCCH, PDCCH, etc.) and information elements can be identified by any suitable name, and therefore, the various names assigned to these various channels and information elements are not restrictive in any way.

[0083] As used in this disclosure, the terms “determining” and “determining” may encompass a wide variety of actions. “Determining” may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiry (e.g., searching in a table, database, or other data structure), and ascertaining. “Determining” may also include, for example, receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, and accessing (e.g., accessing data in memory). Furthermore, "judgment" and "decision" can include considering something as having been "judged" or "decided" after resolving, selecting, choosing, establishing, comparing, etc. In other words, "judgment" and "decision" can include considering something as having been "judged" or "decided" after some action. Also, "judgment (decision)" can be reinterpreted as "assuming," "expecting," or "considering."

[0084] In this disclosure, the phrase "based on" does not mean "based solely on" unless otherwise specified. In other words, the phrase "based on" means both "based solely on" and "based at least on."

[0085] Any reference to elements using the designations “first,” “second,” etc., as used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient way to distinguish between two or more elements. Accordingly, references to the first and second elements do not imply that only two elements may be employed, or that the first element must precede the second element in any way.

[0086] Where the terms “include,” “including,” and variations thereof are used in this disclosure, these terms are intended to be inclusive, as is the term “comprising.” Furthermore, the term “or” as used in this disclosure is not intended to mean exclusive OR.

[0087] In this disclosure, if articles are added through translation, such as a, an, and the in English, this disclosure may include the fact that the noun following these articles is plural.

[0088] In this disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "combine" may be interpreted similarly to "different." [Explanation of symbols]

[0089] 10... Recommendation control unit, 11... Index derivation unit, M... Learning model, 12... Control unit, 13... Recommendation execution unit, 20... Customer data management server, 30... Virtual space control server, 40... Terminal, 1001... Processor, 1002... Memory, 1003... Storage, 1004... Communication device, 1005... Input device, 1006... Output device, 1007... Bus.

Claims

1. An indicator derivation unit that derives an engagement indicator that represents the level of intimacy between a service and a customer, Based on the changes and absolute values ​​of the engagement indicators derived by the indicator derivation unit, the control unit switches the recommendation method for the customer regarding the service to one of a first recommendation method that makes recommendations for the purpose of expanding coverage and a second recommendation method that makes recommendations based on the customer's preference information, and displays the recommendation information selected by the switched recommendation method on the customer's terminal. Equipped with, The index derivation unit is, Using data on the intention to make a next purchase obtained from each of several customers of the aforementioned service as the dependent variable, and using the attribute and behavioral data of each of the aforementioned customers as the independent variables, a learning model for estimating the intention to make a next purchase is generated by machine learning. By inputting the customer's attributes and behavioral data into the generated learning model, the engagement metrics output from the learning model are derived as engagement metrics for the customer. Recommendation control device.

2. The control unit communicates with a virtual space control server that controls the virtual space displayed on the customer's terminal, and if it determines, based on the information obtained from the virtual space control server, that there are avatars corresponding to each of the multiple customers in the virtual space, it switches the recommendation method to a recommendation method common to the multiple customers based on the lowest engagement metric among the engagement metrics of the multiple customers. The recommendation control device according to claim 1.

3. The control unit communicates with a virtual space control server that controls the virtual space displayed on the customer's terminal, and if it determines, based on the information obtained from the virtual space control server, that there are avatars corresponding to multiple customers in the virtual space, it switches the recommendation method independently for each customer based on the engagement index for each customer derived by the index derivation unit. The recommendation control device according to claim 1.

Images