JP2022186308A - Message transmission scenario generation device, message transmission scenario generation method and message transmission scenario generation program - Google Patents

Abstract

To adapt a transmission channel and transmission timing for transmitting a message to an intended receiver to information related to the content of the message.SOLUTION: According to a disclosed technique, a message transmission scenario generation device (100) and the like are provided including a message reception part (104) for receiving information including a message from a source (102), a receiver information acquisition part (106) for acquiring information related to an intended receiver of the message, a message type specification part (108) for specifying a message type based on information including the information related to the content of the message, and a message transmission scenario generation part (112) for outputting information including a transmission channel and transmission timing for transmitting the message to the receiver in response to an input of the information including the information related to the message type and the receiver to a learned model (114).SELECTED DRAWING: Figure 1

Description

The present invention relates to a message transmission scenario generation device, a message transmission scenario generation method, and a message transmission scenario generation program.

As used herein, "message" refers to information sent from an originator to an intended recipient (including groups of intended recipients), regardless of form or amount of information. Also, "message communication scenario" means information that indicates how a message or combination of messages may be sent from an originator to an intended recipient. Here, "communication" is not limited to telecommunication or digital communication, but also includes face-to-face communication through visits or communication through the mail.

In recent years, a variety of communication channels have evolved for sending messages.

For example, messaging services for transmitting messages, which may contain information in text form, to communication terminals are widespread. Message communication techniques include those that use ordinary mail addresses as destinations and those that use telephone numbers or addresses similar to telephone numbers as destinations. An address that conforms to a telephone number is, for example, an address linked in some way to a mobile telephone number.

SMS (Short Messaging Service) is a typical example of message communication using a telephone number or an address equivalent to a telephone number as a destination. SMS can transmit a string of characters of limited length (which may also include symbols) to a telephone number as the destination. SMS makes it possible to send messages directly to mobile phones, and since it is generally provided as a basic service for mobile phone contracts, SMS provides immediate and widespread communication of relatively short messages to mobile phone users. It can be said that it is a message transmission channel that can be used. However, SMS cannot transmit rich text or multimedia content including still images, sounds, or moving images.

Some messaging that uses a phone number or phone number-like address as a destination can carry rich text and multimedia content. Examples of such messaging include MMS (Multimedia Messaging Service) and RCS (Rich Communication Services).

Other communication channels in the form of digital communication include communication through social media and communication through notifications on application programs installed on communication terminals. To use these channels, the intended recipient must be a user of the applicable social media or application program, provided that the intended recipient is such a user and Effective message transmission when the originator (the originator of the message itself, or an entity that has the originator of the message as a customer and undertakes the transmission of the message from the originator, etc.) is in a state where it is possible to transmit to the recipient. can be a channel. Furthermore, there is also communication through a chatbot as an aspect of communication in the form of dialogue. Chatbots can also be a strong choice as a messaging channel, especially if the messages are to be effectively communicated in a conversational exchange.

On the other hand, traditional non-digital communication channels include telephone, mail, leaflet distribution, and visits. These traditional messaging channels generally involve some physical movement and do not have immediacy unlike digital communication, but they can be effective messaging channels depending on the content of the message. For example, the telephone is effective as a means of transmitting messages that require flexible communication based on the response of the other party, and the postal mail is effective as a means of reliably transmitting relatively complex messages. can have

Thus, there are various options for the transmission channel when sending messages. Then, when the message is addressed to the intended recipient, depending on which delivery channel is selected from among those alternatives, the message will be read by that recipient (and where appropriate, there will also be a The likelihood that the recipient will take the expected further action) may vary.

In addition to the transmission channel, transmission timing is also a consideration when sending a message. when a message is sent to its intended recipient, e.g., by what time of day the message is sent or in relation to the subject matter the message is addressed at The likelihood that a message will be read by the recipient (and, where appropriate, the recipient will then take further actions expected with respect to the message) may vary.

However, the originator of the message (either the originator of the message itself, or an entity that has the originator of the message as a customer and undertakes the origination of the message from the originator, etc.), however, often selects recipients as the target of the origination of the message. Although various ideas have been devised, messages are sent at a uniform transmission timing and transmission channel without trying to optimize the transmission channel and transmission timing for sending messages. Such an approach cannot be fully expected to effectively convey the message to the intended recipient, even if the selection of the intended recipient is appropriate.

Under such circumstances, there exists a technique having the following aspect as a prior art (Patent Document 1). That is, extracting the particular user attributes associated with a particular user who will be the recipient of the transmitted content, and applying machine learning techniques to the user data, including the user attributes for a set of other users. communication channels for extracting learned data and transmitting said content to said recipients based on said learned data and specific user attributes, and It is a technique for specifying communication specifications including timings of transmission (claims 1, 4 and 5 and paragraph [0108], etc. of the document).

In this technology, as a result of executing a machine learning model using the learned data and the specific user attribute, the option that is predicted to be the most successful in achieving the target result from among a plurality of options regarding communication specifications can be specified (such as paragraphs [0041] and [0108] of that document). Here, examples of communication specifications to be output include communication channel (options include e-mail, SMS message, and application-based communication), transmission timing (options include, for example, a time range starting from the current time). regularly spaced set of times), and the type of content (options include, for example, content identifying a single product most recently viewed on a website by a particular user, among various users (paragraphs [0108], [0059], [0046] of that document) identifying four of the most viewed products on the website over the past week, and including identifying discounts. , and [0063], etc.). Also, examples of targeted outcomes are detecting clicks on links within a correspondence, detecting that a correspondence has been opened, detecting that a correspondence results in a purchase, chat sessions and to detect that the form has been completed (paragraph [0090] of the document).

WO2019/067257

The disclosed technology enables the delivery channel and delivery timing of a message to be adapted to information related to the content of said message. A message delivery scenario generator, a message delivery scenario An object of the present invention is to provide a generation method and a message transmission scenario generation program.

According to the disclosed technique, a message receiving unit receives information including a message from an originator, a recipient information obtaining unit obtains information about the intended recipient of the message, and information related to the content of the message is obtained. a message type identification unit that identifies a message type based on information included in the message type identification unit; and a transmission channel that sends the message to the recipient in response to input of information including information about the message type and the recipient to a trained model; A message transmission scenario generation unit that outputs information including transmission timing is provided.

Also according to the disclosed technique is a computer-implemented method for generating a message delivery scenario, comprising: receiving information containing a message from an originator; obtaining information about an intended recipient of said message; identifying a message type based on information including information related to content of the message; and receiving the message in response to inputting information including information about the message type and the recipient to a trained model. and outputting information including transmission channels and transmission timings to be sent to people.

Further, according to the disclosed technique, receiving information including a message from an originator, obtaining information about the intended recipient of the message, and determining a message type based on the information including information related to the content of the message. Identifying and executing a process of outputting information including a transmission channel and transmission timing for transmitting the message to the recipient in response to input of information including information on the message type and the recipient to the trained model. A messaging scenario generator is provided that allows

According to the disclosed technology, a messaging scenario generator, message A transmission scenario generation method and a message transmission scenario generation program can be provided.

FIG. 1 is a functional block diagram of the message transmission scenario generation device of the embodiment. FIG. 2 is a flowchart illustrating an example of processing in the embodiment. FIG. 3 is a flow chart showing an example of processing (subroutine) for receiving information including a message from a sender in the embodiment. FIG. 4 is a flowchart illustrating an example process (subroutine) for obtaining information about the intended recipient of a message, in an embodiment. 5A and 5B are flowcharts showing an example of processing (subroutine) for identifying the message type based on information including information related to the content of the message in the embodiment. Among them, FIG. 5A shows an example of identifying a message type by determining whether an extracted character string extracted from a message contains a keyword for identifying a message type, and FIG. 5B shows information including a message and learned data. An example of specifying the message type based on each is shown below. FIG. 6 shows processing (subroutine) for outputting information including transmission channels and transmission timings for sending messages to recipients in accordance with input of information including information on message types and recipients to a trained model in an embodiment. 3 is a flow chart showing an example of FIG. 7 is a table showing examples of generated messaging scenarios with information including delivery channels and delivery timings for sending messages to recipients, in an embodiment. FIG. 8 is a diagram showing the hardware configuration of the embodiment.

In the disclosed embodiment, an "originator" is a subject that initiates a message, and can be any organization, individual, or group of individuals. Typically, apart from the originator, there is an entity that has the originator as a customer and undertakes the transmission of the message from the originator, and the entity is the message transmission scenario generation device, the message transmission scenario generation method, or the message transmission scenario generation. A program (hereinafter sometimes simply referred to as "message transmission scenario generation device, etc.") is used. However, it is not limited to this, and any entity including the originator itself may use the message transmission scenario generation device or the like.

In the disclosed embodiments, a messaging scenario generator, a messaging scenario generating method, or a messaging scenario generating program generates intended recipients (hereinafter simply "recipients ) may be used to transmit the message, or the message transmission scenario may be generated. When a messaging scenario generator or the like causes a message to be conveyed, typically through equipment or personnel of a telecommunications carrier, etc., for message transmission, such as sending a message addressed to the intended recipient. process is carried out. However, a messaging scenario generator or the like may have the function of performing the process for direct messaging. Here, "carriers, etc." refer to entities capable of carrying out processes for message transmission, and typically have devices connected to the recipient's communication terminal via a communication network. Is a carrier. An example of a "communications carrier, etc." is a carrier in mobile communications (mobile network operator: MNO (Mobile Network Operator)). In addition, MVNO (Mobile Virtual Network Operator) that uses the communication line of this carrier is also an example of the "telecommunications carrier, etc." In addition, businesses that operate social media, businesses that operate messaging services provided accompanying social media, etc., provide communication terminals and OS or application programs, and one function of the OS or application programs Business operators that operate messaging services as well as operators that operate services such as chatbots are also examples of "telecommunications carriers, etc." as used herein. "Telecommunications carriers, etc." also includes entities that carry out processes for message transmission through telephone, postal mail, leaflet distribution, visits, etc., when the message transmission is conventional messaging that does not rely on digital communication. included.

<Overview of Embodiment>
In the disclosed embodiments, a computer that implements a message transmission scenario generation device, a message transmission scenario generation method, or a computer that executes processing by a message transmission scenario generation program receives information including a message from an originator, and determines the intent of the message. obtaining information about the recipient, identifying a message type based on information including information related to the content of the message, and inputting information including information about the message type and the recipient to a trained model In response, output information including the delivery channel and delivery timing for sending the message to the recipient.

The idea underlying the disclosed embodiments is that among the various options for delivery channel and delivery timing, it is assumed that the message is more likely to be effectively delivered to the intended recipient. The idea is to select an option and send a message with the selected transmission channel and transmission timing.

In addition, the disclosed embodiments are based on the following findings regarding how to select the transmission channel and transmission timing. That is, when sending a message to the intended recipient, the likelihood that the message will be effectively communicated to that recipient, specifically, for example, that the message will be read by that recipient (if appropriate). In addition, there is a relatively high probability that the recipient will take further actions that are expected with respect to the message. It is the knowledge that it differs depending on the This also means that there is a correlation between information about message type and recipient, and the delivery channel and delivery timing over which the message is more likely to be effectively delivered to that recipient. leads to the knowledge that In light of this knowledge, using machine learning technology, based on the message type and information about the recipient, there is a relatively high probability that the message will be effectively conveyed to the recipient. can be estimated. The above "message type" is a type specified based on information including information related to the content of the message, and is a type based on the content of the message, a type indicating the purpose of sending the message to the recipient, It may include a type or the like indicating the type of recipient.

Here, based on the content of the message itself and information about the recipient, rather than the message type, the transmission channel and transmission timing that make it relatively likely that the message will be effectively delivered to the recipient are estimated. It is also possible. However, the contents of the messages themselves are diverse, and in some cases, the characteristics of the messages are not necessarily clearly expressed in the contents of the messages themselves. Therefore, using a message type that reflects the characteristics of the message, based on the message type and information on the recipient, a transmission channel and transmission timing that relatively increase the possibility that the message will be effectively transmitted to the recipient are proposed. By estimating , it is possible to improve the accuracy of the estimation.

In order to support the findings above, we first consider the “situation related to the beneficiary”. When a message is sent to the intended recipient, the transmission channel and the transmission timing that make it more likely that the message will be effectively delivered to the recipient will differ depending on the circumstances of the recipient. It is as follows.

With regard to the relationship between the transmission channels and the intended recipient's situation, for example, what transmission channels are available on the communication terminal owned by the intended recipient, and whether the intended recipient uses which transmission channel. Various recipient circumstances, such as whether they tend to respond to messages more often, will influence which delivery channel options are appropriate for that recipient.

The relationship between delivery timing and intended recipient status is as follows. For example, if the message is an advertisement related to a product, even if the intended recipient receives the message via messaging during working hours, the message will not be opened on the spot, and the opening will be postponed. As a result, it is quite possible that it will remain unread. In contrast, if the intended recipient received such a message while at home on a holiday, it would generally be more likely to be opened and read. And the working hours mentioned above will vary depending on the intended recipient, whether they tend to read their e-mails collectively at certain times of the day, and if so, when. Behavioral tendencies, such as bandages, that can affect delivery timing also vary with the intended recipient.

As can be seen, the delivery channel and delivery timing over which the message is more likely to be effectively delivered to the intended recipient will vary depending on the recipient's circumstances.

Next, let us consider “message type” as a matter that further supports the above findings. When a message is sent to its intended recipient, the transmission channel and timing for which the message is likely to be effectively The following is an example showing that it differs depending on the type.

For example, consider the following situation, which corresponds to the case where transmission timing is configured in a plurality of stages. That is, suppose the message is disaster-related evacuation information, and the disaster is a flood that can hit an area in a few days. In this scene, when the possibility of actual flood damage increases later, a separate message, such as an evacuation recommendation, will be sent. Send a message with a map to residents in the affected areas to inform them of evacuation sites in advance. Therefore, it can be said that the message type based on the content of the message is disaster-related, and particularly a message related to flood damage several days before the expected occurrence. In this situation, if the recipient is a user of an application program intended to convey disaster prevention information and has a history of frequent use of the program, the message is disaster-related. In terms of type, the program may be the most compelling choice of delivery channel. In this case, the trained model may select that program as the delivery channel depending on the input of information about the message type and recipient. Even if the usage history of that particular recipient is unknown, the trained model can make inferences based on universal trained data. For example, if users of the program generally tend to use the program frequently, the trained model can be used to obtain past messages to various people, not limited to the recipient, when the message type is related to disasters. Using learned data about the message open rate for sending, and based on the information that the recipient is a user of the program, the program as a delivery channel that is relatively likely to be effectively delivered to select. As for the transmission timing, immediate message issuance is selected in the above-mentioned first-stage message transmission in light of the disaster-related message type. The timing of the second stage of message transmission is planned when the risk of actual flooding increases and when a specific warning is issued.

On the other hand, if the message type is an advertisement for a product, the delivery channel and delivery timing over which the message is more likely to be effectively delivered to the recipient can vary widely. Here, if the recipient is a frequent user of a particular social media and has subscribed to receive messages from the manufacturer of the product through that social media, then the most prevalent channel of communication is An option could be its social media. In this case, the trained model may select that social media as the delivery channel depending on the input of information about the message type and recipient. On the other hand, if the recipient has a history of high open rates for RCS messages and is willing to receive RCS messages from the manufacturer of the product, RCS is the most likely delivery channel choice. obtain. (It should be noted that even if the recipient's open rate history for RCS messages is not available, RCS messages can be delivered directly to mobile phones and can be notified of receipt on pop-up screens, and Given its high visibility, RCS can be a viable delivery channel of choice if the recipient is able to receive RCS messages from the manufacturer of the product.) In this case, the trained model may select RCS as the delivery channel depending on the message type and input of information about the recipient. As for the delivery timing, if the message type is an advertisement about a product, it is generally considered that if the recipient receives the message while at home on a holiday, the probability of the message being opened and read is relatively high. . If the recipient actually has that tendency, the trained model selects the recipient's holiday and home timing as the transmission timing according to the input of information on the message type and recipient. can. In other words, it is possible to analyze the location information of the mobile terminal used by the recipient, and in fact, if the message type is an advertisement about a product, it may be Saturday or Sunday (which is a holiday for the recipient). Thus, if a message tends to have a high open rate when the message is sent when the mobile device used by the recipient is located at home, then the particular opportunity for such a high open rate is Selected as the most dominant choice for timing. Even if the location information of the mobile terminal used by the recipient is not in a state that can be analyzed, and from the accumulated information on the recipient, it is difficult to determine when the recipient is on holiday. Even if the estimation is not possible, the trained model can make an estimation based on the universal trained data. That is, the trained model is based on trained data on the message open rate for past message transmissions to various people, not limited to the recipient, when the message type is an advertisement about a product, and the message is effective. It is possible to output a transmission timing that has a relatively high probability of being transmitted.

As described above, the delivery channel and delivery timing that make it relatively likely that the message will be effectively delivered to the intended recipient differ depending on the recipient's situation and the "message type". There is a correlation between delivery timing and information about message type and recipients.

According to disclosed embodiments constructed on the basis of the above observations, the delivery channel and delivery timing represent a "message type", which is information related to the content of the message, and the context of the intended recipient. Depending on the "recipient information", it can be optimally selected through a trained model. Thus, for example, the likelihood that a message will be read by the recipient (and where appropriate the recipient will then take further actions expected with respect to the message) may be maximized.

On the other hand, the above-mentioned prior art is expected to be the most successful in achieving the target results among multiple options regarding communication specifications (including communication channels and transmission timing), as described in "Background Art". However, it does not specify communication channels and transmission timings that take into account the desired outcome, depending on the content sent to a particular recipient user. Therefore, in the prior art, when selecting and specifying the communication channel and transmission timing from each option, the option predicted to be most successful in achieving the target result is the content or message sent to the recipient. characteristics, such as the content characteristics of the message.

<Functional block diagram>
FIG. 1 is a functional block diagram showing the configuration of the message transmission scenario generation device 100 of the embodiment.
The device has a message receiver 104 that receives information, including messages, from an originating server 102 that originates. The message receiving unit 104 receives from the originating server 102, in addition to the message, information specifying the intended recipient, furthermore, for example, information related to the message type, information to be input to the trained model ( (including various information described below as "other input information"), as well as information indicative of a bias with respect to the transmission channel and/or transmission timing. The information identifying the intended recipient may be information indirectly identifying the recipient, such as the recipient's address, which may be the recipient's telephone number or analogous address. you can It should be noted that it is not essential for the apparatus of the embodiment that the message receiver 104 receives information identifying the intended recipient from the originating server 102 . For example, information about a group of intended recipients is maintained within the device, and the message receiver 104 receives from the originating server 102 no information identifying individual intended recipients. The device may have other aspects, such as receiving only the designation of the intended group of recipients.

The messaging scenario generator 100 of the embodiment also includes a recipient information acquisition unit 106 that acquires information about the intended recipient of the message, and identifies the message type based on information including information related to the content of the message. It has a message type identification unit 108 that
The device further receives the message in response to input of information including the message type specified by the message type specifying unit 108 and the information on the recipient acquired by the recipient information acquiring unit 106 to the trained model 114. A message transmission scenario generator 112 is provided for outputting information including transmission channels and transmission timings to be sent to people. Here, the message transmission scenario generation unit 112 also inputs information (including various information described later as “other input information”) other than the message type and recipient information to the trained model 114. It may be configured as follows. In that case, some of the information that constitutes the "other input information" may be obtained from the originating server 102 via the message receiving unit 104 (e.g., information regarding the cost of delivering the message, the originator or recipient information about the weighting of the message for the recipient), some of which may be obtained from the carrier server 120 (eg, information about where the terminal used by the recipient is located).

In the message transmission scenario generation device 100 of the illustrated embodiment, the information on the addressee is obtained by accessing the personal information database 118 storing information on individuals from the addressee information obtaining unit 106 . Access to the personal information database 118 is then performed, for example, based on information identifying the intended recipient that the message receiver 104 receives from the originating server 102 . The personal information database 118 may be included in the message transmission scenario generation device 100, or may not be included in the device and may be external to the device.
The message transmission scenario generation device 100 of the embodiment further has a message type related information storage unit 110 in which information related to message types is stored. Identify the message type by accessing. In the message type related information storage unit 110, according to the embodiment, information including a message type specifying keyword or learned data is stored as information related to the message type.

The message transmission scenario generation device 100 of the embodiment may further include a message transmission scenario recording database 116 that records information including the transmission channel and transmission timing output by the message transmission scenario generation unit 112 . By recording information including transmission channels and transmission timings in the database 116, it is possible to accumulate past output history information. Further, by performing such recording, the apparatus can refer to the content of the output in the previous stage when the transmission timing is composed of two or more stages. That is, when the information is input again to the trained model in order to update the message transmission scenario in issuing a message in the later stage, the device receives the content of the output in the previous stage (transmission channel and transmission timing). information) can be referenced, and the content can also be utilized as input.

As described above, the messaging scenario generator may cause messages to be delivered to intended recipients according to the generated messaging scenario, or may limit the generation of the messaging scenario. When the message transmission scenario generation device 100 causes a message to be transmitted, the transmission of the message addressed to the intended recipient is performed by supplying necessary information to the carrier's server 120, which is the carrier's device. A process for message delivery, such as sending, can be performed. Also, the messaging scenario generator 100 can send the generated messaging scenario to the source server 102 in the form of a reply or report.

The function of each part of the messaging scenario generator 100 of the embodiment will be further clarified by the following description in this specification.
In the following description, basically, the description up to the item of "other output information" is common to the message transmission scenario generation device, the message transmission scenario generation method, and the message transmission scenario generation program. A message transmission scenario generation device, a message transmission scenario generation method, or a message transmission scenario generation program may be collectively referred to as a “message transmission scenario generation device or the like”.

<Transmission channel>
In the disclosed embodiments, examples of communication channels are SMS, MMS, RCS, email, social media, notifications on application programs, chatbots, phone calls, mail, flyers, and visits. A message transmission scenario generation device or the like may select and output at least one of these transmission channels.
As explained for the case where the message is disaster-related evacuation information, there may be cases where the transmission timing is composed of multiple stages. One may be selected.

<Transmission timing>
In the message delivery scenario of the disclosed embodiments, the delivery timing for delivering a message or combination of messages may be configured in one or more stages. The message may be one message or a combination of messages prepared for each transmission timing configured in multiple stages.

The message transmission scenario generation device and the like of the disclosed embodiments can identify the transmission timing in various ways. Specifically, the timing of each step can be specified by, for example, the following aspects or a combination of those aspects. specific by day, specific by day of the week, specific by weekday or not, specific by time of day, specific by time of day, specific by immediate or not, specific by time range, within a predetermined time range specification by being at a random time in the above, specification related to a predetermined event, specification related to the presence or absence of behavior of the recipient, specification related to the location of the terminal used by the recipient, specification related to the location of the terminal used by the recipient, The identification related to the communication status of the terminal used and the identification related to the information about the recipient.

<Information about recipient>
In the messaging scenario of the disclosed embodiments, the intended recipient of the message can be one or more. For example, if the message is the subject of a broadcast from a business or public organization, the intended recipients may be numerous.
Accordingly, information about the intended recipient may include information identifying one or more recipients.
Information about the intended recipient includes information about various matters that may affect the intended recipient's response, particularly the relative likelihood that the message will be effectively delivered to the intended recipient. Various information can be included that can be useful input information in estimating high transmission channels and transmission timings.

Information about the intended recipient may include information about the status of responses to previous messages sent to that recipient.

Here, the information regarding the response status may include, for example, the following information or a combination thereof. information about opening a message, clicking or tapping on any part of the message, accessing web pages in response to the message, launching application programs in response to the message, and chat sessions in response to the message. Information about the initiation of a message, information about providing information in response to a message, information about registration in response to a message, information about identity verification in response to a message, information about purchasing in response to a message, information about the use of services in response to a message, Information about consent to the handling of information in response to messages, information about payment of money in response to messages, information about inquiries related to messages, information about refusal to receive messages, information about answering telephone calls, responding to mail and information on response to visits and information on the frequency of responses.
The information on the response status is at least one of information on the transmission channel and information on the transmission timing associated with the response status so that it can be known, for example, by what transmission channel and transmission timing the response status is to the transmission of the message. or

The information about the response status may include information about the response status for messages related to messages that have been sent in the past. A message related to such a message and sent in the past may be, for example, a message sent in a previous stage when transmission timing is configured in two or more stages. The response status to past messages related to the intended message may have a high correlation with the response status to the intended message and thus, as part of the information about the intended recipient, the message is intended to be. This can be useful input information in estimating delivery channels and delivery timings that are more likely to be effectively delivered to the intended recipient.

Information about the intended recipient may include information about the history of the recipient's behavior, whether or not in response to past messages sent to that recipient. Specifically, such information on the recipient's behavioral history includes at least the history of consent to information handling, the history of identity verification, the history of purchases, the history of browsing web pages, and the history of service usage. It may contain information about one. Such a recipient's behavioral history can be correlated with the status of responses to the message to be sent. For example, if the intended recipient has a history of consenting to the company that originated the message regarding the handling of information related to a certain transmission channel, the transmission of the message from the originator through that transmission channel , the recipient may exhibit a high open rate. Thus, a history of consent to information processing is part of the information about the intended recipients, as well as the delivery channels and timings that make it relatively likely that the message will be effectively delivered to the intended recipients. can be useful input information for estimating The same applies to other specific examples.

Information about the intended recipient may include information about attributes of the recipient. Specifically, the information on the recipient's attributes includes the owned terminal, transmission channels available on the owned terminal, applications used, date of birth, address, place of work, working hours, hobbies, preferences, family structure, contract It may include information about at least one of a situation, a financial situation associated with the message, and a behavioral history feature. For example, what the owned terminal is may be correlated with the transmission channels that tend to be frequently used by that terminal. Information such as the delivery channels available on the device, the applications in use, or even the generation as determined by the recipient's date of birth also make it more likely that the message will be effectively delivered to the intended recipient. It can be useful input information for channel estimation. The address or place of work, combined with, for example, the location of the mobile device used by the recipient, can be used to estimate delivery timing when the message is more likely to be effectively delivered to the intended recipient. can be useful input information in

Information about the intended recipient may include anonymized or partially anonymized information. Such anonymized or partially anonymized information may consist of anonymized, partially anonymized, or non-anonymized information provided by the source. obtain. Anonymized or partially anonymized information can be devoid of information that can lead to an individual's identification or refinement, such as name or date of birth, address, and place of work, for example.
Information about the intended recipient may also include encrypted or partially encrypted information. For example, information about the intended recipient provided by the source may be cryptographically processed by cryptographic techniques before or after it is provided. As a result, the message transmission scenario generation device or the like can use the information in an encrypted state. Cryptographic processing is applied not only to the information about the intended recipient, but also to all information used in the message transmission scenario generator, etc., including various data used in trained models (e.g. teacher data). can be made
In this way, by handling the above information in an anonymized and/or encrypted state, it becomes possible to effectively use the above information while securely managing a large amount of personal information. . By doing so, the functions of the embodiment can be further enhanced.

In disclosed embodiments, information about the intended recipient may be obtained by accessing a personal information database that stores information about individuals. In the message transfer scenario generation device, the information about the intended recipient can be obtained by accessing the personal information database from the recipient information acquisition unit. When accessing the personal information database in the message transmission scenario generation device, the message receiving unit first receives information identifying the intended recipient from the sender, and access to the personal information database is based on the information identifying the recipient. It may be executed based on

<Trained model>
In the disclosed embodiment, in response to information input to the trained model, including information about message type and intended recipient, output of information including transmission channel and transmission timing for sending the message to the recipient. done.
A neural network, for example, can be used for the trained model.
Regarding learning, a trained model can be constructed using known machine learning techniques in the field of artificial intelligence.

For example, a supervised learning method may be employed as the machine learning method.
In one embodiment, the trained model, when input with information including information about the message type and intended recipient, outputs information including the delivery channel and delivery timing for sending the message to that recipient. is subjected to machine learning processing using training data.
Here, the training data includes at least one of information regarding the recipient and information regarding persons other than the recipient. That is, the training data may or may not contain information about the recipient.
The trained model is trained based on at least one of information on response statuses to past message transmissions to the recipient and/or information on response statuses to past message transmissions to persons other than the recipient. good.

In the case of adopting a supervised learning technique, teacher data can be prepared, for example, as follows. That is, when a message of a certain message type is sent to a certain recipient with a certain transmission channel and transmission timing, the target response situation for the message, for example, the message is opened and read by the recipient , or, on top of that, if the response situation is such that the recipient has taken further actions expected with respect to the message, treat it as a correct answer, and the message type and information about the recipient, and , the transmission channel, and the transmission timing can be used as (at least part of) teacher data for supervised learning.

The machine learning method is not limited to the supervised learning method, and for example, a reinforcement learning method may be employed. In that case, an evaluation value may be automatically given to the response situation to the sending of the message according to a predetermined criterion, and reinforcement learning may be performed in which the evaluation value is used as a reward. The evaluation values are, for example, "0" for unopened messages, "1" for opened messages, "2" for accesses to web pages that respond to messages, and Alternatively, a value such as "5" for use of the service and "10" for registration of recipients who responded to the message may be automatically assigned.

<Identification of message type>
In the disclosed embodiment, the message transmission scenario generation device or the like identifies the message type based on information including information related to the content of the message.

Here, the message type may be determined based at least in part on information received from the originator related to the message type.
To this end, in the message transmission scenario generation device, the message receiving unit receives information related to the message type from the originator, and the message type identifying unit determines the message based at least in part on the information related to the message type. The type can be specified.

For example, for message types, options may be systematically prepared in advance in the form of classifications. In that case, the classification may have a hierarchical structure. As an example, in the classification of message types, under the major category of "disaster-related", the middle category of "flood-related" is placed in parallel with "earthquake-related", etc., and under "flood-related" A small classification of "several days before expected occurrence" may be provided in parallel with "immediately after W1 warning issuance".
There is no need to preliminarily prepare options in the form of classification for the message type. In that case, the sender may be allowed to freely specify the message type.

The message transmission scenario generation device or the like provides the sender with message type options (e.g., a hierarchical drop-down menu as a man-machine interface) so that the sender can enter information related to the message type. in the form of

According to one embodiment, the message type is identified by determining whether or not the extracted character string extracted from the message includes a prepared message type identifying keyword (which may include a combination of keywords). may be made. For example, a message may contain keywords such as "evacuation", "residents", "flood", "rain", "forecast", and "notice". , it can be determined that the message falls under the categories of "disaster-related" and "flood-related".
In the message transmission scenario generation device, the device has a message type related information storage section storing information related to the message type, and the message type identification section accesses the message type related information storage section. can be configured to identify the message type. In this case, the information related to the message type includes a keyword for specifying the message type, and the message type specifying unit determines whether or not the extracted character string extracted from the message includes the keyword for specifying the message type. identify.

According to another embodiment, the message type may be identified using machine learning techniques and based on information including the message and learned data.
For this purpose, the message transmission scenario generation device has a message type related information storage unit storing information related to message types, and the message type identification unit stores information related to the message types. The message type can be identified by accessing the information storage unit. In this case, the information related to the message type includes learned data, and the message type identification unit identifies the message type based on the information including the message and learned data.

As a specific method for specifying the message type, for example, the Naive Bayes method can be used. Alternatively, by subjecting a neural network to machine learning processing using teacher data that assigns message types to a large number of messages, learned data is configured, and information including input messages and the learned data is processed. Based on this, the message type may be specified.

<Adjustment by bias>
In the disclosed embodiments, the delivery channel and delivery timing for sending the message to the intended recipient is the delivery channel and delivery timing output from the trained model adjusted for the originator's preferences. You may make it output.
For this purpose, in the message transmission scenario generation device, the message reception unit receives information indicating a bias regarding at least one of the transmission channel and the transmission timing from the sender, and the message transmission scenario generation unit generates at least It can be configured to output information including the partially adjusted transmission channel and transmission timing.

Regarding delivery channels, for each specific option of delivery channel (e.g., SMS, RCS, email, social media, and notification on application program), for the output from the trained model: Bias from '+5' to '-5' via '0' to determine whether to apply a bias in the direction that the option is more selected, no bias, or bias in the direction that the option is less selected On a numerical scale, sources can be continuously selected. According to that aspect, for example, if the bias numerical scale for SMS is '0', '+4' for RCS, and '-1' for email, then the output from the trained model is biased for SMS. is not applied, while RCS is fairly strongly biased in a preferred orientation and e-mail is weakly biased in a non-selected orientation, and a transmission channel adjusted in response to such bias is output. will be

Concerning the transmission timing, each of the specific options for transmission timing (for example, whether it is a weekday or not, and whether it is a weekday or not, and whether it is specified by the time zone in the day, the time zone from 7:00 pm to 9:00 pm on weekdays, etc.) , the output from the trained model can be biased in the direction that the option is more likely to be selected, not biased, or biased in the direction that the option is less selected, from "+5" The source may be allowed to select continuously on a bias numerical scale from '0' to '-5'. According to that aspect, bias adjusted transmit timing similar to that described for the transmit channel is output.

<Specific example of message type>
There are various possible message types as described in the item "Summary of Embodiment", but for example, when preparing message types, or for specifying message types, machine learning technology is used, and specific examples of message types that can be used when machine learning processing is performed using teacher data to which message types are assigned will be further described. Message types include various types that can affect the response of the intended recipient, in particular the delivery channel and delivery timing that make it relatively likely that the message will be effectively delivered to the intended recipient. Various types can be included that can be useful input information in estimating .

Message types may include types that indicate the purpose for which the message is to be sent to the intended recipient.
Here, "the purpose of sending to the recipient" means responding to a request from the recipient, disseminating information to a group of recipients, including the intended recipient, and providing information to the recipient on a regular basis. , making a request to the payee, making a request to the payee, giving notice to the payee, and advertising to the payee.

For example, sending a message with the purpose of "responding to a request from a recipient" (e.g., a message to provide information in response to an earlier request (including inquiry) made by the intended recipient). If so, the message is the message that the recipient expects to receive. Therefore, if the message is seen by the recipient, it is very likely that the message will be opened and read by the recipient. Therefore, direct delivery to a mobile phone is possible, reception notification on a pop-up screen is also possible, and SMS or RCS, or push-type notification on an application program, which is easy for the recipient to see. (If the recipient has the application program installed), a high opening rate can be expected, and if necessary, a high probability of being redirected to a web page from the link provided in the message can be achieved. A preferred transmission channel. With regard to delivery timing, since the message is expected by the recipient, immediate delivery as soon as the message is prepared is the preferred delivery timing, where a high open rate can be expected.

Thus, according to the disclosed embodiments, the delivery channel and delivery timing can be optimally selected by making the message type as input to the trained model, as well as information about the intended recipient. As such, the likelihood that the message will be read by the recipient, or that the recipient will then take further action as expected with respect to the message, may be maximized. In some cases, such as the above example of a message with the purpose of "responding to a request from the addressee," the presence of that purpose that characterizes the message may not necessarily be clearly expressed in the content of the message itself. By using a message type indicating the purpose of "response to a request from a recipient", the characteristics of such a message are accurately reflected in the input information of the trained model. In this way, by estimating the transmission channel and transmission timing that make it relatively likely that the message will be effectively delivered to the recipient, based on the message type and information on the recipient, the accuracy of the estimation is improved. can be further improved compared to the case where the content of the message itself is used as the input for the trained model.

Here, the mode in which the message type includes a type indicating the purpose of sending the message to the intended recipient has been described in detail by taking the case of a message with the purpose of "responding to a request from the recipient" as an example; Other examples of message types given in the specification are also those types that can influence the intended recipient's response, such that the likelihood of the message being effectively delivered to the intended recipient is relatively high. It can be useful input information in estimating the transmission channel and transmission timing, which will be high.

The message type can include a type that indicates the type of intended recipient of the message.
Here, "type of recipient" refers to the customer of the sender, the potential customer of the sender, the member of a specific organization, the person who should perform the authentication procedure, the person who should give consent, and the use of a specific product or service. potential users of specific products or services, persons who receive information on a regular basis, persons who receive benefits or benefits, persons who receive demands for payment, delivery destinations, visiting work destinations, surveys and/or residents of a specific area.

For example, in the case of sending a message whose expected recipient type is "customer of originator", the history of responses to messages from the same originator in the past can be obtained from the information on the recipient. By sending the message on the transmission channel and transmission timing on which the response was made, the likelihood of a response can be particularly enhanced.

Also, for example, in the case of sending a message whose expected recipient type is "delivery destination" or "visiting work destination," the message will contain information about the service of delivery or on-site work that the recipient will receive. Since it is a message, it will be opened and read by the recipient if the message is seen by the recipient, as in the case of sending a message with the purpose of "responding to a request from the recipient" described above. Probability is high. Therefore, direct delivery to a mobile phone is possible, reception notification on a pop-up screen is also possible, and SMS or RCS, or push-type notification on an application program, which is easy for the recipient to see. (if the recipient has the application program installed) would be the preferred delivery channel. As for delivery timing, immediate delivery as soon as the message is prepared is the preferred delivery timing, where a high open rate can be expected.

Message types may include types based on the content of the message.
More specifically, the message type can include at least one of the following. i.e., a type that indicates that the message is a message that provides disaster-related notifications; a type that indicates that the message is a message that provides health-related notifications; a type indicating that the message is a notification message related to a problem or failure of a product or service; A type that indicates that the message is related to authentication procedures, a type that indicates that the message is related to authentication procedures and is a message regarding the type of authentication procedure required for the recipient, and a type that indicates that the message is a recipient. A type that indicates that the message is related to the manifestation of consent necessary for the use of the service, and that the message is a message that includes an inquiry as to whether the recipient agrees to the handling of information related to the recipient. The type indicating that the message is about information related to the recipient and for which the recipient or an entity in a position to give consent to the handling of the information related to the recipient has given consent to the handling , a type indicating that the message is a message giving notice related to the provision of benefits or benefits, a type indicating that the message is a message urging the recipient to pay money, a type indicating that the message provides a service to the recipient A type indicating that the message is a message to contact about the suspension of delivery, a type to indicate that the message is a message to contact about delivery or visiting work to the recipient, and a message that the message is a message requesting responses to surveys The type shown.

For example, an example of a case where the message is a message for "disaster-related notification" has been described in the above "Overview of Embodiments" section.

Also, for example, if the message is a message that provides a "health-related notification" or a "message related to the recipient's evaluation," the message is generally of interest to the recipient. Since the message is of high interest to the recipient, if the message is seen by the recipient, the message will be received, as in the case of sending a message with the purpose of "responding to a request from the recipient" described above. It is highly likely that it will be opened and read by a person. Therefore, direct delivery to a mobile phone is possible, reception notification on a pop-up screen is also possible, and SMS or RCS, or push-type notification on an application program, which is easy for the recipient to see. (if the recipient has the application program installed) may be the preferred delivery channel. Appropriate measures may be taken as necessary when exchanging personal information, such as directing the recipient to a website with a high level of security from the message.

However, if the message is a "health-related notice" and is a notice about a checkup with relatively complicated content, and the end of the checkup period is two months later, While the level of certainty required for message delivery is high, the level of immediacy required for message delivery is not high. In such cases, in the disclosed embodiments, based on more granular message types that reflect those characteristics of the message, or other information input to the trained model, specifically Based on the evaluation from the viewpoint of "immediacy required for message transmission" and "reliability required for message transmission" among "information on weighting of message for sender or recipient" described later , as a delivery channel, mail may be selected instead of or in addition to SMS or RCS as described above or push notifications on the application program. In that case, as the transmission timing, a time earlier than the start time of the medical examination period by more than the mailing period and related procedure period is selected.

Even if the message is a "notification related to a problem or failure in a product or service," the message is generally of interest to the recipient and is of interest to the recipient. If it is seen by the recipient, it is highly likely that it will be opened and read by the recipient. Thus, SMS or RCS, or push notification on an application program (if the recipient has the application program installed), as well as if the message is a "health related notification" case) may be the preferred transmission channel. However, since such messages are "notices related to product or service problems or failures," they may be relatively complex notifications with content that requires precision in their delivery. Then, it is possible that the level of certainty required for message transmission is high, and the level of immediacy required for message delivery is also high. In this case, in the disclosed embodiment, other information input to the trained model, specifically, "message transmission Based on the evaluation from the viewpoint of "immediacy required for communication" and "reliability required for message transmission", two stages of transmission timing were selected, and in the second stage, reliability of transmission was selected. For convenience, mail may be selected as the delivery channel. That is, in the disclosed embodiment, the first stage of immediate timing communication includes SMS or RCS, or push on application programs, for the communication of breaking news about product or service problems or failures. Notification of type is selected as the communication channel, and in the second stage of communication, mail may be selected as the communication channel for reliable communication of detailed information regarding product or service problems or failures.

Considering the case where the message is a "message related to an authentication procedure", the "authentication procedure" typically refers to the identity verification process when a person who intends to receive services using information technology registers himself/herself. occurs in In this case, the recipient has a strong motivation to quickly complete the authentication procedure, complete the service registration, and start using the service. Therefore, if the message is seen by the recipient, it is highly likely that the message will be opened and read by the recipient. Therefore, as in the case of sending a message with the purpose of "accepting a request from the recipient" described above, SMS or RCS, or push-type notification on an application program (where the recipient program) may be the preferred communication channel. As for the transmission timing, immediate issuance is a suitable transmission timing that can be expected to have a high opening rate. Since there is a possibility that a suitable transmission channel will change depending on the "form of authentication procedure required for the recipient", a message type may be provided that reflects the form of such authentication procedure.

Consider the case where the message concerns "consent". A specific example is if the message is "a message related to the recipient's declaration of consent necessary for the use of the service", the message is "a message regarding whether or not the recipient agrees to the handling of information related to the recipient". and the message is "information related to the recipient and the recipient or an entity in a position to give consent to the handling of the information related to the recipient has given consent to the handling informational message”. Such "consent"-related messages are typically messages relating to the commencement or continuation of use of a service by a person who intends to initiate use of the service or who is using the service. . In this case, the recipient has a strong motivation to complete the "consent" procedure and start or continue using the service. Therefore, if the message is seen by the recipient, it is highly likely that the message will be opened and read by the recipient. Therefore, as in the case of sending a message with the purpose of "accepting a request from the recipient" described above, SMS or RCS, or push-type notification on an application program (where the recipient program) may be the preferred communication channel. As for the transmission timing, immediate issuance is a suitable transmission timing that can be expected to have a high opening rate.

If the message is ``a message giving notice related to the provision of benefits or benefits'', if the message is ``a message urging the recipient to pay money'', and if the message is ``providing a service to the recipient Consider the case where the Such messages are of interest to the recipient, as is the case with messages making "health-related notices" or "problems or failures of products or services," as previously described. Therefore, once those messages are seen by the recipient, they are likely to be opened and read by the recipient. And such messages can be relatively complex and have content that must be accurately conveyed. Then, in the case considered here, the messages that make "health-related notices" or "product or service problem or failure-related notices" as previously described with respect to the choice of delivery channel and delivery timing. A similar situation exists when . Therefore, the message type corresponding to the case considered here is also a type that can affect the intended recipient's response, and the likelihood that the message will be effectively delivered to the intended recipient is relatively low. It can be useful input information in estimating the transmission channel and transmission timing, which are expected to be expensive.

<Other input information for the trained model>
In disclosed embodiments, the trained model may be populated with information other than information about message types and recipients. Such, so to speak, "other input information" is information that can influence the intended recipient's response, and is relative to the likelihood that the message will be effectively conveyed to the intended recipient. can be used as useful input information in estimating the transmission channel and transmission timing, which can be high at .

Specifically, the information input to the trained model may include at least one of the following. That is, all or part of the message, the number of stages of transmission timing, information on the location of the terminal used by the recipient, information on the communication status of the terminal used by the recipient, information from the platform that manages the communication of the recipient information, information in the form of information in a message including at least one of text, image, video, sound, and speech by a visitor, the manner assumed as the manner in which the recipient perceives the information in the message , recognition by reading, recognition by visual recognition, recognition by listening to one-way speech directed at the recipient, recognition by interacting through a terminal, and recognition by interacting face-to-face or by telephone information on aspects including at least one of recognition, information on the cost of conveying the message, information on how to ascertain the status of the recipient's response to the message, and information on the weighting of the message for the originator or recipient.

Here, the "information on the weighting of the message" means the immediacy required for sending the message, the urgency for the recipient to act in response to the message, the importance of the content of the message, and the certainty required for the transmission of the message. may include an evaluation in terms of at least one aspect of gender.

The above specific example will be further described.
Regarding the message, in addition to the message type, the message itself, that is, "all or part of the message" may also be input to the trained model. Thereby, it is possible to provide information that complements the message type as the input information.
Regarding the "number of stages of transmission timing", it is possible to allow the sender to specify the number of stages. The delivery timing is thereby output such that the message delivery scenario has a number of stages of delivery corresponding to the number of stages of message delivery intended by the originator.

"Information from the Recipient's Communication Management Platform" means that the intended recipient is using a communication management platform provided by the organization to which he/she belongs to conduct message communication in the course of business. This information is useful for selecting transmission channels in such cases. For example, provided that the company for which the intended recipient works provides a platform for managing communications, and that employees (including the intended recipient) use business information, , to allow sending and receiving through a particular messaging application program. In such cases, messaging targeted at the platform may be candidates for appropriate selection of delivery channels.

"Information about the form of information in the message, including at least one of text, image, video, sound, and speech by the visitor" can be useful input information in selecting a delivery channel. For example, if a message includes an animation, the message delivery scenario generator, etc. of the disclosed embodiments can include the animation in the message, rather than delivery channels such as SMS or mail, which cannot include animation in the message. Select a delivery channel, such as RCS. The form of information in the message is also useful as information that can be historical information that constitutes information about the recipient. For example, some recipients may have a tendency to respond well to messages containing animations. In that case, the message transmission scenario generation device or the like of the disclosed embodiment preferentially selects a transmission channel such as RCS that can include a moving image in a message over a transmission channel such as SMS or mail that cannot include a moving image in a message. In addition, if the message received from the caller does not include a video, the use of the video can be recommended as a "recommendation regarding the form of information in the message" to be described later.

"The manner in which recipients are expected to perceive information in messages" can also be useful input information in selecting a delivery channel. "Information in message" specifically includes information related to the message type. Examples of such aspects include an aspect including "recognition by reading with the eye", an aspect including "recognition by visual recognition", and an aspect including "recognition by listening to one-way voice directed to the recipient". , an aspect including "recognition by interacting through a terminal", and an aspect including "recognition by interacting face-to-face or over the phone". For example, in the case where "a mode assumed as a mode in which a recipient recognizes information in a message" includes "recognition by interacting through a terminal", the message transmission scenario generation device, etc. of the disclosed embodiments Increase the priority of chatbot selection as a channel. The form of information in the message is also useful as information that can be historical information that constitutes information about the recipient. This is because the ease of comprehension of the message content for the recipient may change depending on the mode of information recognition. For example, if the intended recipient has a tendency to respond favorably to messages with aspects that include "visual recognition," then for that recipient a visually appealing message can be advantageously sent by selecting, for example, RCS as the transmission channel. If the intended recipient has a tendency to respond well to messages of the type involving "awareness by interacting with a terminal", then for that recipient the communication channel may be through, for example, a chatbot. Messaging can be effective.

Regarding the "cost of transmitting a message," for example, if the cost of transmitting a message that the sender can tolerate is under strict constraints, by inputting that information into the trained model, the transmission channel will receive a corresponding shipping fee. less likely to choose mail that costs

``Information on the weighting of the message for the sender or recipient'' includes, as described above, the ``immediacy'' required for sending the message, the ``urgency'' for the recipient to act in response to the message, and the content of the message. and an evaluation in terms of at least one of the "importance" of the message and the "certainty" required for the delivery of the message. Regarding the point that "information on the weighting of the message for the originator or recipient" can be useful input information in selecting the transmission channel and transmission timing, it should be noted that when the message is a "health-related notice" and " We have already mentioned the case where it is a "notice related to a problem or failure of a product or service."

Here, in order to further explain the above-mentioned "immediacy", "urgency", "importance" and "certainty", the example used in the section "Summary of Embodiments", "Message is a disaster In the case of relevant evacuation information, when the disaster is a flood that may hit an area in a few days'. At this time, it is desirable to deliver advance notice of the necessity of evacuation or evacuation information such as evacuation sites to the residents of the target area as quickly as possible, so the level of "immediateness" required for sending messages is high. . On the other hand, regarding the "urgency" of the recipient's action in response to the message, the level of "urgency" is the most Not high, but relatively high. Regarding the "importance" of the content of the message, the level of "importance" is high because the content of the message is a matter related to human life. The level of "certainty" required for message transmission is relatively high. Such information on "immediacy", "urgency", "importance", and "certainty" is useful as input information for the trained model in selecting transmission channels and transmission timings. In this example, the delivery channel must be a delivery channel that can send the message immediately, but since the "urgency" is not the highest level, the delivery channel does not necessarily send the message directly to the mobile terminal. There may be none, and from a "certainty" point of view, a transmission channel may be selected that allows transmission of a certain amount of information. On the other hand, when a flood warning is actually issued, the level of "urgency" rises to the highest level. , a notification on an application program with the purpose of transmitting disaster prevention information can be selected as the transmission channel.

<Other output information>
In the disclosed embodiments, the messaging scenario generator or the like may output information other than the transmission channel and transmission timing. Such so-called "other output information" may include recommendations regarding the form of information in the message. Such "recommendations regarding the form of information in messages" are at least one of the following: recommendations regarding text formatting; recommendations regarding the use of images; recommendations regarding the use of video; recommendations regarding the use of sound; and recommendations regarding the use of visitor speech. can include

As mentioned above, for example, some recipients may have a tendency to exhibit good responsiveness to messages containing animations. In such a case, the message transmission scenario generation device, etc. of the disclosed embodiment may be added to the "information on the response status to the past message transmission to the recipient" included in the "information on the recipient" that is input to the trained model. Based on this, it is possible to make "recommendations regarding the form of information in the message" while changing the priority regarding the selection of the delivery channel for the recipient concerned. For example, a message transmission scenario generation device or the like preferentially selects a transmission channel such as RCS that can include a moving image in a message over a transmission channel such as SMS or postal mail that cannot include a moving image in the message. If the received message does not include video, the use of video can be recommended as a "recommendation regarding the form of information in the message." At that time, it is also possible to recommend the use of a specific moving image such as a moving image that has been used in the past.

<Overview of processing>
FIG. 2 is a flow chart outlining an example of process S200 in one embodiment of the messaging scenario generation program.
Each step of the operation flow disclosed in this specification and drawings may be performed in reverse order as long as there is no contradiction. Also, multiple steps may be performed simultaneously. Each step may be implemented by executing a program stored in memory. Also, part of each step may be implemented by an operating system or hardware.

In step S202, the message transmission scenario generation program of the embodiment causes the computer to execute a process of receiving information including a message from the originator. In addition, in the following description, the phrase "to cause a computer to execute a process" will be omitted as appropriate.
In step S204, the program of an embodiment obtains information about the intended recipient of said message.
In step S206, the message type is specified based on information including information related to the content of the message.
In step S208, information including the transmission channel and transmission timing for transmitting the message to the recipient is output in response to the input of information including the message type and information on the recipient to the trained model.
In step S210, the program of the embodiment causes the computer to execute processing for transmitting the message addressed to the recipient.
In step S212, the program of the embodiment causes the computer to execute a process of recording information including the output transmission channel and transmission timing. This process can be implemented by recording in the messaging scenario recording database 116 .
With that, the processing ends.

<Processing to receive information including messages>
FIG. 3 is a flow chart showing an example of a further embodiment (subroutine S300) of the process of receiving information including a message from the sender in step S202.
The message transmission scenario generation program of the embodiment receives a message from the originator in step S302.

In step S304, information identifying the intended recipient is received from the originator (reference number 102 in FIG. 1). The information identifying the intended recipient may be information indirectly identifying the recipient, such as the recipient's address, which may be the recipient's telephone number or a telephone number-like address. good. Note that it is not essential for the programs of the embodiments to receive information identifying the intended recipient from the originator. For example, if the computer on which the program operates contains information about a group of intended recipients, and the computer receives from the originator no information that identifies the individual intended recipients, The program may have other aspects, such as receiving only designations of groups of intended recipients.
In subsequent steps S306, S308, and S310, respectively, information related to the message type, information to be input to the trained model (including various information already described as "other input information"), and transmission channel and/or information may be received from the source indicating a bias with respect to transmission timing. In the program of the embodiment, it is not essential to receive this information from the originator.
With this, the processing of S300, which is a subroutine, ends.

<Processing to acquire information about the recipient>
FIG. 4 is a flowchart illustrating a further example embodiment (subroutine S400) of the process of obtaining information about the intended recipient in step S204 above.
In step S402, the message transmission scenario generation program of the embodiment accesses the personal information database (reference number 118 in FIG. 1) storing information about individuals. Access to the personal information database is then performed, for example, based on information identifying the intended recipient received from the originator at S304. The personal information database may be included in the computer that the program executes processing, or may be outside the computer without being included in the computer.
In step S404, information about the intended recipient is obtained from the personal information database accessed in step S402.
With this, the processing of S400, which is a subroutine, ends.

<Processing to identify the message type>
FIG. 5, which consists of FIGS. 5A and 5B, is a flow chart showing an example of a further embodiment (subroutines S500A and S500B) of the process of identifying the message type in step S206. S500A shown in FIG. 5A and S500B shown in FIG. 5B correspond to different embodiments.

In step S502, the message transmission scenario generation program of the embodiment shown in FIG. 5A extracts a character string from the message previously received from the sender in S302.
In step S504, access is made to the message type related information storage section (reference number 110 in FIG. 1) in which information related to the message type is stored. The message transmission scenario generation program of the embodiment shown in FIG. 5A uses a message type identification keyword as information related to the message type stored in the message type related information storage unit.
In the subsequent step S506, the message type is specified based on whether or not the extracted character string extracted from the message in S502 includes the prepared keyword for specifying the message type (which may include a combination of keywords). I do.
An example of specifying the message type based on such determination is as described in the item "Identification of message type" above.
Thus, the processing of S500A, which is a subroutine, ends.

The message transmission scenario generation program of the embodiment shown in FIG. 5B uses machine learning technology to identify message types based on information including messages and learned data.
In step S512, the message transmission scenario generation program of this embodiment accesses the message type related information storage section (reference number 110 in FIG. 1) in which information related to message types is stored. The message transmission scenario generation program of this embodiment uses learned data as message type related information stored in the message type related information storage unit.
In the subsequent step S514, the message type is specified based on the information including the message received from the originator in S302 and the learned data.
An example of a specific method for specifying the message type is as described in the item "Identification of message type" above.
Thus, the processing of S500B, which is a subroutine, ends.

It should be noted that if, in step S306 of subroutine S300, the messaging scenario generator receives information related to the message type from the originator, the message type can also be identified based at least in part on that information.

<Process of outputting information including transmission channel and transmission timing>
FIG. 6 is a flow chart illustrating a further example embodiment (subroutine S600) of the step S208 outputting information including the delivery channel and delivery timing for sending the message to the intended recipient.
The message transmission scenario generation program of the embodiment inputs information (various types of "other input information") other than the message type and recipient information to be input to the trained model (reference number 114 in FIG. 1). including information), the information is acquired in step S602. In that case, some of the information that constitutes these "other input information" may be obtained from the originating server (reference number 102 in FIG. 1) (e.g. information about the cost of delivering the message, the originator or recipient information about the weighting of the message for the recipient), in part may be obtained from the carrier server (reference numeral 120 in FIG. 1) (eg, information about where the terminal used by the recipient is located).

In step S604, information is input to the trained model, including information about the message type and intended recipients. The input information may include any information other than the message type and recipient information obtained in the previous step S602. The trained model is as described in the item "trained model" above.
In step S606, the delivery channel and delivery timing for sending the message to the intended recipient is output.
It should be noted that if, in step S310 of subroutine S300, the message delivery scenario generation program receives information from the originator indicating a bias regarding the delivery channel and/or delivery timing, then in step S606 the program will at least It can be configured to output information including the partially adjusted transmission channel and transmission timing. Such adjustment is specifically as described in the item "Adjustment by Bias" above.

In the message transmission scenario generation program of the embodiment, in step S608, as output information other than the transmission channel and transmission timing, a recommendation regarding the form of information in the message is output. Such "recommendations regarding the form of information in messages" are at least one of the following: recommendations regarding text formatting; recommendations regarding the use of images; recommendations regarding the use of video; recommendations regarding the use of sound; and recommendations regarding the use of visitor speech. can include Examples of recommended modes are as described in the above item "other output information".
With this, the processing of S600, which is a subroutine, ends.

<Example of message transmission scenario>
FIG. 7 shows a generated message having information including a transmission channel and transmission timing for sending the message to the recipient in the message transmission scenario generation device, message transmission scenario generation method, or message transmission scenario generation program of the disclosed embodiments. 10 is a table showing examples of transmission scenarios;
As previously mentioned, in the messaging scenarios of the disclosed embodiments, the delivery timing for delivering a message or combination of messages may be configured in one or more stages. The message may be one message or a combination of messages prepared for each transmission timing configured in multiple stages. When transmission timing consists of multiple stages, and the message for the transmission of the second and subsequent stages is not prepared at the time of generating the message transmission scenario prior to the transmission of the first stage, The message transmission scenario may be generated with the messages for the transmission of the second and subsequent stages as blanks, placeholders, or automatically generated provisional messages. In that case, the message for the transmission of the second and subsequent stages can be prepared later, and the message transmission scenario can be updated using the prepared message as input information.
Although FIG. 7 includes an example in which the transmission timing consists of a plurality of stages, the transmission timing may be limited to one stage.

FIG. 7 illustrates a messaging scenario for message 1 (M1) through message 6 (M6). Among them, for message 1, different scenarios are generated for different intended recipients R11 and R12. We use that message 1 to exemplify that although there may be many intended recipients for each message, different scenarios are typically generated for different recipients of the same message. ing.
Illustrative line numbers 1 and 2 show the messaging scenario for message 1, line number 1 being the scenario for recipient R11, and line number 2 being the scenario for recipient R12. Below, line numbers 3-7 indicate messaging scenarios for messages 2-6, respectively.

Message 1 will be explained.
The hypothetical scene for message 1 corresponds to one of the hypothetical scenes described in the "Summary of Embodiments" above. Specifically, it is assumed that the message is disaster-related evacuation information, and that the disaster is a flood that may hit an area in a few days. In this scene, at a later stage when the possibility of actual flood damage has increased (for example, the stage when the "W1 warning" is issued; the second stage), a prepared standard message that advises evacuation As the first stage of message transmission, we will notify the forecast of storms and advance notice of the necessity of evacuation, and inform residents of the relevant area in advance of evacuation sites, maps and explanations. We are planning to send a message with a message to the residents of the area.

In this example, the "message type" is, for example, as follows. As a type indicating the purpose of sending a message to a recipient, for example, "to disseminate information to a group of recipients including the recipient" can be applicable. For example, "residents of a specific area" may be applicable as a type indicating the type of expected recipient of the message. Types based on the contents of the message include, for example, "notifications related to flooding" among "notifications related to disasters", "notifications with map information", and "notifications issued several days before the expected disaster occurrence". “notice to be given” may apply.
The message transmission scenario generation device or the like of the embodiment identifies the message type based on information including information related to the content of the message. According to one embodiment, the message type is identified by determining whether the string extracted from the message contains a provided message type identification keyword. According to another embodiment, based on the messages and learned data, their message types are identified.

Next, according to the acquired "recipient information", recipient R11 among the intended recipients is a user of "application program 1" whose purpose is to transmit disaster prevention information, Has a history of frequent use of the program. On the other hand, among the intended recipients, recipient R12 does not have information that he is a user of an application program with the purpose of transmitting disaster prevention information, such as application program 1, and uses RCS on his mobile terminal. there is no record of it. However, there is a record of the e-mail address and a history of using SMS on the mobile terminal in possession.

``Other input information'' other than ``message type'' and ``recipient information'' is not essential. has been received. According to it, the level of "immediacy" required for sending a message is the highest ("5" on a scale of 5), and the level of "urgency" for the recipient to act in response to the message is the highest. It is not a standard, but a relatively high level ("4" on a 5-point scale). It should be noted that the level of "urgency" when the second level message is issued is the highest ("5" on a scale of 5).

The message transmission scenario generation device, etc. of the embodiment selects the transmission channel and transmission timing for transmitting the message to the recipient in accordance with the input of information including the "message type" and the "information on the recipient" to the trained model. Output information containing In the case of this example, the above-mentioned "information on message weighting for the originator" is also input as "other input information".

In this example, from the trained model, regarding the transmission timing, the transmission timing of the first stage "transmission 1" is "immediately", and the transmission timing of the second stage "transmission 2" is "immediately after W1 warning is issued". output is done.

Similarly, from the trained model, regarding the transmission channel, for the intended recipient R11, both transmissions 1 and 2 output "application program 1", and for the intended recipient R12, transmission 1 "e-mail" is selected and "SMS" is selected and output for transmission 2, respectively. The selection of these delivery channels is based on the "message type" and the "recipient information".

Specifically, the recipient R11 is a user of the application program 1 that matches the message type, and has a history of frequently using the program. have selected. Note that even if the use history of the application program 1 of the recipient R11 is unknown, the trained model can make an estimation based on the universal trained data. For example, if users of the program generally tend to use the program frequently, the trained model can be used to obtain past messages to various people, not limited to the recipient, when the message type is related to disasters. Based on the information that the recipient R11 is a user of the program, using the learned data on the message open rate for sending, the program is selected as a transmission channel with a relatively high possibility of effectively transmitting the message. can be selected.

In the case of recipient R12, there is no information that he is a user of an application program that matches the message type of disaster related, and there is no record of using RCS on his mobile terminal, but there is a record of his e-mail address. Based on the fact that there is also a history of using SMS on a mobile terminal with a , "e-mail" is selected for transmission 1, and "SMS" is selected for transmission 2, respectively.

In the message transmission scenario generation device and the like of the embodiments, the transmission channel and transmission timing may be selected based on "other input information" other than the message type and recipient information. In the case of this example, the selection of the transmission channel is also consistent with the evaluation of immediacy and urgency in the "information on the weighting of the message for the sender" as "other input information" ( See the explanation in the section "Other input information for the trained model".)
In the message transmission scenario generation device of the embodiment, etc., the transmission channel and/or transmission timing output in response to the input of the information on the message type and recipient is "information on the weighting of the message for the originator". If there is a mismatch between the transmission channel and/or the transmission timing that can be selected based on the "other input information", the message transmission scenario generation device etc. is configured to output a warning to call attention. good too.
Based on the selection of the learned model as described above, the message transmission scenario generation device or the like of the disclosed embodiment outputs the selected transmission channel and transmission timing.

Regarding this example, the message transmission scenario generation device or the like of the disclosed embodiments also outputs a recommendation regarding the form of information in the message as "other output information" other than the transmission channel and transmission timing. Specifically, when the message type is "notification related to flood damage" in response to past message transmission, when "image 1" is used in the message, for example, the web page that responded to the message There is a general tendency that the frequency of accesses to the site increases, and that is what the trained model has learned. Therefore, in the case of this example, according to the input of the message type, the message transmission scenario generation device or the like of the disclosed embodiment outputs a recommendation to use "image 1" for the message. The recommendations output here are communicated to the originating server, allowing the originator to immediately implement the improvement of adding "image 1" to the message.
In addition, in order to ensure the transmission of messages without delay, the message transmission scenario generation device, etc., after transmitting the recommendation to the transmission source server, if the transmission source does not respond to adopt the recommendation within a predetermined time, the original message may be configured to send out.

Message 2 of line number 3 is an example of a case where the message type has a type indicating that the message is a message for notifying a product failure, and the illustrated message transmission scenario is generated. .
Message 3 of line number 4 has a message type indicating that the message is "a message related to an authentication procedure" and "a message related to the type of authentication procedure requested to the recipient". For example.
Message 4 of line number 5 is an example of a case where the message type has a type indicating that it is a "message related to manifestation of consent necessary for use of recipient's service".
Message 5 of line number 6 is an example in which the message type has a type indicating that the message is a "message demanding payment of money to the recipient".

Next, message 6 on line number 7 will be described.
Message 6 is an advertisement for the product.
In this example, the "message type" is, for example, as follows. As a type indicating the purpose of sending the message to the recipient, for example, "to advertise to the recipient" may correspond. As a type indicating the type of expected recipient of the message, for example, the "customer of the sender" may correspond. As a type based on the contents of the message, for example, "message for advertising related to products" may correspond. The message type may include information about the product, such as the manufacturer of the product, the type of the product, the product name of the product, and the like.
As described above, the message transmission scenario generation device or the like of the embodiment identifies the message type based on information including information related to the content of the message.

Next, according to the obtained "recipient information", the intended recipient R61 is a frequent user of "social media 1" and is the manufacturer of the product through that social media. You have registered to receive messages from the original. If it is possible to analyze the location information of the mobile terminal used by the recipient R61, and the message type is an advertisement about a product, on Saturday or Sunday (which is a holiday for the recipient), If a message is sent when a person's mobile device is located at home, there is a tendency for the message open rate to be high.

As the "other input information" other than the "message type" and "information on the recipient", the location information of the mobile terminal used by the recipient R61 is obtained from the communication carrier server.

As described above, the message transmission scenario generation device, etc. of the embodiment transmits a message to a recipient according to input of information including the "message type" and the "information on the recipient" for the trained model. Output information including channel and transmission timing.
In this example, from the trained model, "social media 1" is selected as the transmission channel, and "when the terminal used by the recipient is located at home on Saturday or Sunday" as the transmission timing. output.

In this example, from the acquired "recipient information", if the message type is an advertisement about a product, it is Saturday or Sunday (which is a holiday for the recipient) and the mobile terminal used by the recipient R61 It has been known that there is a tendency for higher message open rates when messages are sent when is located at home. However, even if the location information of the mobile terminal used by the recipient is not in a state that can be analyzed, and from the accumulated information about the recipient, when is the recipient's holiday? Even if it is not possible to estimate , the trained model can make an estimation based on universal trained data. That is, the trained model is based on trained data on the message open rate for past message transmissions to various people, not limited to the recipient, when the message type is an advertisement about a product. It is possible to output a transmission timing that has a relatively high probability of being transmitted.

Based on the selection of the learned model as described above, the message transmission scenario generation device or the like of the disclosed embodiment outputs the selected transmission channel and transmission timing.
Regarding this example, the message transmission scenario generation device or the like of the disclosed embodiments also outputs a recommendation regarding the form of information in the message as "other output information" other than the transmission channel and transmission timing. Specifically, when the message type is "product-related advertisement message" and the message is a message from the manufacturer of the product, in the response status to the past message transmission, the message "Video 1 is used, there is a general trend that the frequency of visits to web pages in response to messages and purchases in response to messages increases, which the trained model has learned. Therefore, in the case of this example, the message transmission scenario generation device or the like of the disclosed embodiment outputs a recommendation to present a link to "video 1" in the message. The recommendations output here are communicated to the originating server, allowing the originator to immediately implement an improvement that adds a link to "Movie 1" to the message.
In addition, in order to ensure that the message is sent without delay, as described above, the message transmission scenario generation device, etc. does not make a response that the transmission source adopts the recommendation within a predetermined time after transmitting the recommendation to the transmission source server. In this case, it may have a configuration in which the original message is sent.

<Hardware configuration>
FIG. 8 is a hardware configuration diagram of the message transmission scenario generation device 800 of the embodiment, or the device 800 that executes the message transmission scenario generation method or the message transmission scenario generation program. The device 800 has a CPU 802 , a ROM 804 , a RAM 806 , a key operation section 808 , a communication section 810 , a display section 812 and an external storage control section 814 . The device 800 can operate by reading the message transmission scenario generation program stored in the memory 816 by the external storage control unit 814 . A messaging scenario generation program may also be stored in ROM 804 and RAM 806 . That is, programs implementing embodiments may be stored in non-transitory, tangible, and machine-readable storage media such as memory 816, ROM 804, and/or RAM 806. be done. The device 800 operates under the control of a CPU 802 that executes a message transmission scenario generation program.

Although the above has been described based on the embodiments, the embodiments described in this specification are only examples, and should not be construed as limiting the invention specified in the claims. not.

The following notes are disclosed with respect to the above-described embodiments.

(Appendix 1)
a message receiver that receives information, including a message, from an originator;
a recipient information obtaining unit for obtaining information about intended recipients of said message;
a message type identification unit that identifies a message type based on information including information related to the content of the message;
a message transmission scenario generation unit that outputs information including a transmission channel and transmission timing for transmitting the message to the recipient in response to input of information including information on the message type and the recipient to the trained model;
A messaging scenario generator comprising:

(Appendix 2)
The transmission channel, at each transmission timing,
At least one of SMS, MMS, RCS, e-mail, social media, notifications on application programs, chatbots, phone calls, postal mail, flyer distribution, and visits are selected The messaging scenario of Appendix 1 generator.

(Appendix 3)
The transmission timing is composed of one or more stages, and the timing of each stage is
Specific by day, specific by day of the week, specific by weekday or not, specific by time of day, specific by time of day, specific by time or not, specific by time range, random within a predetermined time range identification by being at a certain time, identification related to a predetermined event, identification related to the presence or absence of an action of the recipient, identification related to the location of the terminal used by the recipient, identification used by the recipient 3. The messaging scenario generation device according to appendix 1 or 2, comprising at least one of a specification related to communication status of a terminal and a specification related to information about said recipient.

(Appendix 4)
The information about said recipient is
4. A messaging scenario generator according to any one of appendices 1 to 3, including information identifying said at least one recipient.

(Appendix 5)
The information about said recipient is
5. The message delivery scenario generation device according to any one of Appendices 1 to 4, including information on response statuses to past messages sent to the recipient.

(Appendix 6)
The information about the response status is
Information about opening a message, clicking or tapping on a part within a message, accessing a web page in response to a message, launching an application program in response to a message, starting a chat session in response to a message information about providing information in response to messages; information about registration in response to messages; information about identity verification in response to messages; information about purchasing in response to messages; Information about consent to the handling of information in response, information about payment of money in response to messages, information about inquiries related to messages, information about refusal to receive messages, information about answering telephone calls, information about responding to mail , and at least one of information on response to visits and information on frequency of responses.

(Appendix 7)
The information about said recipient is
7. The messaging scenario generation device according to any one of Appendices 1 to 6, including information relating to a history of behavior of said recipient, whether or not it is a response to past messages sent to said recipient.

(Appendix 8)
The information about said recipient is
8. A messaging scenario generator according to any one of appendices 1 to 7, including information about attributes of said recipient.

(Appendix 9)
The information about said recipient is
Generating a messaging scenario according to any one of Appendices 1 to 8, including anonymized or partially anonymized information and/or encrypted or partially encrypted information. Device.

(Appendix 10)
The trained model is
Machine learning using teacher data so as to output information including the transmission channel and the transmission timing for sending the message to the recipient when information including the message type and information on the recipient is input. processed,
10. The message transmission scenario generation device according to any one of Appendices 1 to 9.

(Appendix 11)
The teacher data is
11. The message delivery scenario generation device according to appendix 10, including at least one of information about the recipient and information about a person other than the recipient.

(Appendix 12)
The trained model is
Learning is performed based on at least one of information regarding response statuses to past message transmissions to said recipient and information regarding response statuses to past message transmissions to persons other than said recipient. The message transmission scenario generation device according to any one of the above.

(Appendix 13)
The message receiving unit
further receiving information related to the message type from the originator;
The message type identification unit is
13. The messaging scenario generator of any one of clauses 1-12, wherein the message type is determined based at least in part on information associated with the message type.

(Appendix 14)
further comprising a message type related information storage unit storing information related to the message type;
The message type identification unit is
14. The message transmission scenario generation device according to any one of appendices 1 to 13, wherein the message type is specified by accessing the message type related information storage unit.

(Appendix 15)
Information related to the message type includes:
including a message type identification keyword,
The message type identification unit is
15. The message transmission scenario generation device according to appendix 14, wherein the message type is specified by determining whether or not the extracted character string extracted from the message includes the message type specifying keyword.

(Appendix 16)
Information related to the message type includes:
contains trained data,
The message type identification unit is
15. The message transmission scenario generation device according to appendix 14, wherein the message type is specified based on information including the message and the learned data.

(Appendix 17)
The message receiving unit
further receiving information from the source indicating a bias with respect to at least one of the transmission channel and the transmission timing;
The message transmission scenario generation unit
17. The messaging scenario generator of any one of claims 1 to 16, outputting information including the transmission channel and the transmission timing at least partially adjusted according to the bias.

(Appendix 18)
The message type is
18. The message delivery scenario generation device according to any one of Appendices 1 to 17, including a type indicating the purpose of sending the message to the recipient.

(Appendix 19)
The message type is
19. The message transmission scenario generation device according to any one of Appendices 1 to 18, including a type indicating a type of expected recipient of the message.

(Appendix 20)
The message type is
20. The message transmission scenario generation device according to any one of Appendices 1 to 19, including a type based on the content of the message.

(Appendix 21)
The message type is
a type indicating that the message is a message providing disaster-related notification; a type indicating that the message is a message providing health-related notification; a type indicating that the message is a notification message related to problems or failures of products or services; a type indicating that the message is a message that provides information or advertisements related to products or services , a type indicating that the message is a message related to an authentication procedure, a type indicating that the message is a message related to an authentication procedure and is a message related to a form of authentication procedure requested of the recipient; A type indicating that the message is a message related to the consent required for the use of the service of the recipient, and whether or not the recipient consents to the handling of information related to the recipient in the message. The message is information related to the recipient, and the recipient or a subject who is in a position to give consent to the handling of information related to the recipient consents to the handling a type indicating that the message is a message relating to information giving a A type indicating that the message is a message, a type indicating that the message is a message for contacting the recipient regarding suspension of service provision, and a message for contacting the recipient regarding delivery or on-site work. and at least one of a type indicating that the message is a message requesting an answer to a survey.

(Appendix 22)
The information input to the trained model is
Manage all or part of the message, the number of stages of the transmission timing, information about the location of the terminal used by the recipient, information about the communication status of the terminal used by the recipient, and communication of the recipient information from the platform, information in the form of information in said message including at least one of text, images, video, sound, and speech by a visitor, said recipient perceiving information in said message. Recognition by visual reading, recognition by visual recognition, recognition by listening to one-way voice directed to the recipient, recognition by interacting through a terminal, and face-to-face or information on aspects including at least one of recognition by telephone interaction, information on costs of conveying said message, information on how to check the status of said recipient's response to said message, and said originator or said recipient. 22. The messaging scenario generator according to any one of the appendices 1 to 21, further comprising at least one piece of information regarding the weighting of the messages for.

(Appendix 23)
The message weighting information includes:
from at least one of the immediacy required for sending the message, the urgency of the recipient acting in response to the message, the importance of the content of the message, and the certainty required for delivery of the message 23. The messaging scenario generation device according to appendix 22.

(Appendix 24)
The information output by the message transmission scenario generation unit includes:
24. A messaging scenario generator according to any one of clauses 1 to 23, further comprising recommendations regarding the form of information in said message.

(Appendix 25)
A computer-executed method for generating a message-passing scenario, comprising:
receiving information, including a message, from an originator;
obtaining information about intended recipients of said message;
identifying a message type based on information including information related to content of the message;
and outputting information including a transmission channel and transmission timing for transmitting the message to the recipient in response to input of information including information on the message type and the recipient to a trained model. Method.

(Appendix 26)
receive information, including messages, from originators;
obtaining information about the intended recipient of said message;
identifying a message type based on information including information related to the content of the message;
Outputting information including a transmission channel and transmission timing for transmitting the message to the recipient according to input of information including information on the message type and the recipient to a trained model Message transmission for causing a computer to execute processing Scenario generator.

(Appendix 27)
Said message is
25. The message transmission scenario generation device according to any one of appendices 1 to 24, wherein the message is one message or a combination of messages prepared for each transmission timing composed of a plurality of stages.

(Appendix 28)
The information about the response status is
6. The message delivery scenario generation device according to appendix 5, including at least one of information on a delivery channel and information on delivery timing associated with the response status.

(Appendix 29)
The information about the response status is
6. The message transmission scenario generation device according to appendix 5, wherein the message is related to the message and contains information about response status to a message sent in the past.

(Appendix 30)
Messages related to said message and sent in the past are:
29. The message transfer scenario generation device according to appendix 29, wherein the message is a message sent in a previous step when the transfer timing is composed of two or more steps.

(Appendix 31)
Information about the history of behavior of said recipient,
The message transmission scenario generation device according to appendix 7, including information on at least one of a history of consent to information handling, a history of personal authentication, a history of purchase, a history of browsing web pages, and a history of service usage.

(Appendix 32)
The information about the attribute of the recipient is
Possessed terminal, transmission channels available on the possessed terminal, applications used, date of birth, address, place of work, working hours, hobbies, preferences, family structure, contract status, financial status related to the message, and 9. The messaging scenario generation device according to appendix 8, including information about at least one feature related to behavioral history.

(Appendix 33)
The information about said recipient is
The message transmission scenario generation device according to any one of Appendices 1 to 24 and Appendices 27 to 32, which is obtained by accessing a personal information database storing information about an individual from the addressee information obtaining unit.

(Appendix 34)
The message receiving unit
receiving further information from the originator identifying an intended recipient of the message;
Access to the personal information database includes:
34. The messaging scenario generator according to appendix 33, executed based on information identifying said recipient.

(Appendix 35)
The purpose of sending to said recipient is
responding to requests from said recipients; disseminating information to a group of recipients including said recipients; providing information to said recipients periodically; making requests to said recipients; 19. The messaging scenario generator of clause 18, comprising at least one of making a request to a person, giving notice to the recipient, and advertising to the recipient.

(Appendix 36)
The types of recipients are:
Customer of said originator, potential customer of said originator, member of a specific organization, person to undergo certification procedures, person to give consent, user of a particular product or service, potential customer of a particular product or service users, persons who regularly receive information, persons who receive benefits or benefits, persons who receive demands for payment of money, delivery destinations, visiting work destinations, persons who receive requests to respond to surveys, and 20. The messaging scenario generator according to appendix 19, including at least one of the residents of a specific area.

(Appendix 37)
Recommendations regarding the form of said information are:
25. The messaging scenario generator of Claim 24, comprising at least one of text formatting recommendations, image usage recommendations, video usage recommendations, sound usage recommendations, and visitor speech recommendations.

(Appendix 38)
The transmission channel, at each transmission timing,
At least one of SMS, MMS, RCS, email, social media, notification on application program, chatbot, phone, mail, leaflet distribution, and visitation is selected The messaging scenario of Appendix 25 generation method.

(Appendix 39)
The transmission timing is composed of one or more stages, and the timing of each stage is
Specific by day, specific by day of the week, specific by weekday or not, specific by time of day, specific by time of day, specific by time or not, specific by time range, random within a predetermined time range identification by being at a certain time, identification related to a predetermined event, identification related to the presence or absence of an action of the recipient, identification related to the location of the terminal used by the recipient, identification used by the recipient 39. The method of claim 25 or 38, comprising at least one of a specification related to a communication state of a terminal and a specification related to information about said recipient.

(Appendix 40)
The information about said recipient is
40. A method of creating a messaging scenario according to any one of clauses 25 or 38-39, including information identifying at least one said recipient.

(Appendix 41)
The information about said recipient is
41. The method of creating a messaging scenario according to any one of Appendixes 25 or 38-40, including information about response statuses to past messages sent to the recipient.

(Appendix 42)
The information about the response status is
Information about opening a message, clicking or tapping on a part within a message, accessing a web page in response to a message, launching an application program in response to a message, starting a chat session in response to a message information about providing information in response to messages; information about registration in response to messages; information about identity verification in response to messages; information about purchasing in response to messages; Information about consent to the handling of information in response, information about payment of money in response to messages, information about inquiries related to messages, information about refusal to receive messages, information about answering telephone calls, information about responding to mail , and information on response to visits, and information on frequency of responses.

(Appendix 43)
The information about said recipient is
43. The messaging scenario of any one of Clauses 25 or 38-42, including information about a history of the recipient's behavior, whether or not in response to past messages sent to the recipient. generation method.

(Appendix 44)
The information about said recipient is
44. A method of creating a messaging scenario according to any one of clauses 25 or 38-43, including information about attributes of said recipient.

(Appendix 45)
The information about said recipient is
45. The message of any one of Clauses 25 or 38-44, comprising anonymized or partially anonymized information and/or encrypted or partially encrypted information. Propagation scenario generation method.

(Appendix 46)
The trained model is
Machine learning using teacher data so as to output information including the transmission channel and the transmission timing for sending the message to the recipient when information including the message type and information on the recipient is input. processed,
46. The method of generating a message delivery scenario according to any one of appendices 25 or 38-45.

(Appendix 47)
The teacher data is
47. The method of claim 46, including information about the recipient and/or information about a person other than the recipient.

(Appendix 48)
The trained model is
Learning is performed based on at least one of information regarding response statuses to past message transmissions to the recipient and/or information regarding response statuses to past message transmissions to persons other than the recipient. 47. The method of generating a message delivery scenario according to any one of claims 47 to 47.

(Appendix 49)
Receiving information containing said message includes:
receiving further information related to the message type from the originator;
Identifying the message type includes:
49. The method of clause 25 or any one of clauses 38-48, comprising identifying the message type based at least in part on information associated with the message type.

(Appendix 50)
Identifying the message type includes:
50. The method of claim 25 or any one of clauses 38-49, comprising identifying the message type by accessing stored information associated with the message type.

(Appendix 51)
Information related to the message type includes:
including a message type identification keyword,
Identifying the message type includes:
51. The method of generating a message transmission scenario according to appendix 50, further comprising determining whether or not the extracted character string extracted from the message includes the message type specifying keyword to specify the message type.

(Appendix 52)
Information related to the message type includes:
contains trained data,
Identifying the message type includes:
51. The message delivery scenario generation method according to appendix 50, including specifying the message type based on information including the message and the learned data.

(Appendix 53)
Receiving information containing said message includes:
further comprising receiving information from the source indicating a bias with respect to at least one of the transmission channel and the transmission timing;
Outputting information including the transmission channel and transmission timing,
53. The method of claim 25 or any one of clauses 38-52, comprising outputting information including the transmission channel and the transmission timing at least partially adjusted in response to the bias.

(Appendix 54)
The message type is
54. A method of creating a messaging scenario according to any one of clauses 25 or 38-53, including a type indicating the purpose for which said message is to be sent to said recipient.

(Appendix 55)
The message type is
55. The method of generating a message delivery scenario according to any one of attachments 25 and 38 to 54, including a type indicating a type of expected recipient of the message.

(Appendix 56)
The message type is
56. The method of creating a message delivery scenario according to any one of Appendix 25 or Appendixes 38-55, including a type based on content of the message.

(Appendix 57)
The message type is
a type indicating that the message is a message providing disaster-related notification; a type indicating that the message is a message providing health-related notification; a type indicating that the message is a notification message related to problems or failures of products or services; a type indicating that the message is a message that provides information or advertisements related to products or services , a type indicating that the message is a message related to an authentication procedure, a type indicating that the message is a message related to an authentication procedure and is a message related to a form of authentication procedure requested of the recipient; A type indicating that the message is a message related to the consent required for the use of the service of the recipient, and whether or not the recipient consents to the handling of information related to the recipient in the message. The message is information related to the recipient, and the recipient or a subject who is in a position to give consent to the handling of information related to the recipient consents to the handling a type indicating that the message is a message relating to information giving a A type indicating that the message is a message, a type indicating that the message is a message for contacting the recipient regarding suspension of service provision, and a message for contacting the recipient regarding delivery or on-site work. and at least one of a type indicating that the message is a message requesting an answer to a survey.

(Appendix 58)
The information input to the trained model is
Manage all or part of the message, the number of stages of the transmission timing, information about the location of the terminal used by the recipient, information about the communication status of the terminal used by the recipient, and communication of the recipient information from the platform, information in the form of information in said message including at least one of text, images, video, sound, and speech by a visitor, said recipient perceiving information in said message. Recognition by visual reading, recognition by visual recognition, recognition by listening to one-way voice directed to the recipient, recognition by interacting through a terminal, and face-to-face or information on aspects including at least one of recognition by telephone interaction, information on costs of conveying said message, information on how to check the status of said recipient's response to said message, and said originator or said recipient. 58. The method of generating a messaging scenario according to any one of clauses 25 or 38-57, further comprising at least one of information relating to the weighting of the message for.

(Appendix 59)
The message weighting information includes:
from at least one of the immediacy required for sending the message, the urgency of the recipient acting in response to the message, the importance of the content of the message, and the certainty required for delivery of the message 59. The message passing scenario generation method of clause 58.

(Appendix 60)
In outputting information including the transmission channel and transmission timing, the output information is
60. A method of creating a messaging scenario according to any one of clauses 25 or 38-59, further comprising recommendations regarding the form of information in said message.

(Appendix 61)
Said message is
61. The message transmission scenario generation method according to any one of attachments 25 or 38 to 60, wherein the message is one message or a combination of messages prepared for each transmission timing configured in a plurality of stages.

(Appendix 62)
The information about the response status is
42. The method of generating a message delivery scenario according to claim 41, including at least one of information on delivery channel and information on delivery timing associated with response status.

(Appendix 63)
The information about the response status is
42. The method of generating a message delivery scenario according to appendix 41, wherein the message is related to the message and includes information about the status of responses to messages sent in the past.

(Appendix 64)
Messages related to said message and sent in the past are:
64. The message delivery scenario generation method according to appendix 63, wherein the message is a message sent in a previous stage when the delivery timing is composed of two or more stages.

(Appendix 65)
Information about the history of behavior of said recipient,
44. The message transmission scenario generation method according to appendix 43, including information on at least one of a history of consent to information handling, a history of personal authentication, a history of purchase, a history of browsing web pages, and a history of service usage.

(Appendix 66)
The information about the attribute of the recipient is
Possessed terminal, transmission channels available on the possessed terminal, applications used, date of birth, address, place of work, working hours, hobbies, preferences, family structure, contract status, financial status related to the message, and 45. The method of claim 44, including information about at least one historical behavioral characteristic.

(Appendix 67)
Obtaining information about the recipient includes:
67. The method of generating a message delivery scenario according to any one of Appendixes 25 or 38-66, comprising obtaining the information about the recipient by accessing a personal information database storing information about the individual.

(Appendix 68)
Receiving information containing said message includes:
receiving information from the originator further identifying an intended recipient of the message;
Access to the personal information database includes:
68. A method of generating a messaging scenario according to clause 67, performed based on information identifying said recipient.

(Appendix 69)
The purpose of sending to said recipient is
responding to requests from said recipients; disseminating information to a group of recipients including said recipients; providing information to said recipients periodically; making requests to said recipients; 55. The method of claim 54, comprising at least one of: making a request to a person, warning the recipient, and advertising to the recipient.

(Appendix 70)
The types of recipients are:
Customer of said originator, potential customer of said originator, member of a specific organization, person to undergo certification procedures, person to give consent, user of a particular product or service, potential customer of a particular product or service users, persons who regularly receive information, persons who receive benefits or benefits, persons who receive demands for payment of money, delivery destinations, visiting work destinations, persons who receive requests to respond to surveys, and 56. The method of creating a messaging scenario according to Clause 55, including at least one inhabitant of a particular region.

(Appendix 71)
Recommendations regarding the form of said information are:
61. The method of claim 60, comprising at least one of text formatting recommendations, image usage recommendations, video usage recommendations, sound usage recommendations, and visitor speech usage recommendations.

(Appendix 72)
The transmission channel, at each transmission timing,
At least one of SMS, MMS, RCS, e-mail, social media, notification on application program, chatbot, phone, mail, leaflet distribution, and visit is selected The messaging scenario of Appendix 26 generation program.

(Appendix 73)
The transmission timing is composed of one or more stages, and the timing of each stage is
Specific by day, specific by day of the week, specific by weekday or not, specific by time of day, specific by time of day, specific by time or not, specific by time range, random within a predetermined time range identification by being at a certain time, identification related to a predetermined event, identification related to the presence or absence of an action of the recipient, identification related to the location of the terminal used by the recipient, identification used by the recipient 73. The messaging scenario generating program according to clause 26 or 72, comprising at least one of a specification related to communication status of a terminal and a specification related to information about said recipient.

(Appendix 74)
The information about said recipient is
74. The messaging scenario generator of any one of clauses 26 or 72-73, including information identifying at least one said recipient.

(Appendix 75)
The information about said recipient is
75. A messaging scenario generator according to any one of Appendixes 26 or 72-74, including information regarding response status to past messages sent to the recipient.

(Appendix 76)
The information about the response status is
Information about opening a message, clicking or tapping on a part within a message, accessing a web page in response to a message, launching an application program in response to a message, starting a chat session in response to a message information about providing information in response to messages; information about registration in response to messages; information about identity verification in response to messages; information about purchasing in response to messages; Information about consent to the handling of information in response, information about payment of money in response to messages, information about inquiries related to messages, information about refusal to receive messages, information about answering telephone calls, information about responding to mail , and information about response to visits, and information about frequency of responses.

(Appendix 77)
The information about said recipient is
77. The messaging scenario of any one of Clauses 26 or 72-76, including information about a history of behavior of said recipient, whether or not in response to past messages sent to said recipient. generation program.

(Appendix 78)
The information about said recipient is
78. A messaging scenario generator according to any one of clauses 26 or 72-77, including information about attributes of said recipient.

(Appendix 79)
The information about said recipient is
79. The message of any one of Clauses 26 or 72-78, comprising anonymized or partially anonymized information and/or encrypted or partially encrypted information. Transmission scenario generator.

(Appendix 80)
The trained model is
Machine learning using teacher data so as to output information including the transmission channel and the transmission timing for sending the message to the recipient when information including the message type and information on the recipient is input. processed,
80. The message transfer scenario generation program according to any one of appendices 26 or 72-79.

(Appendix 81)
The teacher data is
81. The message delivery scenario generation program according to appendix 80, including at least one of information about the recipient and information about a person other than the recipient.

(Appendix 82)
The trained model is
Learning is performed based on at least one of information regarding response statuses to past message transmissions to the recipient and information regarding response statuses to past message transmissions to persons other than the recipient. Appendix 26 or Appendix 72 81. The messaging scenario generating program according to any one of 81.

(Appendix 83)
A process for receiving information comprising said message comprises:
further comprising receiving information related to the message type from the originator;
The process of identifying the message type includes:
83. The messaging scenario generation program according to any one of Appendixes 26 or 72-82, comprising identifying the message type based at least in part on information associated with the message type.

(Appendix 84)
The process of identifying the message type includes:
84. The message delivery scenario generation program according to any one of Appendix 26 or Appendixes 72-83, comprising a process of identifying the message type by accessing stored information related to the message type.

(Appendix 85)
Information related to the message type includes:
including a message type identification keyword,
The process of identifying the message type includes:
85. The message transmission scenario generation program according to appendix 84, including a process of determining whether or not the extracted character string extracted from the message includes the message type specifying keyword to specify the message type.

(Appendix 86)
Information related to the message type includes:
contains trained data,
The process of identifying the message type includes:
85. The message transmission scenario generation program according to appendix 84, including a process of identifying the message type based on information including the message and the learned data.

(Appendix 87)
A process for receiving information comprising said message comprises:
further comprising receiving information from the source indicating a bias with respect to at least one of the transmission channel and the transmission timing;
The process of outputting information including the transmission channel and transmission timing includes:
87. The message delivery scenario generation program according to any one of Clauses 26 or 72-86, comprising outputting information including the transmission channel and the transmission timing at least partially adjusted in response to the bias.

(Appendix 88)
The message type is
88. The messaging scenario generator of any one of Appendixes 26 or 72-87, including a type indicating the purpose of sending the message to the recipient.

(Appendix 89)
The message type is
89. The message transmission scenario generation program according to any one of Appendix 26 or Appendixes 72 to 88, including a type indicating a type of expected recipient of the message.

(Appendix 90)
The message type is
90. The message delivery scenario generation program according to any one of Appendix 26 or Appendixes 72-89, including a type based on content of the message.

(Appendix 91)
The message type is
a type indicating that the message is a message providing disaster-related notification; a type indicating that the message is a message providing health-related notification; a type indicating that the message is a notification message related to problems or failures of products or services; a type indicating that the message is a message that provides information or advertisements related to products or services , a type indicating that the message is a message related to an authentication procedure, a type indicating that the message is a message related to an authentication procedure and is a message related to a form of authentication procedure requested of the recipient; A type indicating that the message is a message related to the consent required for the use of the service of the recipient, and whether or not the recipient consents to the handling of information related to the recipient in the message. The message is information related to the recipient, and the recipient or a subject who is in a position to give consent to the handling of information related to the recipient consents to the handling a type indicating that the message is a message relating to information giving a A type indicating that the message is a message, a type indicating that the message is a message for contacting the recipient regarding suspension of service provision, and a message for contacting the recipient regarding delivery or on-site work. and at least one of a type indicating that the message is a message requesting an answer to a survey.

(Appendix 92)
The information input to the trained model is
Manage all or part of the message, the number of stages of the transmission timing, information about the location of the terminal used by the recipient, information about the communication status of the terminal used by the recipient, and communication of the recipient information from the platform, information in the form of information in said message including at least one of text, images, video, sound, and speech by a visitor, said recipient perceiving information in said message. Recognition by visual reading, recognition by visual recognition, recognition by listening to one-way voice directed to the recipient, recognition by interacting through a terminal, and face-to-face or information on aspects including at least one of recognition by telephone interaction, information on costs of conveying said message, information on how to check the status of said recipient's response to said message, and said originator or said recipient. 92. The messaging scenario generator of any one of clauses 26 or 72-91, further comprising at least one of information relating to weighting of the messages for.

(Appendix 93)
The message weighting information includes:
from at least one of the immediacy required for sending the message, the urgency of the recipient acting in response to the message, the importance of the content of the message, and the certainty required for delivery of the message 93. The messaging scenario generator of Claim 92, including an evaluation of .

(Appendix 94)
In the process of outputting information including the transmission channel and transmission timing, the output information is
94. The messaging scenario generator of any one of clauses 26 or 72-93, further comprising recommendations regarding the form of information in the message.

(Appendix 95)
Said message is
95. The message transmission scenario generation program according to any one of attachments 26 or 72 to 94, which is one message or a combination of messages prepared for each transmission timing composed of a plurality of stages.

(Appendix 96)
The information about the response status is
76. The message delivery scenario generation program according to appendix 75, including at least one of information on delivery channel and information on delivery timing associated with response status.

(Appendix 97)
The information about the response status is
76. The message delivery scenario generation program according to appendix 75, wherein the message related to the message includes information about the response status to the message sent in the past.

(Appendix 98)
Messages related to said message and sent in the past are:
98. The message delivery scenario generation program according to appendix 97, wherein the message was sent in a previous stage when the delivery timing is composed of two or more stages.

(Appendix 99)
Information about the history of behavior of said recipient,
78. The message transmission scenario generation program according to appendix 77, including information on at least one of a history of consent to information handling, a history of personal authentication, a history of purchase, a history of browsing web pages, and a history of service usage.

(Appendix 100)
The information about the attribute of the recipient is
Possessed terminal, transmission channels available on the possessed terminal, applications used, date of birth, address, place of work, working hours, hobbies, preferences, family structure, contract status, financial status related to the message, and 79. The messaging scenario generating program of Claim 78, including information about at least one of the behavioral history characteristics.

(Appendix 101)
The process of obtaining information about the recipient includes:
101. The message transmission scenario generation program according to any one of Appendices 26 or 72 to 100, including a process of acquiring information about the recipient by accessing a personal information database storing information about an individual.

(Appendix 102)
A process for receiving information comprising said message comprises:
receiving from the originator further information identifying an intended recipient of the message;
Access to the personal information database includes:
102. The messaging scenario generation program according to appendix 101, which is executed based on the information identifying the recipient.

(Appendix 103)
The purpose of sending to said recipient is
responding to requests from said recipients; disseminating information to a group of recipients including said recipients; providing information to said recipients periodically; making requests to said recipients; 89. The messaging scenario generating program of Clause 88, comprising at least one of making a request to a person, warning the recipient, and advertising to the recipient.

(Appendix 104)
The types of recipients are:
Customer of said originator, potential customer of said originator, member of a specific organization, person to undergo certification procedures, person to give consent, user of a particular product or service, potential customer of a particular product or service users, persons who regularly receive information, persons who receive benefits or benefits, persons who receive demands for payment of money, delivery destinations, visiting work destinations, persons who receive requests to respond to surveys, and 89. The messaging scenario generator of Claim 89, including at least one of the residents of a particular locality.

(Appendix 105)
Recommendations regarding the form of said information are:
95. The messaging scenario generator of Claim 94, comprising at least one of text formatting recommendations, image usage recommendations, video usage recommendations, sound usage recommendations, and visitor speech recommendations.

100 Message transmission scenario generation device 102 Source server 104 Message reception unit 106 Recipient information acquisition unit 108 Message type identification unit 110 Message type related information storage unit 112 Message transmission scenario generation unit 114 Trained model 116 Message transmission scenario recording database 118 Individual Information database 120 Communication carrier server

According to the disclosed technique, a message receiver receives information including a message from an originator, a recipient information retriever retrieves information about the intended recipient of said message, and at least information related to the content of said message. a message type identifying unit that identifies a message type based on the above; inputs information including information about the message type and the recipient to a trained model ; a message delivery scenario generator for outputting information including a delivery channel and delivery timing for sending the message to the recipient.

Also according to the disclosed technique is a computer-implemented method for generating a message delivery scenario, comprising: receiving information containing a message from an originator; obtaining information about an intended recipient of said message; identifying a message type based on at least information related to content of the message; inputting information including information about the message type and the recipient to a trained model; and outputting from the trained model. and outputting information including a delivery channel and delivery timing for sending the message to the recipient in response to a message delivery scenario generation method.

Further, according to the disclosed technology, receiving information including a message from an originator, obtaining information about the intended recipient of said message, and identifying a message type based on at least information relating to the content of said message. , inputting information including information about the message type and the recipient to a trained model, and including a transmission channel and transmission timing for sending the message to the recipient according to the output from the trained model. A message transmission scenario generation program is provided that causes a computer to execute a process of outputting information.

Claims (26)

a message receiver that receives information, including a message, from an originator;
a recipient information obtaining unit for obtaining information about intended recipients of said message;
a message type identification unit that identifies a message type based on information including information related to the content of the message;
a message transmission scenario generation unit that outputs information including a transmission channel and transmission timing for transmitting the message to the recipient in response to input of information including information on the message type and the recipient to the trained model;
A messaging scenario generator comprising: The transmission channel, at each transmission timing,
2. The messaging of claim 1, wherein at least one of SMS, MMS, RCS, e-mail, social media, notifications on application programs, chatbots, phone calls, mail, flyers, and visits are selected. Scenario generator. The transmission timing is composed of one or more stages, and the timing of each stage is
Specific by day, specific by day of the week, specific by weekday or not, specific by time of day, specific by time of day, specific by time or not, specific by time range, random within a predetermined time range identification by being at a certain time, identification related to a predetermined event, identification related to the presence or absence of an action of the recipient, identification related to the location of the terminal used by the recipient, identification used by the recipient 3. A messaging scenario generator according to claim 1 or 2, comprising at least one of a specification related to communication status of a terminal and a specification related to information about said recipient. The information about said recipient is
4. A messaging scenario generator according to any one of claims 1 to 3, comprising information identifying at least one said recipient. The information about said recipient is
5. The messaging scenario generating device according to any one of claims 1 to 4, further comprising information on response statuses to past messages sent to said recipient. The information about the response status is
Information about opening a message, clicking or tapping on a part within a message, accessing a web page in response to a message, launching an application program in response to a message, starting a chat session in response to a message information about providing information in response to messages; information about registration in response to messages; information about identity verification in response to messages; information about purchasing in response to messages; Information about consent to the handling of information in response, information about payment of money in response to messages, information about inquiries related to messages, information about refusal to receive messages, information about answering telephone calls, information about responding to mail , and information on response to visits, and information on frequency of responses. The information about said recipient is
7. The message transmission scenario generation device according to any one of claims 1 to 6, comprising information on a history of behavior of said recipient, regardless of whether or not it is a response to past messages sent to said recipient. . The information about said recipient is
8. A messaging scenario generator according to any one of claims 1 to 7, comprising information about attributes of said recipient. The information about said recipient is
9. A messaging scenario according to any one of claims 1 to 8, comprising anonymized or partially anonymized information and/or encrypted or partially encrypted information. generator. The trained model is
Machine learning using teacher data so as to output information including the transmission channel and the transmission timing for sending the message to the recipient when information including the message type and information on the recipient is input. processed,
10. A messaging scenario generation device according to any one of claims 1 to 9. The teacher data is
11. The message delivery scenario generation device according to claim 10, comprising at least one of information about the recipient and information about a person other than the recipient. The trained model is
12. Learning is performed based on at least one of information on response statuses to past message transmissions to said recipient and information on response statuses to past message transmissions to persons other than said recipients. The message transmission scenario generation device according to any one of the above. The message receiving unit
further receiving information related to the message type from the originator;
The message type identification unit is
13. A messaging scenario generator as claimed in any one of the preceding claims, wherein the message type is identified based at least in part on information associated with the message type. further comprising a message type related information storage unit storing information related to the message type;
The message type identification unit is
14. The message transmission scenario generation device according to any one of claims 1 to 13, wherein the message type is specified by accessing the message type related information storage unit. Information related to the message type includes:
including a message type identification keyword,
The message type identification unit is
15. The message transmission scenario generation device according to claim 14, wherein the message type is specified by determining whether or not the extracted character string extracted from the message includes the message type specifying keyword. Information related to the message type includes:
contains trained data,
The message type identification unit is
The message transmission scenario generation device according to claim 14, wherein the message type is specified based on information including the message and the learned data. The message receiving unit
further receiving information from the source indicating a bias with respect to at least one of the transmission channel and the transmission timing;
The message transmission scenario generation unit
17. A messaging scenario generator according to any one of the preceding claims, outputting information comprising said transmission channel and said transmission timing at least partially adjusted according to said bias. The message type is
18. A messaging scenario generator according to any one of claims 1 to 17, comprising a type indicating the purpose of sending said message to said recipient. The message type is
19. The message transmission scenario generation device according to any one of claims 1 to 18, further comprising a type indicating a type of expected recipient of said message. The message type is
20. The message transmission scenario generation device according to any one of claims 1 to 19, further comprising a type based on the content of said message. The message type is
a type indicating that the message is a message providing disaster-related notification; a type indicating that the message is a message providing health-related notification; a type indicating that the message is a notification message related to problems or failures of products or services; a type indicating that the message is a message that provides information or advertisements related to products or services , a type indicating that the message is a message related to an authentication procedure, a type indicating that the message is a message related to an authentication procedure and is a message related to a form of authentication procedure requested of the recipient; A type indicating that the message is a message related to the consent required for the use of the service of the recipient, and whether or not the recipient consents to the handling of information related to the recipient in the message. The message is information related to the recipient, and the recipient or a subject who is in a position to give consent to the handling of information related to the recipient consents to the handling a type indicating that the message is a message relating to information giving a A type indicating that the message is a message, a type indicating that the message is a message for contacting the recipient regarding suspension of service provision, and a message for contacting the recipient regarding delivery or on-site work. and a type indicating that the message is a message requesting an answer to a survey. The information input to the trained model is
Manage all or part of the message, the number of stages of the transmission timing, information about the location of the terminal used by the recipient, information about the communication status of the terminal used by the recipient, and communication of the recipient information from the platform, information in the form of information in said message including at least one of text, images, video, sound, and speech by a visitor, said recipient perceiving information in said message. Recognition by visual reading, recognition by visual recognition, recognition by listening to one-way voice directed to the recipient, recognition by interacting through a terminal, and face-to-face or information on aspects including at least one of recognition by telephone interaction, information on costs of conveying said message, information on how to check the status of said recipient's response to said message, and said originator or said recipient. 22. A messaging scenario generator according to any one of claims 1 to 21, further comprising at least one information relating to the weighting of said messages for. The message weighting information includes:
from at least one of the immediacy required for sending the message, the urgency of the recipient acting in response to the message, the importance of the content of the message, and the certainty required for delivery of the message 23. The messaging scenario generator of claim 22, comprising an evaluation of . The information output by the message transmission scenario generation unit includes:
24. A messaging scenario generator according to any one of the preceding claims, further comprising recommendations regarding the form of information in said message. A computer-executed method for generating a message-passing scenario, comprising:
receiving information, including a message, from an originator;
obtaining information about intended recipients of said message;
identifying a message type based on information including information related to content of the message;
and outputting information including a transmission channel and transmission timing for transmitting the message to the recipient in response to input of information including information on the message type and the recipient to a trained model. Method. receive information, including messages, from originators;
obtaining information about the intended recipient of said message;
identifying a message type based on information including information related to the content of the message;
Outputting information including a transmission channel and transmission timing for transmitting the message to the recipient according to input of information including information on the message type and the recipient to a trained model Message transmission for causing a computer to execute processing Scenario generator.

Images