JP2023004580A - Program, method and system - Google Patents

Abstract

To provide a system easily setting up specific conditions to be required when selecting a user.SOLUTION: A program according to the present disclosure causes a processor of a computer to execute the steps of: acquiring operation logs of user terminals reaching a specific event; generating a user behavior model for estimating a possibility of whether or not a user reaches the specific event by using the acquired operation log; setting conditions relating to the possibility of whether or not to reach the event and setting an action to be executed for a user satisfying the conditions; and executing the action for a user having a high possibility of reaching the event, calculated by using the user behavior model.SELECTED DRAWING: Figure 6

Description

The present disclosure relates to programs, methods, and systems that provide appropriate information to users.

Conventionally, there has been known a system in which a system administrator prompts a user to perform an action expected of the user by performing some action on the user when the user's action in the WEB environment satisfies a specific condition.
For example, in Patent Document 1, when a user satisfies a specific condition such as browsing a specific website, an advertisement for a product is presented to the user for the purpose of increasing the user's willingness to purchase. A system is disclosed.

JP 2020-144562 A

However, in the system described in Patent Literature 1, it is necessary to set specific conditions necessary for selecting a user for each subsequent action. Setting appropriate conditions requires specialized knowledge for each field to which the system is applied. For this reason, it takes effort to set the conditions necessary for user selection.

Therefore, an object of the present disclosure is to provide a system that can easily set specific conditions necessary to select a user.

According to one aspect of the present disclosure, a program instructs a processor of a computer to acquire an operation log of a user terminal leading to a specific event; a step of generating a user behavior model for estimating a degree; a step of setting a condition regarding the degree of whether or not an event will occur; a step of setting an action to be executed for a user who satisfies the condition; Also, a step of executing an action for a user who is highly likely to reach an event is executed.

According to the present disclosure, it is possible to easily set specific conditions and corresponding actions required to select users.

BRIEF DESCRIPTION OF THE DRAWINGS It is a figure explaining the outline|summary of the information provision system of this invention. 3 is a block diagram showing a functional configuration of a user terminal that configures the information providing system; FIG. It is a block diagram which shows the functional structure of the processing server which comprises an information provision system. It is a figure which shows the specific example of the question-and-answer database which a memory|storage part memorize|stores. FIG. 4 is a diagram showing a specific example of an operation log database stored in a storage unit; FIG. 4 is a table for explaining segment conditions; FIG. 10 is a diagram illustrating a process of setting segment conditions; FIG. 10 is a diagram for explaining action execution processing; It is a figure explaining the threshold value set in multiple numbers. It is a block diagram showing a functional configuration of a processing server according to the second embodiment. It is a figure explaining the process of setting of segment conditions which concerns on 2nd Embodiment. FIG. 11 is a diagram illustrating processing for executing an action according to the second embodiment; FIG.

<First Embodiment>
An information providing system 1 according to a first embodiment of the present disclosure will be described below with reference to the drawings. It should be noted that the embodiments described below do not unduly limit the content of the present disclosure described in the claims. Moreover, not all the configurations described in the embodiments are essential constituent elements of the present disclosure. Further, in all the drawings for explaining the embodiments, common constituent elements are given the same reference numerals, and repeated explanations are omitted.

<Overview>
FIG. 1 is a diagram explaining an outline of an information providing system 1 of the present invention.
As shown in FIG. 1, the information providing system 1 according to the present embodiment analyzes the user's behavior in the WEB environment and provides information for prompting the user to perform the behavior expected by the system administrator. System.

In the following description, after first describing the configuration of the information providing system 1, each process performed by the information providing system 1 will be described in detail.

<Overall composition>
An overall configuration of an information providing system 1 according to this embodiment will be described.
As shown in FIG. 1, an information providing system 1 includes a plurality of users (in FIG. 1, user terminals 10A and 10B used by the users, hereinafter collectively referred to as "user terminals 10") and a processing server 20. include. A plurality of user terminals 10 and the processing server 20 are connected via a network 80 so as to be able to communicate with each other. Network 80 is configured by a wired or wireless network.

The user terminal 10 is a device operated by each user. The user terminal 10 is implemented by a stationary PC (Personal Computer), a laptop PC, or the like. In addition, the user terminal 10 may be, for example, a mobile terminal such as a tablet compatible with a mobile communication system or a smart phone.

The user terminal 10 is communicably connected to the processing server 20 via the network 80 . The user terminal 10 is a wireless base station 81 compatible with communication standards such as 5G and LTE (Long Term Evolution), and a wireless LAN (Local Area Network) standard such as IEEE (Institute of Electrical and Electronics Engineers) 802.11. It is connected to the network 80 by communicating with a communication device such as a wireless LAN router 82 . As shown in FIG. 1 , the user terminal 10 includes a communication IF (Interface) 12 , an input device 13 , an output device 14 , a memory 15 , a storage section 16 and a processor 19 .

The communication IF 12 is an interface for inputting and outputting signals for the user terminal 10 to communicate with an external device.
The input device 13 is a device (for example, a keyboard, a touch panel, a touch pad, a pointing device such as a mouse, etc.) for receiving an input operation from a user.

The output device 14 is a device (display, speaker, etc.) for presenting information to the user.
The memory 15 temporarily stores programs and data processed by the programs, and is a volatile memory such as a DRAM (Dynamic Random Access Memory).

The storage unit 16 is a storage device for storing data, and is, for example, a flash memory or a HDD (Hard Disc Drive).
The processor 19 is hardware for executing an instruction set described in a program, and is composed of arithmetic units, registers, peripheral circuits, and the like.

The processing server 20 is a device that analyzes the behavior pattern of the user and provides information that prompts the user to behave as expected by the administrator of the processing server 20 . The administrator of the processing server 20 refers to the operator of the service in which the processing server 20 is implemented. For example, when the processing server 20 is installed in an EC site, the administrator refers to the operator of the EC site. In this case, the behavior expected of the user by the administrator refers to the purchase of goods by the user at the EC site.

The processing server 20 is a computer connected to the network 80 . The processing server 20 includes a communication IF 22 , an input device 23 , an output device 24 , a processor 25 , a memory 26 and a storage 27 .

The communication IF 22 is an interface for inputting and outputting signals so that the processing server 20 communicates with an external device.
The input device 23 is a device (for example, a keyboard, a touch panel, a touch pad, a pointing device such as a mouse, etc.) for receiving an input operation from an administrator.
The output device 24 is a device (display, speaker, etc.) for presenting information to the administrator.

The processor 25 is hardware for executing an instruction set described in a program, and is composed of arithmetic units, registers, peripheral circuits, and the like.
The memory 26 temporarily stores programs and data processed by the programs, and is a volatile memory such as a DRAM (Dynamic Random Access Memory).
The storage 27 is a storage device for storing data, such as a flash memory or HDD (Hard Disc Drive).

<Configuration of user terminal 10>
FIG. 2 is a block diagram showing the functional configuration of the user terminal 10 that constitutes the information providing system 1. As shown in FIG. The user terminal 10 is assumed to be, for example, a terminal handled by a general user who accesses an EC site managed by a server administrator and considers purchasing an article. As shown in FIG. 2, the user terminal 10 includes a plurality of antennas (antenna 111, antenna 112), wireless communication units (first wireless communication unit 121, second wireless communication unit 122) corresponding to each antenna, input It includes reception unit 130 (including keyboard 131 and display 132 ), audio processing unit 140 , microphone 141 , speaker 142 , camera 150 , storage unit 160 , and control unit 170 .

The user terminal 10 also has functions and configurations not particularly shown in FIG. 2 (for example, a battery for retaining power, a power supply circuit for controlling power supply from the battery to each circuit, etc.). As shown in FIG. 2, each block included in the user terminal 10 is electrically connected by a bus or the like.

The antenna 111 radiates a signal emitted by the user terminal 10 as radio waves. Antenna 111 also receives radio waves from space and provides a received signal to first wireless communication section 121 .

The antenna 112 radiates a signal emitted by the user terminal 10 as radio waves. Antenna 112 also receives radio waves from space and provides a received signal to second wireless communication section 122 .

The first radio communication unit 121 performs modulation and demodulation processing for transmitting and receiving signals via the antenna 111 so that the user terminal 10 communicates with other radio equipment. The second radio communication unit 122 performs modulation/demodulation processing and the like for transmitting and receiving signals via the antenna 112 so that the user terminal 10 communicates with other radio equipment. The first radio communication unit 121 and the second radio communication unit 122 are communication modules including a tuner, an RSSI (Received Signal Strength Indicator) calculation circuit, a CRC (Cyclic Redundancy Check) calculation circuit, a high frequency circuit, and the like. The first radio communication unit 121 and the second radio communication unit 122 perform modulation/demodulation and frequency conversion of radio signals transmitted/received by the user terminal 10 and give received signals to the control unit 170 .

The input receiving unit 130 has a mechanism for receiving user input operations. Specifically, input accepting unit 130 includes a keyboard 131 and a display 132 . Note that the input reception unit 130 may be configured as a touch screen that detects the user's touch position on the touch panel by using, for example, a capacitive touch panel.

The keyboard 131 accepts input operations of the user of the user terminal 10 . The keyboard 131 is a device for inputting characters, and outputs input character information to the control unit 170 as an input signal.

The display 132 displays data such as images, moving images, and text under the control of the control unit 170 . That is, the display 132 also functions as an output unit that outputs information to the user. The display 132 is implemented by, for example, an LCD (Liquid Crystal Display) or an organic EL (Electro-Luminescence) display.

The audio processing unit 140 modulates and demodulates an audio signal. Audio processing section 140 modulates the signal supplied from microphone 141 and supplies the modulated signal to control section 170 . Also, the audio processing unit 140 provides an audio signal to the speaker 142 .
The audio processing unit 140 is implemented by, for example, a processor for audio processing. Microphone 141 receives a voice input and provides voice signal corresponding to the voice input to voice processing section 140 . The speaker 142 converts the audio signal given from the audio processing unit 140 into audio and outputs the audio to the outside of the user terminal 10 .

The camera 150 is a device for receiving light with a light receiving element and outputting it as a photographed image. The camera 150 is, for example, a depth camera that can detect the distance from the camera 150 to the subject.

The storage unit 160 is configured by, for example, a flash memory or the like, and stores data and programs used by the user terminal 10 . In one aspect, storage unit 160 stores user information 161 .

The user information 161 is information of a user who utilizes the information providing system 1 using the user terminal 10 . The user information 161 includes contact information such as the user's name, the user's mail address, and the IP address of the user terminal.

The control unit 170 reads programs stored in the storage unit 160 and executes commands included in the programs, thereby controlling operations of the user terminal 10 . The control unit 170 is, for example, an application program installed in the user terminal 10 in advance. The control unit 170 functions as an operation reception unit 171 , a transmission/reception unit 172 , a data processing unit 173 , and a display processing unit 174 by operating according to programs.

The operation accepting unit 171 performs processing for accepting a user's input operation to the input device such as the keyboard 131 .

The transmitting/receiving unit 172 performs processing for the user terminal 10 to transmit/receive data to/from an external device such as the processing server 20 according to a communication protocol.

The data processing unit 173 performs processing according to a program to perform calculations on data input by the user terminal 10, and outputs the calculation results to a memory or the like.

The display processing unit 174 performs processing for presenting information to the user. The display processing unit 174 performs processing for displaying a display image on the display 132, processing for outputting sound to the speaker 142, processing for causing the camera 150 to vibrate, and the like.

<Functional Configuration of Processing Server 20>
FIG. 3 is a block diagram showing the functional configuration of the processing server 20 forming the information providing system 1. As shown in FIG. The processing server 20 is a server device that performs processing as the information providing system 1 .
The processing server 20 is connected, for example, via a network 80 to an EC site management terminal managed by an EC site operating company.
As shown in FIG. 3 , the processing server 20 functions as a communication section 201 , a storage section 202 and a control section 203 .

The communication unit 201 performs processing for the processing server 20 to communicate with an external device.

The storage unit 202 stores data and programs used by the processing server 20 . The storage unit 202 stores at least a question-and-answer database, an operation log database, and a user behavior model.

The question-and-answer database is a database that stores conversational sentences used in an automatic response system using a chatbot. That is, the information providing system 1 may have a chatbot function.
In the question-and-answer database, for example, one title and one answer pattern exist for one theme, and one or more question patterns assumed for that theme are stored.

FIG. 4 is a diagram showing a specific example of the question-and-answer database stored in the storage unit 202. As shown in FIG. The question-and-answer database has a "title" field, a "question pattern" field, and an "answer pattern" field.

The "title" field stores, for example, information about the theme of the conversation that simply indicates the content of the question pattern and answer pattern.
The "question pattern" field stores information about question sentences expected to be input by the user.
The "answer pattern" field stores information about an answer sentence to be given to the user in correspondence with the title.

As shown in FIG. 4, one question pattern does not have to correspond to one answer pattern. For example, when making a purchase on an EC (Electronic Commerce) site, if the user wants to know how to change the address, various types of information such as "I want to change the address." It is useful to register a plurality of question patterns because the way of asking can be considered.

The question-and-answer database may also store categories. Categories are used to generate the page structure of the FAQ page. Also, when the administrator edits the question-and-answer database, it may be used to generate the page structure on the administrator's setting screen. Note that the page structure may use not only grouping by category but also order and hierarchy.

The operation log database stored in the storage unit 202 is a database that stores user operation logs acquired by the processing server 20 . FIG. 5 is a diagram showing a specific example of the operation log database stored in the storage unit 202. As shown in FIG.
As shown in FIG. 5, the operation log database includes an "operation log ID" field, a "user ID" field, an "operation target" field, an "operation log" field, and a "date and time" field. there is

The “operation log ID” field stores an ID for identifying the user's operation log acquired by the processing server 20 .

The "user ID" field stores the user ID of the user who performed the operation corresponding to the operation log ID.

The "operation target" field stores the URL of the site where the operation corresponding to the operation log ID was performed.

The "operation log" field stores an operation log corresponding to the operation log ID. The operation log is chronological information, and includes, for example, login operations, page transition operations, page stay operations, product selection operations by clicking, etc., together with time information.

The "date and time" field stores the date and time of the operation corresponding to the operation log ID.

The user behavior model stored in the storage unit 202 is a learned model for estimating the extent to which the user, who is the subject of evaluation, will reach a specific event. The user behavior model learns the user's operation log and events that occurred in the operation log as training data.
The user behavior model uses a user's operation log as input data, and uses an event occurrence probability indicating whether or not the user will reach a predetermined event as output data. Here, the predetermined event includes an action desirable for the EC site administrator, such as a user purchasing a product at the EC site. As the predetermined event, for example, an action that is undesirable for the administrator, such as the user's withdrawal from the mail-order service that provides the EC site, may be set.

The user behavior model according to this embodiment is, for example, a parameterized synthetic function obtained by synthesizing a plurality of functions. A parameterized composite function is defined by a combination of multiple adjustable functions and parameters. The user behavior model according to this embodiment may be any parameterized composite function that satisfies the above requirements, but is assumed to be a multi-layered network model (hereinafter referred to as a multi-layered network). A user behavior model using a multilayer network has an input layer, an output layer, and at least one intermediate layer or hidden layer provided between the input layer and the output layer. User behavior models are envisioned for use as program modules that are part of artificial intelligence software.

As the multi-layered network according to the present embodiment, for example, a deep neural network (DNN), which is a multi-layered neural network targeted for deep learning, can be used. As the DNN, for example, a convolution neural network (CNN) targeting images may be used.

3, the processor 29 of the processing server 20 performs processing according to a program, so that various modules such as a transmission/reception control module 2031, an operation log acquisition module 2032, a model generation module 2033, an action setting module 2034, and a It functions as an action estimation module 2035 and an action execution module 2036 .

The transmission/reception control module 2031 controls processing for the processing server 20 to transmit and receive signals from an external device according to a communication protocol.

Also, the operation log acquisition module 2032 acquires a user's operation log. The operation log acquisition module 2032 acquires the user's operation log by receiving the user's operation log acquired by the user terminal 10 from the user terminal 10 . The operation log acquisition module 2032 stores the acquired user operation log in the operation log database.

Model generation module 2033 generates a user behavior model. The model generation module 2033 uses the user's operation log to generate a user behavior model. Specifically, the model generation module 2033 creates an operation log related to a user who has reached a certain event (purchased a certain product), an operation log related to a user who has not reached the event (has not purchased the product), to get At this time, as the operation log, the operation log from the time of visiting the EC site until the occurrence of an event, or the operation log from the time of visiting the EC site to immediately before the occurrence of the event is acquired.

As an example, the model generation module 2033 annotates information regarding whether or not an event has occurred for each operation log of a user who has reached a certain event and a user who has not reached a certain event, and learns the user behavior model. Let The user behavior model outputs the event occurrence probability leading to the event when receiving the input of the operation log.

The action setting module 2034 sets segment conditions for executing actions. Specifically, the action setting module 2034 sets the content of the segment indicating the condition for specifying the user input from the administrator terminal of the EC site and the action to be executed for the user who satisfies the specific condition to the segment. Set as a condition. FIG. 6 is a table for explaining segment conditions.

As shown in FIG. 6, the action setting module 2034 can set a plurality of segment conditions by using a plurality of user behavior models. Examples of segment contents include the following.
・Presenting product advertisements to users who are likely to purchase products ・Popup chatbot windows for users who are likely to use chatbots ・For users who are likely to add products to their favorites, other users Presenting evaluation comments from users ・Providing discount information to users who tend to browse product pages for a long time ・Providing Q&A guidance to users who are likely to browse Q&A.
・Present related products to users who are likely to purchase related products. be able to.

The segment condition may be, for example, the following contents, and can be arbitrarily set according to the administrator's request.
・For users who are likely to withdraw from the membership, information is provided to persuade them to withdraw. ・For users who are likely to enter evaluation comments, guidance is provided on granting benefits by entering comments. ・For users who are likely to return products, Present the return method ・Display the Q&A page for users who are likely to check the Q&A

Behavioral estimation module 2035 estimates the extent to which the user will lead to the event. Specifically, the behavior estimation module 2035 inputs the user's operation log to the user behavior model, and estimates whether the user will reach the event based on the event occurrence probability output from the user behavior model. .

The action execution module 2036 executes the presentation of information (action) for the user who is highly likely to lead to the event as estimated by the behavior estimation module 2035 so as to lead the user to the event. Specifically, the action execution module 2036 executes actions according to the set segment conditions. For example, in the example of FIG. 6, a product advertisement is presented as an action to the user who is likely to purchase the product.

<Process flow>
Next, processing of the information providing system 1 will be described. First, the process of setting segment conditions will be described. FIG. 7 is a diagram for explaining the process of setting segment conditions.

As shown in FIG. 7, the processing server 20 acquires operation logs of multiple users (step S200). Specifically, the operation log acquisition module 2032 of the processing server 20 periodically acquires operation logs of all users.

After step S201, the processing server 20 generates a user behavior model (step S201). Specifically, the model generation module 2033 of the processing server 20 classifies all the acquired operation logs of users into operation logs of users who have reached a certain event and operation logs of users who have not reached a certain event. do.

Next, the model generation module 2033 annotates the classified operation log with information regarding whether an event has occurred, and makes the user behavior model learn. A user behavior model is thereby generated.
It should be noted that the model generation module 2033 can predict the content of the segment condition to be input in subsequent processing and create a plurality of user behavior models in advance.

Next, the processing server 20 sets segment conditions (step S202). Specifically, the content of the segment and the content of the action to be executed are set based on the administrator's input operation from the management terminal of the EC site. It should be noted that the setting of actions may be performed, for example, by referring to past measures obtained statistically. This completes the segment condition setting process.
Moreover, in setting the segment conditions, the segment conditions corresponding to a plurality of user behavior models created in advance may be set by a procedure in which the administrator selects. In this case, the action to be executed may be newly set by input from the administrator.

Note that segment conditions may be set before the user behavior model is generated (step S201). In this case, the model generation module 2033 generates a user behavior model according to the item input as the content of the segment.

Next, the processing of action execution for the user will be described. FIG. 8 is a diagram for explaining action execution processing.
As shown in FIG. 8, the user terminal 10 inputs an operation (step S110). Specifically, the user operates the user terminal 10, visits the EC site, and considers purchasing a product.

After step S110, the processing server 20 acquires the user's operation log from the user terminal 10 (step S210). Specifically, the operation log acquisition module 2032 acquires a series of user operation logs and stores them in the operation log database. The operation log is acquired, for example, at intervals of one minute. By acquiring it periodically, the user's operation log can be evaluated periodically.

After step S210, the processing server 20 calculates the user's event occurrence probability (step S211). Specifically, the behavior estimation module 2035 of the processing server 20 refers to the operation log database and inputs a series of user operation logs into the user behavior model. Thereby, the user's event occurrence probability is calculated. By calculating the user's event occurrence probability in this way, it is possible to quantitatively evaluate whether or not the user will reach the event. The behavior estimation module 2035 periodically calculates the event occurrence probability for the user's operation log that is acquired periodically. The behavior estimation module 2035 may calculate the event occurrence probability periodically, for example, at the same interval as the acquisition of the operation log. Note that the action estimation module 2035 may periodically calculate the event occurrence probability at an interval longer than the acquisition of the operation log.

Also, when a plurality of user behavior models are used, in step S211 shown in FIG. 8, the behavior estimation module 2035 inputs user operation logs for each of the plurality of user behavior models. Thereby, the action estimation module 2035 calculates the event occurrence probability output from each user action model.
Further, subsequent processing is performed on the calculated event occurrence probabilities of the plurality of types of events.

After step S211, the processing server 20 determines an action (step S212). Specifically, the action execution module 2036 of the processing server 20 compares the event occurrence probability with a preset threshold, and determines the corresponding action. At this time, if there is no applicable action, no action is determined.

After step S212, the processing server 20 executes an action on the user terminal 10 (step S213). Specifically, the action execution module 2036 of the processing server 20 executes the determined action. For example, a product advertisement is presented to the user terminal 10 . Information presented to the user terminal 10 is transmitted to the user terminal 10 via the transmission/reception control module 2031 of the processing server 20 . Then, the action execution module 2036 executes the action immediately after calculating the event occurrence probability in step 211 . Timely information can be provided to the user by immediately executing the action in response to the calculation of the event occurrence probability which is periodically and repeatedly executed.

After step S<b>213 , the user terminal 10 causes the display processing unit 174 to display the content of the action from the processing server 20 on the display 132 . For example, the discount information transmitted from the processing server 20 is displayed on the display 132 . This completes the action execution process for the user.

As described above, according to the information providing system 1 of the present disclosure, the user behavior model generated from the user's operation log is used to estimate the degree of whether or not the user's event will occur. take action on it. Therefore, it is possible to easily set specific conditions necessary for selecting a user to execute an action.

In the above description, the occurrence of an event is described as "purchasing a product", but the present invention is not limited to this example. For example, the occurrence of an event may be set as "inquiry by chatbot", and the chatbot function may be proposed to a user who is likely to make an inquiry by chatbot, prior to the inquiry from the user. In this case, the executed action refers to the display of the question input window of the chatbot function.

In this case, the operation log is classified into the operation log of the user who made an inquiry using the chatbot function (which led to an event) and the operation log of the user who did not make an inquiry using the chatbot function (did not lead to the event).
Then, annotate information about whether or not an event has occurred to the operation log for which the inquiry was made and the operation log for which the inquiry was not made, and let the user behavior model learn. Accordingly, the behavior estimation module 2035 can calculate the probability that the user will make an inquiry using the chatbot as the event occurrence probability by inputting the user's operation log into the user behavior model. Then, when the event occurrence probability exceeds a preset threshold value, a process of popping up a question input window of the chatbot function is executed as the set action.

It is also possible to create a user behavior model in which an "inquiry about a specific theme in a chatbot" is set as an event occurrence. In this case, the operation log of the user who made an inquiry about a specific theme in the chatbot (which led to an event) and the user who did not make an inquiry about the specific theme in the chatbot (did not lead to an event) operation log and classify.
Then, for the classified two types of operation logs, annotate information about whether or not an event has occurred, and let the user learning model learn.
Conversation cards to be exchanged in the chatbot are set in the action to be executed set as the segment condition. That is, in the conversation card regarding a specific theme, the presentation of the answer sentence set as the answer pattern is set as the action.

Specifically, for a user who seems to be having trouble logging in, the chatbot function is immediately activated in response to the user's operation in the form of displaying answer patterns related to the theme of "I can't log in." is applicable.

<Modification 1>
Further, in the information providing system 1, the processing server 20 may select an action to be executed according to multiple set thresholds. Segment conditions in this case will be described. FIG. 9 is a diagram explaining a plurality of set segment conditions. In this example, a segment condition targeting users who are likely to purchase a product will be described.

As shown in FIG. 9, for users with an event occurrence probability of 90% or more, an action is set to present a product advertisement to encourage the user to purchase the product. This is because viewing the product advertisement may increase the willingness to purchase.
Next, for example, for a user with an event occurrence probability of 80% or more and less than 90%, an action is set to present highly rated comments by other users in order to encourage the purchase of the product. This is because there is a possibility that the purchase intention will increase due to the high evaluation of others.

Next, for example, for users with an event occurrence probability of 70% or more and less than 80%, an action is set to present discount information in order to encourage them to purchase products. This is because there is a possibility that they will purchase the product when the price becomes low.
Next, for example, for users with an event occurrence probability of 60% or more and less than 70%, an action is set to present a similar product to encourage them to purchase the product. This is because comparison with similar products may increase the willingness to purchase.

Next, for example, for a user with an event occurrence probability of 50% or more and less than 60%, in order to encourage the purchase of the product, an action is to present a related product to be used together with the product that the user is considering purchasing. is set. This is because confirmation of related products may increase the willingness to purchase.
In this way, by selecting an action to be executed according to a plurality of set thresholds, it is possible to give a degree of freedom to measures leading to action generation.

<Second embodiment>
Next, the information providing system 1 according to the second embodiment will be described. In the information providing system 1 according to the second embodiment, the processing server 20B automatically sets the contents of actions to be executed in the segment conditions. FIG. 10 is a block diagram showing the functional configuration of the processing server 20B according to the second embodiment. In addition, the same code|symbol is assigned about the structure same as 1st Embodiment, and the description is abbreviate|omitted.

The storage unit 202 of the processing server 20B further stores action setting models.

The action setting model is a trained model that estimates an action that should be performed to bring about a specific event to a user who is highly likely to reach the event. The action setting model learns the user's operation log and the process performed in the operation log immediately before the occurrence of the event as teacher data.

The action setting model uses a user's operation log as input data, and an action that the user should perform to bring about a predetermined event as output data. An action setting model is generated by the model generation module 2033 .
Next, a process for setting segment conditions according to the second embodiment will be described. FIG. 11 is a diagram for explaining the segment condition setting process according to the second embodiment.

As shown in FIG. 11, after generating a user behavior model (step S201), an action setting model is generated (step 201B). Specifically, the model generation module 2033 acquires the action history of the user who reached a certain event immediately before the event occurred. The immediately preceding action history is information relating to user actions that were performed before the occurrence of an event and that are recognized to be directly related to the occurrence of the event. Whether or not there is a direct relationship with the occurrence of an event is determined by the timing and content of processing. This action history includes information output from the processing server 20 to the user terminal 10 and performed by the user terminal 10 for display processing, in addition to input operations from the user.

Next, the model generation module 2033 identifies and associates the feature amount of the operation history up to the occurrence of the user's event with the correlated immediately preceding action history, and makes the user action model learn them. This creates an action setting model.

After step 201B, the administrator sets segment conditions. At this time, the content of the segment and the threshold are set according to the content of the user behavior model to be used. On the other hand, among the segment conditions, the contents of actions are not set at this point.
Next, processing for executing an action according to the second embodiment will be described. FIG. 12 is a diagram illustrating action execution processing according to the second embodiment.

As shown in FIG. 12, the processing server 20B determines an action (step S212B) after the action estimation module 2035 calculates the event occurrence probability (step S211). Specifically, the action setting module 2034 of the processing server 20 obtains the output of the action to be executed by inputting the user's operation log to the action setting model. As a result, the contents of actions that have not been set in the segment conditions are determined.
In other words, the action setting module 2034 sets the action to be executed from the action history of the other user who reached the specific event, just before the event.

As a specific example of the second embodiment, the user's action history learned by the action setting module 2034 includes conversation cards exchanged in the chatbot under the segment condition targeting users who are likely to purchase. The action to be executed includes presentation of the answer card received by the other user who caused the event immediately before the event occurred. That is, based on a case where a certain user made a purchase after confirming an answer card in a chatbot, a similar answer card is presented to a user who has made a similar action history to encourage purchase. can be done.

After step S212B, the action execution module 2036 executes the action (step S213) and the user terminal 10 displays the contents of the action (step S111). This completes the action execution process according to the second embodiment.

As described above, according to the information providing system 1 according to the second embodiment, using the action setting model, from the action history immediately before the event of another user who has reached a specific event, Actions to be executed are set. Therefore, the action need not be set by the administrator, and the appropriate action can be set for each user by a simple process of acquiring the user's operation log.

<Other Modifications>

In each of the above-described embodiments, the configuration in which the information providing system 1 is mounted on the EC site is shown, but the information providing system 1 is not limited to such an aspect, and various services provided under the WEB environment can be implemented in

Also, the user behavior model and action setting model do not have to be generated in the processing server 20 . That is, in another server, each trained model may be generated and stored in the storage unit 202 to be used for action execution processing. Also, each trained model may be re-learned using the latest user operation log.

Also, in each of the above-described embodiments, the segment condition is set as a source code to be executed by the program, but the configuration is not limited to this. The segment conditions may be stored in the storage unit 202 of the processing server 20 as a data table.

Also, the question-and-answer database may be realized by a server that implements a program on a computer. Each part constituting the question answering server may be implemented by being distributed over a plurality of computers.

The embodiments of the present disclosure have been described above. It is intended to be included within the scope of the invention corresponding to the examples.

The contents of the present invention are added below.

(Appendix 1)
It will be created after the claim is confirmed.

1 information providing system 10 user terminal 20 question answering server 203 control unit 2031 transmission/reception control module 2032 operation log acquisition module 2033 model generation module 2034 action setting module 2035 action estimation module 2036 action execution module

Claims (10)

computer processor,
a step of acquiring an operation log of a user terminal leading to a specific event;
generating a user behavior model for estimating the extent to which the user will reach the specific event using the acquired operation log;
setting a condition regarding the extent to which the event is reached, and setting an action to be performed for users who satisfy the condition;
executing the action for a user who is likely to reach the event, calculated using the user behavior model;
program to run. In the step of acquiring the operation log, the user's operation log is acquired periodically;
In the step of setting the action,
2. The program according to claim 1, which calculates an event occurrence probability indicating a probability that an event will occur as a condition relating to the degree of occurrence of said event. In the step of performing said action,
3. The program according to claim 1, which presents information for leading the user to the specific event. In the step of setting the action,
4. The program according to any one of claims 1 to 3, wherein said action is set from the action history immediately before said event of another user who led to said specific event. 5. The program according to any one of claims 1 to 4, wherein said actions include presentation of conversation cards exchanged in a chatbot. The user's action history includes conversation cards exchanged in the chatbot,
6. The program according to claim 5, wherein said action includes presenting an answer card received immediately before the event occurrence by another user leading to the event occurrence. The user behavior model includes:
7. The program according to any one of claims 1 to 6, wherein the program is a trained model that receives a user's operation log as an input and outputs a probability of whether the event will occur. User behavior model
8. The program according to claim 7, wherein the program is a trained model that has pre-learned information annotated with information indicating whether or not the specific event has occurred with respect to the user's operation log. the computer's processor
a step of acquiring an operation log of a user terminal leading to a specific event;
generating a user behavior model for estimating the extent to which the user will reach the specific event using the acquired operation log;
setting a condition regarding the extent to which the event is reached, and setting an action to be performed for users who satisfy the condition;
executing the action for a user who is likely to reach the event, calculated using the user behavior model;
how to run. comprising a computer having a processor;
The processor
a first module that acquires an operation log of a user terminal leading to a specific event;
a second module that uses the acquired operation log to generate a user behavior model that estimates the extent to which the user will reach the specific event;
a third module for setting a condition regarding the extent to which the event is reached and for setting an action to be performed for a user who satisfies the condition;
a fourth module that executes the action for a user who is likely to reach the event, calculated using the user behavior model;
A system with

Images