JP2021128553A - Apparatus, method and program for information processing - Google Patents

Abstract

To provide an apparatus, method and program for information processing adapted to find out more effectively a defection cause of a user as to online service.SOLUTION: An information processing apparatus 100 comprises a control section 130 including: an acquiring section 132 that acquires sequence data including a utilization of a predetermined target service indicative of a transition of service utilization of each user; and a learning section 134 that makes a model learn to output defection period-of-time information indicative of a period of time from a utilization of a target service by the user to a re-utilization of the target service from the sequence data.SELECTED DRAWING: Figure 4

Description

The present invention relates to an information processing device, an information processing method, and an information processing program.

Technology for grasping the actual situation of user withdrawal from online services is disclosed. For example, the content transition of the online service used by the user is detected, and the transition eigenvector, which is the probability of staying in the state, is calculated from the degree of transition to other content for each content, and any other content is used. A technique is known in which the degree of non-transition to the content is calculated for each content as the withdrawal probability, and the value obtained by multiplying the transition unique vector for each content by the withdrawal probability for each content is calculated as the withdrawal factor value.

Japanese Unexamined Patent Publication No. 2019-15291

However, in the above-mentioned conventional technique, it may not be possible to say that the user's withdrawal mode is estimated in detail. For example, in the above-mentioned prior art, the withdrawal mode of the user is only estimated based on the withdrawal probability for each content. Therefore, it cannot be said that the intention (context) of the user to leave is properly estimated.

The present application has been made in view of the above, and an object of the present application is to find the cause of withdrawal more effectively.

The information processing device according to the present application is sequence data indicating a transition of service usage of each user, and the user can obtain sequence data including the use of a predetermined target service from the acquisition unit and the sequence data. It is characterized by including a learning unit that trains a model to output withdrawal period information indicating a period from the use of the target service to the reuse of the target service.

According to one aspect of the embodiment, the cause of withdrawal can be found more effectively.

FIG. 1 is an explanatory diagram showing an outline of an information processing method according to an embodiment. FIG. 2A is an explanatory diagram showing an outline of a purpose-of-use prediction model. FIG. 2B is an explanatory diagram showing an outline of the withdrawal period prediction model. FIG. 2C is an explanatory diagram showing a transition sequence to each function of one application. FIG. 3 is a diagram showing a configuration example of the information processing system according to the embodiment. FIG. 4 is a diagram showing a configuration example of the information processing device according to the embodiment. FIG. 5 is a diagram showing an example of a user database. FIG. 6 is a diagram showing an example of a sequence database. FIG. 7 is a diagram showing an example of a withdrawal period database. FIG. 8 is a flowchart showing a processing procedure according to the embodiment. FIG. 9 is a diagram showing an example of a hardware configuration.

Hereinafter, the information processing apparatus according to the present application, the information processing method, and a mode for carrying out the information processing program (hereinafter, referred to as “the embodiment”) will be described in detail with reference to the drawings. The information processing apparatus, information processing method, and information processing program according to the present application are not limited by this embodiment. Further, in the following embodiments, the same parts are designated by the same reference numerals, and duplicate description is omitted.

[1. Information processing method overview]
First, with reference to FIG. 1, an outline of an information processing method performed by the information processing apparatus according to the embodiment will be described. FIG. 1 is an explanatory diagram showing an outline of an information processing method according to an embodiment. In FIG. 1, a case where the cause of withdrawal of the target service (the reason why the user stops using the service) is estimated by machine learning will be described as an example.

As shown in FIG. 1, the information processing system 1 includes a terminal device 10, an information processing device 100, and an information providing device 200. The terminal device 10, the information processing device 100, and the information providing device 200 are connected to each other via a network N (see FIG. 3) so as to be able to communicate with each other by wire or wirelessly.

The terminal device 10 is a smart device such as a smartphone or tablet used by the user U, and communicates with an arbitrary server device via a wireless communication network such as 4G (Generation) or LTE (Long Term Evolution). It is an information processing device that can be used. Further, the terminal device 10 is a screen of a liquid crystal display or the like and has a screen having a touch panel function. Accepts various operations. The operation performed on the area of the screen on which the content is displayed may be an operation on the content. Further, the terminal device 10 may be an information processing device such as a desktop PC (Personal Computer) or a notebook PC as well as a smart device.

The information providing device 200 is an information processing device that provides various services to the terminal device 10 of each user U, and is realized by a server device, a cloud system, or the like. The number of information providing devices 200 may be plurality. In addition, the information providing device 200 provides a search function for searching content such as a web page, a website, and / or an image file existing on the Internet in response to a search request from the terminal device 10 of the user U. It may be a server (search engine). For example, the information providing device 200 may provide a search site including a portal site or the like. Here, when the terminal device 10 or the like accesses the information providing device 200, the information providing device 200 manages the state between the information providing device 200 and the web browser from the terminal device 10 or the like via the web browser. You may acquire the communication protocol, or the HTTP Cookie which is the data stored in the web browser used therein. Further, the information providing device 200 may be a relay server (Proxy), a redirector (Redirector), or the like.

The information processing device 100 is an information processing device that analyzes the cause of withdrawal of each user U, and is realized by a server device, a cloud system, or the like. The information processing device 100 may be one of the information providing devices 200. That is, the information processing device 100 may also be an information processing device that provides services.

For example, the user U uses the terminal device 10 to use various services provided by each information providing device 200 (step S1).

Each information providing device 200 aggregates the accesses from the terminal device 10 of each user U and generates an access log of each user U (step S2).

The information processing device 100 acquires the access log of each user U from each information providing device 200 (step S3).

At this time, the information processing apparatus 100 may acquire only the access log of the user U who has consented to the use of the access log in advance. For example, the information processing device 100 may acquire only the access log of the user U registered in advance.

When the information processing device 100 itself provides a service to the terminal device 10 of each user U, the information processing device 100 can generate an access log of each user U regarding the service provided by the information processing device 100. ..

The information processing device 100 creates service usage transition sequence data related to service usage in the same session from the access log of each user U (step S4).

Actually, a series of accesses of the user U within a predetermined period may be used as the use of the service in the same session. In addition to the access performed within the predetermined period, the service may be used in the same session even when the link is followed, for example.

Further, in addition to the access from the same browser and the same application, the access from different browsers and different applications in the same time zone may be used as the service in the same session. That is, in addition to the case where the services are continuously accessed from the same browser or the like, the access that is continuously performed within a predetermined period may be used as the service in the same session.

Further, the service may be any service. For example, the service may be web content or an application (hereinafter referred to as an application). Further, the service may be a function or data of a program provided to the terminal device 10 via an API (Application Programming Interface). Further, the unit of service may be for each web content, each application, each domain, each subdomain, or the like. Further, the information processing apparatus 100 may consider the behavior of the user to be analyzed as a service at an arbitrary particle size, and create service use transition sequence data showing the transition for each service.

Here, the information processing device 100 displays access logs at an arbitrary particle size, such as the order of domains accessed by user U, the order of applications accessed by user U, or the order of web contents accessed by user U. In summary, service usage transition sequence data indicating the contents of the services used by the user U and the order thereof is created.

In the example shown in FIG. 1, the service usage transition sequence data is the "first search site", "local summary site", "first map WEB", "first map application", "first search site", "1st search site". It includes "travel site" and "accommodation reservation site", and indicates that the use of the service is transitioning in this order.

The "first search site" indicates a search service (search engine) provided via a Web browser. In addition, the "local summary site" indicates a summary site for information originating in the area, tourist information, etc. accessed by the user U using a Web browser. Further, the "first map WEB" indicates a map service provided via a Web browser. Further, the "first map application" indicates a map service provided via a map-dedicated application. Further, the "travel site" indicates a travel-related site accessed by the user U using a Web browser. Further, the "accommodation reservation site" indicates a reservation site such as an accommodation facility accessed by the user U using a Web browser.

In generating such service usage transition sequence data, a user ID commonly used by the user U across each service may be used. For example, a table that converts the user IDs individually used by the user U into a common user ID may be used in each service. Further, the user ID may be, for example, identification information of the terminal device 10.

The information processing device 100 inputs service usage transition sequence data into the model, makes the user U learn the intention (purpose of use) of using the service in the same session, and based on the learning result, the withdrawal period (Absence). A model for predicting Time) is created (step S5). Actually, the model for predicting the withdrawal period may be learned from the beginning without learning the purpose of use.

At this time, the information processing apparatus 100 uses the withdrawal period as correct answer data and causes the model to learn by deep learning. For example, the sequence data shows the service used by the user U and the order used by the user U. Therefore, the information processing apparatus 100 learns the model so as to output the withdrawal period when the service and the order are input. Then, the information processing apparatus 100 generates and outputs the predicted withdrawal period based on the model for predicting the withdrawal period.

The information processing device 100 sets a certain percentage of the training data as test data in advance, inputs all the service usage transition sequence data of the test data into the model, and the predicted withdrawal period is long. Sort in order and estimate the cause of withdrawal of the target service (step S6). In addition to the test data, the learning data may be input to predict the withdrawal period.

Here, the information processing apparatus 100 defines an interval (interval) from the end of use of the target service to the start of use of the next target service as the withdrawal period. The withdrawal period is not limited to the time, and may be, for example, the number of services (number of contents) used by the user U from the time of using the latest target service to the time of using the next target service. That is, the withdrawal period is the withdrawal period information indicating the period from the use of the target service by the user U to the re-use of the target service.

Further, in consideration of the session, the usage sequence may be from the time of using the latest target service to the time of using the target service in the next session. That is, the service usage transition sequence data may be created with one cycle from the time when the latest target service is used to the time when the target service is used in the next session.

Unless the user U uses the service next time, the withdrawal period will continue to increase. If the user U never uses the service again, the withdrawal period approaches infinity. In this case, an upper limit may be set for the withdrawal period. For example, when the withdrawal period reaches a predetermined threshold value, the information processing device 100 fixes the withdrawal period to the threshold value and determines that the user U has completely withdrawn from the service (has stopped using the service). You may do so. Further, the binary classification may be performed depending on whether the user U is not using the service next time or other than that.

In addition, if there is a period during which the service cannot be physically used regardless of the circumstances on the user U side such as service malfunction or maintenance, that period may be excluded from the withdrawal period. However, such frequent occurrences of defects and maintenance may cause the user U to be dissatisfied and stop using the service, and therefore may be positively included in the withdrawal period. In this case, for example, as a breakdown of the withdrawal period, the withdrawal period due to the factor on the user U side and the withdrawal period due to the factor on the service side (defect, maintenance, etc.) may be shown.

In addition, the withdrawal period is an index indicating whether or not the user U is satisfied / attractive with the service. For example, when the withdrawal period is short compared to the relative / predetermined threshold value, the user U is actively using the service, and the user U is satisfied / attractive with the service. It is presumed that you are feeling. On the contrary, when the withdrawal period is long, it is presumed that the user U does not use the service very much and the user U is not dissatisfied / attractive with the service. That is, the withdrawal period can also be index information that serves as an index of the user's impression of the target service.

Therefore, the information processing apparatus 100 can determine that the cause of withdrawal of the target service is included in the service use transition sequence data having a long withdrawal period of the target service. By examining the service usage transition sequence data, the cause of withdrawal from the target service can be estimated and the service can be improved.

Here, regarding the service usage transition sequence data, the sequence pattern and the span of the withdrawal period are different between the service used on a daily basis and the service used at a specific time of the year such as Bon Festival / New Year / Golden Week (GW). different.

Also, even with the same map application, there are cases where it is used for the purpose of checking the weather forecast such as rain cloud radar and when it is used for the purpose of checking traffic information such as traffic jam prediction during consecutive holidays, the sequence pattern and the span of the withdrawal period. Is different.

Therefore, in the present embodiment, after first specifying the purpose of use for each sequence, a model for predicting the withdrawal period of the target service is created next.

[1-1. Purpose of use prediction model]
The purpose-of-use prediction model will be described with reference to FIG. 2A. FIG. 2A is an explanatory diagram showing an outline of a purpose-of-use prediction model.

In the present embodiment, the information processing apparatus 100 inputs service use transition sequence data into the model and makes the model learn about the purpose of use of the sequence.

Any type of model can be adopted as the model. For example, the information processing apparatus 100 may use an SVM (Support Vector Machine) or a DNN (Deep Neural Network) as a model. Here, the DNN may be a CNN (Convolutional Neural Network) or an RNN (Recurrent Neural Network). Further, the RNN may be an LSTM (Long short-term memory) or the like. That is, any type of model can be adopted as the model. Further, the model may be a model realized by combining a plurality of models, for example, a model in which a CNN and an RNN are combined.

Further, the information processing apparatus 100 regards the behavior of the user to be analyzed as a service at an arbitrary granularity, creates service use transition sequence data indicating the transition for each service, and features (service) indicated by the created service use transition sequence data. A model may be generated from the characteristics of the transition mode of use).

In this embodiment, a case where LSTM is used will be described as an example of the model. LSTM is a model for time series data (sequential data).

First, the information processing device 100 inputs the service usage transition sequence data into the LSTM, and makes the model learn for each service included in the service usage transition sequence data so as to predict the service to be used next to the service. .. That is, the model is trained to seek a service that has a high probability of being used next to that service.

Here, when the service "first search site" is input to the LSTM, the information processing device 100 causes the model to learn to predict the service "local summary site" as the service to be used next to the service.

In fact, in the input service usage transition sequence data, the service "local summary site" is used next to the service "first search site", so this prediction is presumed to be correct.

In the same procedure, the information processing apparatus 100 trains the model to predict the service to be used next to the service after the service "local summary site".

As described above, in the example shown in FIG. 2A, "first search site", "local summary site", and "first map web". .. .. Enter the "value indicating each service" in the order in which each service is used. At this time, the information processing apparatus 100 performs pre-training (pre-learning) of the model so as to predict the next service of the input service. For example, the model is pre-trained to output a "local summary site" when the "first search site" is entered. In reality, pre-training does not have to be performed.

Ultimately, such a model learns the user's impression (purpose of use, whether or not they are satisfied, or the degree of use, etc.) when using the target service in the sequence data.

A service with a purpose of use close to each service may be separately obtained by the information processing apparatus 100 quantifying the purpose of use of each service.

For example, the information processing apparatus 100 uses a learning method such as a vector representation method (for example, Word2Vec) or a similar method to make the model learn the features of the service use transition sequence data. Then, the information processing apparatus 100 uses the trained model to generate a vector for each service. The information processing device 100 visualizes and analyzes the vector for each service, and estimates the service for the purpose of use close to each service.

Further, a distributed expression (word embedding) indicating the purpose of use of the sequence is generated at the upper right end of the RSTM model shown in FIG. 2A. In the distributed representation, a word is represented by a high-dimensional (about 200-dimensional) real number vector. In the distributed representation, words with similar meanings correspond to close vectors. Here, the distributed representation expresses the purpose of use of the sequence.

[1-2. Withdrawal period prediction model]
The withdrawal period prediction model will be described with reference to FIG. 2B. FIG. 2B is an explanatory diagram showing an outline of the withdrawal period prediction model. It should be noted that the withdrawal period prediction model may be learned from the beginning without learning the purpose of use prediction model.

The information processing device 100 inputs all the service usage transition sequence data learned by the above-mentioned usage purpose prediction model into the model. The model is, for example, LSTM. However, RSTM is only an example.

As a result, the information processing apparatus 100 inputs the service usage transition sequence data into the model and calculates the squared error. Then, the information processing apparatus 100 learns the model so as to reduce the square error.

The squared error is an error between the actual value and the predicted value by the model. That is, the difference between the actual withdrawal period and the withdrawal period predicted by the model is shown.

The information processing device 100 sets a certain percentage of the training data as test data in advance, inputs all the service usage transition sequence data of the test data into the model, and the predicted withdrawal period is long. Sort in order. Moreover, you may use the data used for learning as the input data. For example, the information processing apparatus 100 outputs service use transition sequence data in order from the one with the longest withdrawal period. At this time, the information processing apparatus 100 may limit the number of output service use transition sequence data to a predetermined number. Further, the information processing apparatus 100 sorts (sorts) all the input service usage transition sequence data in ascending / descending order or merges (merges) duplicate data based on the predicted withdrawal period. You may do it. Further, the information processing apparatus 100 may compare the output of the service use transition sequence data in order from the one with the longest withdrawal period with the one that outputs the service use transition sequence data in order from the one with the shortest withdrawal period. As a result, the characteristics of the transition sequence group having a long withdrawal period (high probability of withdrawal), that is, the transition sequence group including the cause of withdrawal are revealed. As a result, it is possible to interpret that there is a cause of withdrawal in this transition sequence group and implement service improvement.

As described above, in the example shown in FIG. 2B, the information processing apparatus 100 is output when each service of sequence data is sequentially input to the pre-trained model as in the example shown in FIG. 2A. The squared error between the predicted value and the actual value is calculated, and the model is retrained so that the squared error becomes small. In reality, pre-training does not have to be performed.

As a result, the information processing apparatus 100 can train the model for estimating the withdrawal period from the sequence data.

[1-3. An example of a fine-grained service usage transition sequence]
Up to this point, the case of inputting coarse-grained service usage transition sequence data such as transitions to different sites has been described, but fine-grained service usage transition sequence data such as transitions to each function of one application has been input. The case can be processed in the same manner.

A case where fine-grained service usage transition sequence data is input will be described with reference to FIG. 2C. FIG. 2C is an explanatory diagram showing a transition sequence to each function of one application.

As shown in FIG. 2C, the learned sequence pattern is input only for the transition sequence to each function in the map service "first map" in which the "gourmet icon" is included in the sequence. ..

In the example shown in FIG. 2C, this sequence pattern starts with "starting the first map application" for the map service "first map", then "map search", "result screen scroll", "gourmet icon", and "result screen". Indicates that the transition is in the order of "scroll", "gourmet detail display", and "telephone".

By doing so, it is possible to estimate the purpose of use of each function of one application and find out the cause of withdrawal in the transition sequence of each function. For example, if the sequence pattern with a long withdrawal period is characterized by the use of a certain function, the user U may find it difficult to use the function as the times change due to the lack of creation of the application. It is presumed that he became dissatisfied with the function. This will clarify the issues for improving the service.

In this embodiment, the cause of withdrawal from the target service is estimated by deep learning. This allows the model to derive the sequence of usage that the user leaves most.

[2. Information processing system configuration example]
Next, the configuration of the information processing system 1 including the information processing device 100 according to the embodiment will be described with reference to FIG. FIG. 3 is a diagram showing a configuration example of the information processing system 1 according to the embodiment. As shown in FIG. 3, the information processing system 1 according to the embodiment includes a terminal device 10, an information processing device 100, and an information providing device 200. These various devices are connected via network N so as to be communicable by wire or wirelessly. The network N is, for example, a LAN (Local Area Network) or a WAN (Wide Area Network) such as the Internet.

Further, the number of each device included in the information processing system 1 shown in FIG. 3 is not limited to that shown in the figure. For example, in FIG. 3, for simplification of the illustration, only one terminal device 10 and one information providing device 200 are shown, but this is merely an example and is not limited to two or more. May be good.

The terminal device 10 is an information processing device used by the user U. For example, the terminal device 10 is a smart device such as a smartphone or tablet terminal, a feature phone, a PC (Personal Computer), a PDA (Personal Digital Assistant), a car navigation system, a wearable device such as a smart watch or a head-mounted display (Wearable Device). , Smart glasses, etc.

Further, the terminal device 10 includes a wireless communication network such as LTE (Long Term Evolution), 4G (4th Generation), 5G (5th Generation: 5th generation mobile communication system), Bluetooth (registered trademark), and wireless LAN (Local). It can be connected to the network N via short-range wireless communication such as Area Network) and communicate with the information processing device 100 and the information providing device 200.

The information processing device 100 and the information providing device 200 are, for example, a PC, a server, a mainframe, a workstation, or the like. The information processing device 100 and the information providing device 200 may be realized by cloud computing.

[3. Information processing device configuration example]
Next, the configuration of the information processing apparatus 100 according to the embodiment will be described with reference to FIG. FIG. 4 is a diagram showing a configuration example of the information processing device 100 according to the embodiment. As shown in FIG. 4, the information processing device 100 includes a communication unit 110, a storage unit 120, and a control unit 130.

(Communication unit 110)
The communication unit 110 is realized by, for example, a NIC (Network Interface Card) or the like. Further, the communication unit 110 is connected to the network N (see FIG. 3) by wire or wirelessly.

(Memory unit 120)
The storage unit 120 is realized by, for example, a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory (Flash Memory), or a storage device such as a hard disk or an optical disk. As shown in FIG. 4, the storage unit 120 has a user database 121, a sequence database 122, and a withdrawal period database 123.

(User database 121)
The user database 121 stores various information about the user U. For example, the user database 121 stores various information such as the attributes of the user U. FIG. 5 is a diagram showing an example of the user database 121. In the example shown in FIG. 5, the user database 121 has items such as "user ID (Identifier)", "age", "gender", "home", "work location", "service", and "access log". Has.

The "user ID" indicates identification information for identifying the user U. Further, "age" indicates the age of the user U identified by the user ID. The "age" may be the specific age of the user U identified by the user ID, such as 35 years old. Further, "gender" indicates the gender of the user U identified by the user ID.

Further, "home" indicates the location information of the home of the user U identified by the user ID. In the example shown in FIG. 5, "home" is illustrated with an abstract code such as "LC11", but may be latitude / longitude information or the like. Further, for example, "home" may be an area name or an address.

Further, the "work location" indicates the location information of the work location (school in the case of a student) of the user U identified by the user ID. In the example shown in FIG. 5, the “work location” is illustrated by an abstract code such as “LC12”, but may be latitude / longitude information or the like. Further, for example, the "work location" may be a region name or an address.

Further, "service" indicates the service currently used by the user U identified by the user ID. The number of "services" may be plural. That is, one user U may use a plurality of services at the same time.

Further, the "access log" is an access log of the user U identified by the user ID, and indicates the access log of the user U acquired from each information providing device 200.

For example, in the example shown in FIG. 5, the age of the user U identified by the user ID "U1" is "20's", and the gender is "male". Further, for example, the user U identified by the user ID "U1" indicates that his / her home is "LC11". Further, for example, the user U identified by the user ID "U1" indicates that the work location is "LC12". Further, for example, it indicates that the user U identified by the user ID "U1" is using "service # 11" and the access log is recorded in "log # 1".

Here, in the example shown in FIG. 5, abstract values such as "U1", "LC11", "LC12", "service # 11", and "log # 1" are used for illustration, but "U1" and "U1", " It is assumed that information such as specific character strings and numerical values is stored in "LC11", "LC12", "service # 11", and "log # 1". Hereinafter, abstract values may be illustrated also in figures relating to other information.

The user database 121 is not limited to the above, and various information may be stored depending on the purpose. For example, the user database 121 may store various information about the terminal device 10 of the user U. In addition, the user database 121 has attributes such as demographic (demographic attribute), psychographic (psychological attribute), geographic (geographical attribute), and behavioral (behavioral attribute) of user U. May remember information about. For example, the user database 121 includes name, family structure, occupation, position, income, qualification, living style (detached house, apartment, etc.), presence / absence of a car, commuting time, commuting route, commuter pass section (station). , Routes, etc.), Frequently used stations (other than the nearest station to your home / work location), Behavior history (Frequently used location information), Lessons (location, time zone, etc.), Hobbies, interests, lifestyle, etc. Information may be stored. In addition, the user database 121 may store information about a search query (search keyword) entered by the user U into a search engine or the like.

(Sequence database 122)
The sequence database 122 stores various information related to the service usage transition sequence data created from the access log related to the service usage in the same session of the user U. FIG. 6 is a diagram showing an example of the sequence database 122. In the example shown in FIG. 6, the sequence database 122 has items such as "user ID", "sequence ID", and "service use transition sequence".

The "user ID" indicates identification information for identifying the user U. Further, the "sequence ID" indicates identification information for identifying the service use transition sequence data. Further, the "service usage transition sequence" indicates each service included in the service usage transition sequence data and the usage order thereof. The information indicating each service may be identification information for identifying each service.

Further, the "service usage transition sequence" includes items such as "service" and "use date and time". “Service” indicates a service used by user U. Further, the "use date and time" indicates the date and time when the user U used the service. The "use date and time" is information indicating the order in which the user U used the service (for example, "1, 2, 3, ...", "A, B, C, ..."). There may be.

For example, in the example shown in FIG. 6, the service usage transition sequence data identified by the sequence ID "SQ1" is obtained from the access log relating to the use of the service in the same session of the user U identified by the user ID "U1". Indicates that it was created. Further, in the service usage transition sequence data identified by the sequence ID "SQ1", the services used by the user U are the "first search site", the "local summary site", the "first map WEB", and the "first". 1 Indicates that the transition is in the order of "map application", "first search site", "travel site", and "accommodation reservation site".

The sequence database 122 is not limited to the above, and various information may be stored depending on the purpose. For example, the sequence database 122 may store information regarding the usage date and time of each service included in the service usage transition sequence data. In addition, the sequence database 122 may store information regarding the usage location of each service included in the service usage transition sequence data.

(Withdrawal period database 123)
The withdrawal period database 123 stores various information regarding the withdrawal period (Absence Time) of the target service. FIG. 7 is a diagram showing an example of the withdrawal period database 123. In the example shown in FIG. 7, the withdrawal period database 123 has items such as “sequence ID”, “target service”, and “withdrawal period”.

The “sequence ID” indicates identification information for identifying the service usage transition sequence data. Further, the “target service” indicates a target service included in the service usage transition sequence data. The “withdrawal period” indicates the withdrawal period of the target service.

For example, in the example shown in FIG. 7, it is shown that the withdrawal period of the target service “first map application” in the service use transition sequence data identified by the sequence ID “SQ1” is the “withdrawal period # 1”.

The withdrawal period database 123 is not limited to the above, and various information may be stored depending on the purpose. For example, the withdrawal period database 123 may store information regarding the cause of withdrawal of the target service (the reason why the user stops using the service). In addition, the withdrawal period database 123 may store information regarding the square error.

(Control unit 130)
Returning to FIG. 4, the description will be continued. The control unit 130 is a controller, and is an information processing device 100 using, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or the like. It is realized by executing various programs (corresponding to an example of an information processing program) stored in an internal storage device using a storage area such as a RAM as a work area. In the example shown in FIG. 4, the control unit 130 includes a reception unit 131, an acquisition unit 132, a creation unit 133, a learning unit 134, an analysis unit 135, and a provision unit 136.

(Reception Department 131)
The reception unit 131 receives the registration of various information about each user U from the terminal device 10 of each user U via the communication unit 110, and stores the registration in the user database 121.

(Acquisition unit 132)
The acquisition unit 132 acquires the access log of each user U from each information providing device 200 via the communication unit 110.

(Creation unit 133)
The creation unit 133 creates service usage transition sequence data related to service usage in the same session from the acquired access log of each user U.

Further, when the terminal device 10 of each user U is using the service provided by the information processing device 100, the creation unit 133 automatically obtains an access log related to the service provided by the information processing device 100. You may create it.

In addition, the creation unit 133 may create service usage transition sequence data including the use of a predetermined target service and the usage mode of another service satisfies a predetermined condition. For example, service usage transition sequence data for learning is created by extracting only sequences that include seasonal services (depending on seasonal factors).

The creation unit 133 may be integrated with the acquisition unit 132. For example, the access log acquired by the acquisition unit 132 may be used as it is as the service usage transition sequence data. Further, it can be considered that the service usage transition sequence data has been acquired by the creation unit 133 creating the service usage transition sequence data.

(Learning Department 134)
The learning unit 134 inputs the service usage transition sequence data into the model, and makes the user U learn the intention (purpose of use) of using the service in the same session. In addition, the learning unit 134 creates a model for predicting the withdrawal period based on the learning result. At this time, the learning unit 134 causes the model to learn the withdrawal period as correct answer data. Although not shown, the learning unit 134 includes a calculation unit 134A and a generation unit 134B.

The calculation unit 134A calculates the squared error of the target service based on the service usage transition sequence data.

The generation unit 134B creates a model for predicting the withdrawal period of the target service based on the square error of the target service. In addition, the generation unit 134B generates the withdrawal period information of the target service based on the model for predicting the withdrawal period.

The model is, for example, RSTM (Long short-term memory) or the like. However, RSTM is only an example. Any type of model can be adopted as the model.

The learning is, for example, deep learning (deep learning) using a deep neural network (DNN). Data mining and other machine learning algorithms may also be used. The learning unit 134 learns the model by the various learning methods described above.

(Analysis Department 135)
The analysis unit 135 inputs all the learned service usage transition sequence data into the model and analyzes the cause of withdrawal of the target service. Although not shown, the analysis unit 135 includes a rearrangement unit 135A and an estimation unit 135B.

The sorting unit 135A sorts all the input service usage transition sequence data in descending order of the predicted withdrawal period. For example, the sorting unit 135A outputs the service usage transition sequence data in order from the one with the longest withdrawal period. That is, the sorting unit 135A may be an output unit. At this time, the sorting unit 135A may limit the number of output service use transition sequence data to a predetermined number. As a result, the characteristics of the transition sequence group having a long withdrawal period (high probability of withdrawal), that is, the transition sequence group including the cause of withdrawal are revealed.

The estimation unit 135B estimates the cause of withdrawal included in the service usage transition sequence data. For example, the estimation unit 135B estimates a service that may be a cause of withdrawal from the target service among the services included in the service usage transition sequence data.

(Providing Department 136)
The providing unit 136 provides information on the cause of withdrawal of the target service to the outside via the communication unit 110. For example, the provider 136 may provide a report on the cause of withdrawal of the target service in response to an external request. The providing unit 137 may function as a display control unit that displays information on the cause of withdrawal from the target service on a display device such as a display.

Further, the providing unit 136 may provide information regarding the service use transition sequence group having a high withdrawal probability to the outside via the communication unit 110. For example, the providing unit 136 may provide information on a service use transition sequence group having a high withdrawal probability to the outside in response to a request from the outside.

Further, the providing unit 136 may provide various services to the terminal device 10 of each user U via the communication unit 110.

[4. Processing procedure]
Next, the processing procedure by the information processing apparatus 100 according to the embodiment will be described with reference to FIG. FIG. 8 is a flowchart showing a processing procedure according to the embodiment. The processing procedure shown below is repeatedly executed by the control unit 130 of the information processing apparatus 100.

The control unit 130 of the information processing device 100 acquires the access log of each user U from each information providing device 200 via the communication unit 110 (step S101).

The control unit 130 creates service usage transition sequence data related to service usage in the same session from the access log of each user U (step S102).

The control unit 130 inputs service usage transition sequence data into the model, and makes the user U learn the intention (purpose of use) of using the service in the same session (step S103). Actually, the model for predicting the withdrawal period may be learned from the beginning without learning the purpose of use.

The control unit 130 inputs service usage transition sequence data into the model, and calculates a squared error (Squared Error), which is an error between the actual value and the predicted value by the model (step S104).

The control unit 130 learns the model so as to reduce the square error, and creates a model for predicting the withdrawal period (step S105).

The control unit 130 sets a certain percentage of the training data as test data in advance, inputs all the service usage transition sequence data of the test data into the model, and orders the predicted withdrawal period in the order of longest. Sort (step S106). As a result, the characteristics of the transition sequence group having a long withdrawal period (high probability of withdrawal), that is, the transition sequence group including the cause of withdrawal are revealed.

The control unit 130 estimates the cause of withdrawal of the target service from the transition sequence group having a long withdrawal period (high withdrawal probability) (step S107).

[5. Modification example]
The terminal device 10 and the information processing device 100 described above may be implemented in various different forms other than the above-described embodiment. Therefore, a modified example of the embodiment will be described below.

In the above embodiment, the service usage transition sequence data may be created from the access log related to the service usage in the same session of the user U having the same attribute. For example, since the service usage transition sequence is considered to differ depending on the attributes of the user U (gender / age group / residential area / family composition / hobbies, etc.), the service usage transition sequence data classified according to the attributes of the user U. May be created. As a result, a model is created for each attribute of the user U.

Further, in the above embodiment, the service use transition sequence data may be created from the access log related to the use of the service in the same session according to the situation (context) when the user U uses the service. For example, the cause of withdrawal from the service is considered to change depending on the situation of user U and changes in the environment (moving, purchasing a car, going on to school / employment, changing jobs, getting married, giving birth / childcare, etc.). Service usage transition sequence data classified according to the situation may be created. As a result, a model is created for each situation in which the user U is placed.

Further, in the above embodiment, a model for predicting the "withdrawal period" is created, but by changing the correct answer data from the "withdrawal period" to the "user's impression of the target service" by the same method, You may train a model that estimates "the impression of the user when the services are used in the order indicated by the sequence data".

For example, if the map service is used from a web search, the satisfaction level is high, but if the map service is used from shopping, it can be estimated that the impression is bad.

Further, for example, the information processing apparatus 100 acquires sequence data and a value indicating an impression when the target service is used in the sequence from the user by using a technique such as cloud sourcing. Here, the value indicating the impression is, for example, a value from 0 point (dissatisfied) to 10 points (satisfied). Then, the information processing apparatus 100 performs pre-training of the model and learning of satisfaction by the above-mentioned technique.

The model subjected to such learning processing can estimate the user's satisfaction with the target service in the sequence indicated by the sequence data from the sequence data.

In addition to the various evaluations entered by the user, the values indicating the impression include, for example, the facial expression and pulse of the user when the target service is used among the sequences shown by the sequence data using a camera or the like. It may be a value estimated from facial expressions and biometric information by acquiring various biometric information.

Further, the information processing apparatus 100 may use the above-mentioned withdrawal period as information suggesting an impression of the user. That is, the information processing apparatus 100 calculates a score indicating the impression of the user by using various known techniques for estimating the impression of the user, and learns the model so as to estimate the calculated score. May be good.

Further, in the above embodiment, the order is learned by inputting the information indicating the service according to the sequence, but the present invention is not limited to this. For example, information indicating the service and a numerical value indicating the order may be input to the model, and a non-recursive model such as CNN may be trained so as to output the withdrawal period and the impression of the user.

[6. effect〕
As described above, the information processing apparatus 100 according to the present application has an acquisition unit and a learning unit. The acquisition unit acquires sequence data indicating the transition of service usage of each user, including the usage of a predetermined target service. The learning unit trains the model to output the withdrawal period information indicating the period from the use of the target service by the user to the re-use of the target service from the sequence data. This makes it possible to find the cause of withdrawal more effectively.

The learning unit trains a model having an LSTM (Long short-term memory) configuration as a model. As a result, the cause of withdrawal from the analysis target service can be estimated based on the sequence data which is the time series data.

The acquisition unit acquires, as sequence data, data indicating a plurality of services used by the user and the order in which each service is used. The learning unit trains the model to output the withdrawal period information when the information indicating each service used by the user is input to the model in the order in which each service is used. This makes it possible to carry out specific learning for more effectively finding the cause of withdrawal.

When the learning unit inputs information indicating the service used by the user, the learning unit trains the model to output information indicating the service to be used next by the user, and then trains the model in the order in which each service is used. Train the model to output the withdrawal period information when inputting to the model. As a result, it is possible to perform pre-training (pre-learning) of the model regarding the purpose of using the target service.

The learning unit trains the model to output the number of other services used by the user from the use of the target service to the re-use of the target service as the withdrawal period information. As a result, as the withdrawal period, the number of contents used during that period can be used instead of the time.

The learning unit trains the model to output the time elapsed from the use of the target service by the user to the re-use of the target service as the withdrawal period information. As a result, as the withdrawal period, the time elapsed during that period can be used.

The acquisition unit acquires sequence data that includes the use of a predetermined target service and that the usage mode of another service satisfies a predetermined condition. This makes it possible to filter the context. For example, it is possible to create sequence data for learning by extracting only sequences that include seasonal services.

Further, the information processing apparatus 100 according to the present application further includes a generation unit and an output unit. The generation unit generates the withdrawal period information corresponding to each sequence data by using the model trained by the learning unit. The output unit outputs sequence data in which the period indicated by the withdrawal period information satisfies a predetermined condition. As a result, sequence data can be output according to the withdrawal period.

The output unit outputs sequence data in order from the one with the longest period indicated by the withdrawal period information. As a result, the sequence data can be sorted in the order of the longest withdrawal period.

The output unit outputs a predetermined number of sequence data in order from the one with the longest period indicated by the withdrawal period information. Thereby, the number of output sequence data can be limited. As a result, only the sequence data having the longer withdrawal period can be targeted.

The acquisition unit acquires sequence data for each function of one application as a service as sequence data. As a result, it is possible to target sequence data with fine particle size. As a result, it is possible to estimate the cause of withdrawal of the function of one application.

From another point of view, the information processing apparatus 100 according to the present application has an acquisition unit and a learning unit. The acquisition unit acquires sequence data indicating the transition of service usage of each user, including the usage of a predetermined target service. The learning unit trains the model to output index information that is an index of the user's impression of the target service from the sequence data. This makes it possible to estimate the satisfaction / dissatisfaction of the user.

[7. Hardware configuration]
Further, the terminal device 10 and the information processing device 100 according to the above-described embodiment are realized by, for example, a computer 1000 having a configuration as shown in FIG. Hereinafter, the information processing device 100 will be described as an example. FIG. 9 is a diagram showing an example of a hardware configuration. The computer 1000 is connected to the output device 1010 and the input device 1020, and the arithmetic unit 1030, the primary storage device 1040, the secondary storage device 1050, the output I / F (Interface) 1060, the input I / F 1070, and the network I / F 1080 are bused. It has a form connected by 1090.

The arithmetic unit 1030 operates based on a program stored in the primary storage device 1040 or the secondary storage device 1050, a program read from the input device 1020, or the like, and executes various processes. The arithmetic unit 1030 is realized by, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or the like.

The primary storage device 1040 is a memory device that temporarily stores data used by the arithmetic unit 1030 for various calculations, such as a RAM (Random Access Memory). Further, the secondary storage device 1050 is a storage device in which data used by the calculation device 1030 for various calculations and various databases are registered, and is a ROM (Read Only Memory), an HDD (Hard Disk Drive), and an SSD (Solid). State Drive), flash memory, etc. The secondary storage device 1050 may be internal storage or external storage. Further, the secondary storage device 1050 may be a removable storage medium such as a USB memory or an SD (Secure Digital) memory card. Further, the secondary storage device 1050 may be a cloud storage (online storage), a NAS (Network Attached Storage), a file server, or the like.

The output I / F 1060 is an interface for transmitting information to be output to an output device 1010 that outputs various information such as a display, a projector, and a printer. For example, USB (Universal Serial Bus) or DVI. (Digital Visual Interface), HDMI (registered trademark) (High Definition Multimedia Interface), etc. are realized by standard connectors. The input I / F 1070 is an interface for receiving information from various input devices 1020 such as a mouse, keyboard, keypad, buttons, and a scanner, and is realized by, for example, USB.

Further, the output I / F 1060 and the input I / F 1070 may be wirelessly connected to the output device 1010 and the input device 1020, respectively. That is, the output device 1010 and the input device 1020 may be wireless devices.

Further, the output device 1010 and the input device 1020 may be integrated like a touch panel. In this case, the output I / F 1060 and the input I / F 1070 may also be integrated as input / output I / F.

The input device 1020 is, for example, an optical recording medium such as a CD (Compact Disc), a DVD (Digital Versatile Disc), a PD (Phase change rewritable Disk), a magneto-optical recording medium such as an MO (Magneto-Optical disk), or a tape. It may be a device that reads information from a medium, a magnetic recording medium, a semiconductor memory, or the like.

The network I / F 1080 receives data from another device via the network N and sends it to the arithmetic unit 1030, and also transmits the data generated by the arithmetic unit 1030 to the other device via the network N.

The arithmetic unit 1030 controls the output device 1010 and the input device 1020 via the output I / F 1060 and the input I / F 1070. For example, the arithmetic unit 1030 loads a program from the input device 1020 or the secondary storage device 1050 onto the primary storage device 1040, and executes the loaded program.

For example, when the computer 1000 functions as the information processing device 100, the arithmetic unit 1030 of the computer 1000 realizes the function of the control unit 130 by executing the program loaded on the primary storage device 1040. Further, the arithmetic unit 1030 of the computer 1000 may load a program acquired from another device via the network I / F 1080 onto the primary storage device 1040 and execute the loaded program. Further, the arithmetic unit 1030 of the computer 1000 may cooperate with other devices via the network I / F 1080 to call the functions and data of the program from other programs of the other devices and use them.

[8. others〕
Although the embodiments of the present application have been described above, the present invention is not limited to the contents of these embodiments. In addition, the above-mentioned components include those that can be easily assumed by those skilled in the art, those that are substantially the same, that is, those in a so-called equal range. Furthermore, the components described above can be combined as appropriate. Further, various omissions, replacements or changes of the components can be made without departing from the gist of the above-described embodiment.

Further, among the processes described in the above-described embodiment, all or a part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed can be performed. All or part of it can be done automatically by a known method. In addition, the processing procedure, specific name, and information including various data and parameters shown in the above document and drawings can be arbitrarily changed unless otherwise specified. For example, the various information shown in each figure is not limited to the illustrated information.

Further, each component of each of the illustrated devices is a functional concept, and does not necessarily have to be physically configured as shown in the figure. That is, the specific form of distribution / integration of each device is not limited to the one shown in the figure, and all or part of the device is functionally or physically dispersed / physically distributed in arbitrary units according to various loads and usage conditions. Can be integrated and configured.

For example, the information processing device 100 described above may be realized by a plurality of server computers, or may be realized by calling an external platform or the like by API (Application Programming Interface), network computing, or the like depending on the function. Can be changed flexibly.

Further, the above-described embodiments and modifications can be appropriately combined as long as the processing contents do not contradict each other.

Further, the above-mentioned "section, module, unit" can be read as "means" or "circuit". For example, the acquisition unit can be read as an acquisition means or an acquisition circuit.

1 Information processing system 10 Terminal equipment 100 Information processing equipment 110 Communication unit 120 Storage unit 121 User database 122 Sequence database 123 Withdrawal period database 130 Control unit 131 Reception unit 132 Acquisition unit 133 Creation unit 134 Learning unit 135 Analysis unit 136 Providing unit 200 Information provider

Claims (14)

An acquisition unit that acquires sequence data that indicates the transition of service usage by each user and includes the usage of a predetermined target service.
It is characterized by having a learning unit that trains a model to output withdrawal period information indicating a period from the use of the target service to the reuse of the target service from the sequence data. Information processing device. The information processing apparatus according to claim 1, wherein the learning unit trains a model having an LSTM configuration as the model. The acquisition unit acquires, as the sequence data, data indicating a plurality of services used by the user and the order in which each service is used.
The learning unit is characterized in that when information indicating each service used by the user is input to the model in the order in which each service is used, the model is trained to output the withdrawal period information. The information processing apparatus according to claim 1 or 2. When the learning unit inputs information indicating the service used by the user, the learning unit trains the model to output information indicating the service to be used next by the user, and then uses each service. The information processing apparatus according to claim 3, wherein the model is trained to output the withdrawal period information when the information is input to the model in the order of the above. The learning unit trains the model to output the number of other services used by the user from the use of the target service to the re-use of the target service as the withdrawal period information. The information processing apparatus according to any one of claims 1 to 4. The learning unit is characterized in that the learning unit trains the model to output the elapsed time from the use of the target service to the re-use of the target service as the withdrawal period information. The information processing apparatus according to any one of claims 1 to 4. The acquisition unit is any one of claims 1 to 6, which includes the use of the predetermined target service and acquires sequence data in which the usage mode of the other service satisfies a predetermined condition. The information processing device described in. Using the model trained by the learning unit, a generation unit that generates the withdrawal period information corresponding to each sequence data, and a generation unit.
The information processing apparatus according to any one of claims 1 to 7, further comprising an output unit for outputting sequence data in which the period indicated by the withdrawal period information satisfies a predetermined condition. The information processing apparatus according to claim 8, wherein the output unit outputs the sequence data in order from the one with the longest period indicated by the withdrawal period information. The information processing apparatus according to claim 9, wherein the output unit outputs a predetermined number of the sequence data in order from the one with the longest period indicated by the withdrawal period information. The information processing device according to any one of claims 1 to 10, wherein the acquisition unit acquires sequence data for each function of one application as a service as the sequence data. It is an information processing method executed by an information processing device.
An acquisition process that acquires sequence data that includes the use of a predetermined target service, which is sequence data indicating the transition of service usage of each user, and
It is characterized by including a learning step of training a model to output withdrawal period information indicating a period from the use of the target service to the reuse of the target service from the sequence data. Information processing method to do. An acquisition procedure for acquiring sequence data indicating the transition of service usage of each user, including the use of a predetermined target service, and
To make a computer execute a learning procedure for training a model to output withdrawal period information indicating a period from the use of the target service to the reuse of the target service from the sequence data. Information processing program. An acquisition unit that acquires sequence data that indicates the transition of service usage by each user and includes the usage of a predetermined target service.
An information processing device including a learning unit that trains a model to output index information that is an index of the impression of the user with respect to the target service from the sequence data.

Images