WO2021250987A1 - Dispositif de traitement d'informations, procédé de traitement d'informations et support d'enregistrement - Google Patents

Dispositif de traitement d'informations, procédé de traitement d'informations et support d'enregistrement Download PDF

Info

Publication number
WO2021250987A1
WO2021250987A1 PCT/JP2021/014700 JP2021014700W WO2021250987A1 WO 2021250987 A1 WO2021250987 A1 WO 2021250987A1 JP 2021014700 W JP2021014700 W JP 2021014700W WO 2021250987 A1 WO2021250987 A1 WO 2021250987A1
Authority
WO
WIPO (PCT)
Prior art keywords
learning
prediction
unit
teacher
processing
Prior art date
Application number
PCT/JP2021/014700
Other languages
English (en)
Japanese (ja)
Inventor
直樹 柴山
祐己 牧野
達也 物井
瑛 春日
優汰 西村
Original Assignee
株式会社プレイド
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社プレイド filed Critical 株式会社プレイド
Publication of WO2021250987A1 publication Critical patent/WO2021250987A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present invention relates to an information processing device or the like that performs learning processing and prediction processing of machine learning.
  • Patent Document 1 Conventionally, machine learning processing has been used for various predictions.
  • Patent Document 1 For example, there is a technique for predicting a sales index related to the sale of a product handled by a store based on the actual value of another product handled by another store (see Patent Document 1).
  • Patent Document 2 For the purpose of preventing the situation where debt cannot be collected as much as possible, it is possible for the target user to perform debt consolidation using a prediction model that learns the characteristics of user information about the debt consolidation person who performed debt consolidation.
  • Patent Document 2 There was a technique for predicting sex
  • the learning process and the prediction process are fixed, and the learning process and the prediction process that can be flexibly customized cannot be performed.
  • the information processing apparatus of the first invention is a template storage unit that stores a template having one or more learning parameters that are parameters related to learning processing of machine learning and one or more prediction parameters that are parameters related to prediction processing of machine learning.
  • a teacher source data storage unit that stores one or more teacher source data that is the source of teacher data, a target data storage unit that stores prediction target data that is prediction target data, and one or more that the template has. It is one or more teacher source data corresponding to one or more learning parameters, and the teacher data used for the learning process from each one or more teacher source data stored in the teacher source data storage unit.
  • the target data is obtained by using the learning unit that acquires the learning device by performing the learning process on one or more teacher data and the learning unit that has one or more prediction parameters of the template and the learning unit.
  • It is an information processing device including a prediction unit that performs prediction processing on the prediction target data stored in the storage unit and acquires the prediction result, and an output unit that outputs the prediction result.
  • learning processing and prediction processing when learning processing and prediction processing are performed by a machine learning algorithm, learning processing and prediction processing that can be flexibly customized can be performed using a template having various parameters to be used.
  • two or more templates are stored in the template storage unit in association with different administrator identifiers, and the administrator identifier is received.
  • the reception unit is further provided, the learning unit performs learning processing using one or more learning parameters corresponding to the administrator identifiers received by the reception unit, acquires a learning device, and the prediction unit is accepted by the reception unit.
  • This is an information processing device that performs prediction processing and acquires prediction results using one or more prediction parameters corresponding to the administrator identifier.
  • learning processing and prediction processing can be performed using different templates for each administrator who uses the information processing device.
  • a default template is stored in the template storage unit, and the learning unit is a template corresponding to the administrator identifier accepted by the reception unit. Does not exist in the template storage unit, the learning process is performed using one or more learning parameters of the default template, the learner is acquired, and the prediction unit corresponds to the administrator identifier accepted by the reception unit.
  • the information processing device performs prediction processing using one or more prediction parameters of the default template and acquires the prediction result.
  • the information processing apparatus of the fourth invention has two or more used to output prediction results of different objects in the template storage unit for any one of the first to third inventions.
  • the template is stored in association with the template identifier, further includes an identifier receiving unit that accepts the template identifier, and the learning unit is one or more of the templates identified by the template identifier accepted by the identifier receiving unit. It is one or more teacher source data corresponding to the learning parameter, and the teacher data used for the learning process is acquired from each one or more teacher source data stored in the teacher source data storage unit, and the teacher data is converted into one or more teacher data.
  • the learning process is performed to acquire the learning device, and the prediction unit uses one or more prediction parameters of the template identified by the template identifier received by the identifier receiving unit and the learning device acquired by the learning unit.
  • An information processing device that performs prediction processing on the prediction target data stored in the target data storage unit and acquires the prediction result.
  • the teacher source data has the time information indicating the time, and one or more learning parameters are the teacher source.
  • a teacher that includes a learning period parameter that specifies a period for selecting data, and the learning unit acquires one or more teacher data from one or more source data having information when the learning period parameter corresponds to the specified period. It is an information processing apparatus including a data acquisition means and a learning means for performing learning processing on one or more teacher data and acquiring a learning device.
  • the learning period parameter used in the teacher data selection process which is the pre-process of the learning process, can be specified by the template.
  • one or more learning parameters have method parameters related to an evaluation method for evaluating the learning device, and the learning unit. Performs learning processing of two or more different algorithms on one or more teacher data, and evaluates the learning means for acquiring two or more learners and the evaluation based on the method parameters for each of the two or more learners. It is provided with an evaluation means for acquiring evaluation results for each of two or more learning devices and a selection means for selecting one learning device based on the evaluation results, and the prediction unit uses one or more prediction parameters of the template.
  • the prediction target data is acquired from the target data storage unit in which one or more prediction target data which is the prediction target data is stored, and the prediction target data is predicted by using the learning device selected by the selection means. It is an information processing device that performs processing and acquires prediction results.
  • the method parameters used in the post-processing of the learning process can be specified by the template.
  • one or more prediction parameters include a prediction period parameter for specifying a prediction period, and the prediction unit predicts. It is an information processing device that acquires the prediction result of the period according to the period parameter.
  • the forecast period parameter used in the forecast process can be specified by the template.
  • one or more prediction parameters include a prediction timing parameter that specifies the timing for performing the prediction process, and the prediction unit. Is an information processing device that performs prediction processing and acquires prediction results at the timing specified by the prediction timing parameter.
  • the timing of prediction processing can be specified by the template.
  • the template has one or more post-processing parameters which are parameters for post-processing using the prediction result. It further comprises a post-processing unit that processes the prediction result and acquires output information using one or more post-processing parameters, the output unit in place of or in addition to the prediction result. It is an information processing device that outputs output information.
  • the post-processing parameters used in the post-processing after the prediction processing can be specified by the template.
  • the teacher data and the prediction target data include one or more dynamic attribute values of the user, and the teacher data is used.
  • the objective variable to have is a user category
  • the learning unit performs learning processing using one or more teacher data including one or more dynamic attribute values, acquires a learner, and the prediction unit has one or more.
  • the teacher data and the prediction target data include one or more static attribute values of the user, and the learning unit has one or more static attribute values. Learning processing is performed using one or more teacher data including the target attribute value, a learner is acquired, and the prediction unit uses one or more prediction target data including one or more static attribute values. It is an information processing device that performs prediction processing and acquires prediction results including user categories.
  • the teacher source data includes dynamic attribute values related to the operation when the user purchases the product
  • the template One or more learning parameters have a source parameter that identifies the source of the teacher source data, a learning period parameter that specifies the period for selecting the teacher source data, and a total that specifies the aggregation period for acquiring the dynamic attribute value. It has a period parameter and a prediction target parameter that specifies the prediction target data, and one or more prediction parameters that the template has has a prediction user selection parameter that identifies the prediction target user, and the learning unit uses the original parameter as the original parameter.
  • the teacher data acquisition means for calculating the aggregation result acquiring one or more teacher data having the calculated one or more aggregation results as explanatory variables, and using the prediction target parameter as the objective variable, and one or more teacher data.
  • the prediction unit calculates and calculates the aggregation result of one or more dynamic attribute values of the user corresponding to the prediction user selection parameter.
  • This is an information processing device that performs prediction processing and acquires prediction results using an explanatory variable group having the above aggregated results as explanatory variables and a learning device.
  • the post-processing parameter of the template has a threshold parameter indicating a threshold for user classification with respect to the twelfth invention
  • the post-processing unit has a threshold parameter. It further includes a post-processing unit that determines whether or not it corresponds to a specific user using the threshold parameter and acquires output information using the determination result, and the output unit replaces or predicts the prediction result. It is an information processing device that outputs output information in addition to the result.
  • learning processing and prediction processing that can be flexibly customized can be performed using a template.
  • a flowchart illustrating an example of the learning process Flow chart explaining an example of the teacher data acquisition process
  • a flowchart illustrating an example of the output processing A flowchart illustrating an operation example of the terminal device 5.
  • Diagram showing the teacher's original data management table Diagram showing the template management table The figure which shows the same screen example Diagram showing an example of the template Diagram showing the teacher data management table Diagram showing an example of the template
  • Embodiment 1 an information system including a server device that detects and outputs an attribute value of a visitor to a website (hereinafter, appropriately referred to as a “user”) in real time will be described.
  • an information system including a server device that automatically performs an action to a user whose attribute value satisfies the condition will be described.
  • an information system including a server device that performs an action for a specific user selected by the administrator will be described.
  • the server device which associates with the logged-in user by the Cookie ID of the non-logged-in user and acquires the attribute value of the user by using both the operation information before login and the operation information after login.
  • the information system to be logged in will be described.
  • FIG. 1 is a conceptual diagram of the information system A in the present embodiment.
  • the information system A includes one or more user terminals 1, a server device 2, and a management terminal 3.
  • the user terminal 1 and the management terminal 3 are, for example, so-called personal computers, tablet terminals, smartphones, and the like, and their types are not limited.
  • the server device 2 is, for example, an ASP server, a cloud server, or the like. However, the type of the server device 2 does not matter.
  • the user terminal 1 and the server device 2 can communicate with each other via a network such as the Internet. Further, the server device 2 and the management terminal 3 can communicate with each other via a network such as the Internet or a LAN.
  • FIG. 2 is a block diagram of the information system A in the present embodiment.
  • FIG. 3 is a block diagram of the server device 2.
  • the user terminal 1 includes a user storage unit 11, a user reception unit 12, a user processing unit 13, a user transmission unit 14, a user reception unit 15, and a user output unit 16.
  • the server device 2 includes a storage unit 21, a reception unit 22, a processing unit 23, a transmission unit 24, and an output unit 25.
  • the storage unit 21 includes a user information storage unit 211, a dynamic processing information storage unit 212, and an operation information storage unit 213.
  • the receiving unit 22 includes a login instruction receiving unit 221, an operation information receiving unit 222, and a selection instruction receiving unit 223.
  • the processing unit 23 includes a login processing unit 231, a response unit 232, an operation information storage unit 233, an attribute value acquisition unit 234, a thumbnail image acquisition unit 235, a determination unit 236, a condition processing execution unit 237, and an instruction user processing unit 238. ..
  • the attribute value acquisition unit 234 includes score calculation means 2341.
  • the transmission unit 24 includes a processing result transmission unit 241 and a user terminal transmission unit 242.
  • the output unit 25 includes an attribute value output unit 251 and a thumbnail image output unit 252.
  • the management terminal 3 includes a management storage unit 31, a management reception unit 32, a management processing unit 33, a management transmission unit 34, a management reception unit 35, and a management output unit 36.
  • the various types of information include, for example, a user identifier, a user terminal identifier, a user attribute value, and the like.
  • the user identifier is information that identifies the user, and is, for example, an ID.
  • the user identifier may be a telephone number, a credit card number, an email address, or the like.
  • the user terminal identifier is information that identifies the user terminal 1, and is, for example, a cookie ID, a session identifier, an IP address, a MAC address, or the like.
  • the attribute value of the user is, for example, a static attribute value such as the gender and age of the user.
  • the user reception unit 12 receives input of instructions, information, etc. from the user.
  • the instructions, information, and the like are, for example, operation information, login instructions, and the like.
  • the operation information is information related to the operation of the user's website.
  • the operation information is, for example, information indicating that a button has been pressed, information indicating that an anchor has been instructed, information on an operation for jumping to another page, information input in a field, or the like.
  • the operation information is, for example, "rightButtonON” (the right mouse button is pressed), "drug object A" (the object A is dragged), " ⁇ purchased product ID> 123 ⁇ quantity> 3" (123).
  • the operation information here is usually primitive operation information, but it is preferable that the information can be viewed by an administrator to be described later to determine the meaning and significance of the operation. That is, the data structure of the operation information, the particle size of the information, etc. do not matter.
  • the operation information is usually information that identifies the operation performed by the user, but may include information related to the processing performed by the server device 2 due to the user's operation.
  • the login instruction is a login instruction.
  • the login instruction has, for example, a user identifier.
  • the login instruction has, for example, a user identifier and a password.
  • a website may be called a web page.
  • the website is, for example, an EC site. However, the type of website does not matter.
  • the input means for instructions and information may be anything, such as a touch panel, keyboard, mouse, or menu screen.
  • the user reception unit 12 can be realized by a device driver for input means such as a touch panel or a keyboard, control software for a menu screen, or the like.
  • the user processing unit 13 performs various processes.
  • the various processes include, for example, a process of changing the instructions and information received by the user reception unit 12 into instructions and information of a structure to be transmitted, and a process of changing to a structure of outputting the information received by the user reception unit 15. And so on.
  • the user transmission unit 14 transmits various information, instructions, and the like.
  • the various information and instructions are, for example, operation information, login instructions, and the like.
  • the user transmission unit 14 usually transmits information, instructions, and the like to the server device 2.
  • the user transmission unit 14 may transmit the attribute value of the user in addition to the operation information.
  • the user attributes are stored in the user storage unit 11, and are, for example, the gender, age, and the like of the user.
  • the attribute of the user transmitted here is, for example, a static attribute value.
  • the user transmission unit 14 may transmit information, instructions, or the like to a second server device (not shown). In such a case, operation information or the like is transmitted from the second server device to the server device 2.
  • the second server device will be described later.
  • the user receiving unit 15 receives various information.
  • the various information is, for example, a processing result and a login processing result.
  • the processing result is information regarding the processing result in the response unit 232, which will be described later.
  • the processing result is, for example, a destination web page, a panel on which the result of a product purchase instruction is described, an error message, or the like.
  • the result of the login process is information indicating whether the login process was successful or unsuccessful, information on the web page after login, and the like.
  • the user output unit 16 outputs various information.
  • the various types of information are, for example, information that has been changed to a structure that is received by the user receiving unit 15 and output by the user processing unit 13, and is, for example, a processing result and a login processing result.
  • the output is usually a display on a display, but is projected by a projector, printed by a printer, sound output, transmitted to an external device, stored in a recording medium, other processing devices and others. It may be considered that the concept includes the delivery of the processing result to the program or the like.
  • Various information is stored in the storage unit 21 that constitutes the server device 2.
  • the various types of information are, for example, user information described later, dynamic processing information described later, operation information, and information of arithmetic expressions for calculating a score.
  • the user information storage unit 211 stores two or more user information.
  • User information is information about a user and has one or more attribute values.
  • the attribute value of 1 or more is usually a static attribute value, but may include a dynamic attribute value.
  • the static attribute value is usually an attribute value that does not change, but it may be considered as an attribute value that does not change from moment to moment.
  • the static attribute value is, for example, a name, age, address, telephone number, credit card number, e-mail address, user terminal identifier, user identifier, password, or the like.
  • the user terminal identifier is information that identifies the user terminal 1, and is, for example, a cookie ID, a session identifier, an IP address, a MAC address, or the like.
  • the user identifier is information that identifies the user, and is, for example, an ID.
  • the user identifier may be a telephone number, a credit card number, an email address, or the like.
  • the dynamic attribute value is an attribute value that can be dynamically changed by a user operation or the like.
  • Dynamic attribute values include, for example, real-time dynamic attribute values and historical information utilization dynamic attribute values.
  • the real-time dynamic attribute value is an attribute value that changes from moment to moment in real time.
  • the real-time dynamic attribute value is, for example, the staying time of the web page currently viewed by the user, the number of web pages viewed during the current stay, and the like.
  • the dynamic attribute value using historical information is a dynamic attribute value acquired by using the history of operation information at the time of a past visit.
  • the historical information usage dynamic attribute value is, for example, the number of purchases, the purchase price, the total purchase price, the average staying time, the average number of PVs, the number of visits, the score described later, and the like.
  • the dynamic processing information storage unit 212 stores one or more dynamic processing information.
  • the dynamic processing information has a condition and a processing identifier that identifies the processing to be executed when the condition is satisfied.
  • the dynamic processing information may further include information for specifying the processing timing.
  • the condition is a condition for determining that the processing corresponding to the processing identifier is performed.
  • the condition is a condition related to one or more attribute values of the user. It is preferable that the condition is a condition using one or more dynamic attribute values. Further, it is preferable that the condition is a condition using one or more real-time dynamic attribute values or one or more history information utilization dynamic attribute values.
  • the condition may be, for example, " ⁇ age> 20s, ⁇ number of purchases> 5 times or more, ⁇ score> 70 or more".
  • the processing identifier is an ID, a function name, a method name, a program address corresponding to the processing, and the like.
  • the processing identifier may be an executable program.
  • the processing identifier may be any information as long as it is information for executing the processing corresponding to the condition.
  • the operation information storage unit 213 stores one or more operation information for each user.
  • the operation information storage unit 213 stores one or more operation information in association with the user identifier.
  • the received operation information and the operation information stored in the operation information storage unit 213 may be different.
  • the data structure or the like may be different between the operation information received by the operation information receiving unit 222 and the operation information stored in the operation information storage unit 213.
  • the operation information received by the operation information receiving unit 222 is primitive operation information (for example, "rightButtonON")
  • the operation information stored in the operation information storage unit 213 is information on which the meaning and significance of the operation can be determined. (For example, " ⁇ page was displayed") may be used.
  • the processing unit 23 configures the operation information stored in the operation information storage unit 213 by using the received operation information.
  • the receiving unit 22 receives various information, instructions, and the like.
  • the various information and instructions are, for example, login instructions, operation information, and selection instructions.
  • the login instruction receiving unit 221 receives a login instruction from the user terminal 1.
  • the login instruction is a login instruction.
  • the operation information receiving unit 222 receives one or more operation information from the user terminal 1 of the user who is a visitor to the website.
  • the operation information receiving unit 222 does not need to receive the operation information directly from the user terminal 1.
  • the operation information receiving unit 222 may receive operation information based on the information input from the user terminal 1 from a second server device described later.
  • the selection instruction receiving unit 223 receives the selection instruction from the management terminal 3.
  • the selection instruction is an instruction to select a user.
  • the selection instruction usually has a user identifier.
  • the selection instruction may have a user terminal identifier. The details of the operation of the management terminal 3 will be described later.
  • the management terminal 3 receives and outputs one or more attribute values output by the attribute value output unit 251 to each of two or more users.
  • the processing unit 23 performs various processes.
  • the various processes are processes performed by the login processing unit 231 and the like.
  • the processing unit 23 constantly acquires information regarding browsing by a user who is visiting the website. Browsing information includes the length of stay on a website or web page, the number of pages browsed, and the like. That is, for example, the processing unit 23 constantly measures the staying time of the site or web page of the user who is visiting the website. Further, the processing unit 23 performs product purchase processing, payment processing, and the like based on the received operation information. Further, the processing unit 23 updates the dynamic attribute value of the user based on the received operation information.
  • the login processing unit 231 executes the login process for the user of the user terminal 1 in response to the received login instruction. Since the execution of the login process is a known technique, detailed description thereof will be omitted. Further, it is assumed that the execution of the login process usually includes the transmission of the result of the login process to the user terminal 1. Normally, the login processing unit 231 permits login and enables communication with the user terminal 1 when a valid user identifier or the like is received. Normally, when an invalid user identifier or the like is received, the login processing unit 231 disallows login and sends an error message to the user terminal 1. Further, it is preferable that the login processing unit 231 performs a process of associating the user identifier of the login instruction with the user terminal identifier. The user terminal identifier may be included in the login instruction or may be received together with the login instruction.
  • the response unit 232 performs processing according to the received operation information.
  • the processing according to the operation information is, for example, transmission of a web page corresponding to the operation information, purchase processing of a product corresponding to the operation information, payment processing corresponding to the operation information, and the like. Any processing may be performed as long as it follows the operation information.
  • the response unit 232 performs processing and acquires the processing result.
  • the processing result is, for example, a destination web page, a panel on which the result of a product purchase instruction is described, an error message, or the like. Further, since the processing of the response unit 232 is a known technique, detailed description thereof will be omitted.
  • the operation information storage unit 233 stores the operation information received by the operation information receiving unit 222 in the operation information storage unit 213 as a pair with the user terminal identifier.
  • the operation information storage unit 233 stores the operation information received by the operation information receiving unit 222 in the operation information storage unit 213 in pair with the user identifier included in the login instruction. Pairing with a user terminal identifier may be paired with a user identifier.
  • the operation information storage unit 233 stores the operation information received by the operation information receiving unit 222 in association with the user terminal identifier received in pairs with the operation information before the login processing of the login processing unit 231 is executed. Accumulate in 213. It should be noted that receiving in pairs does not have to be received at the same time.
  • the operation information storage unit 233 associates the user terminal identifier or the user identifier received in pairs with the login instruction or the operation information after the login process of the login processing unit 231 is executed, and the operation information receiving unit 222 receives the operation information. Is stored in the operation information storage unit 213.
  • the attribute value acquisition unit 234 acquires one or more attribute values of the user who is visiting the website.
  • one or more attribute values may include the above-mentioned static attribute value and dynamic attribute value.
  • the attribute value acquisition unit 234 acquires one or more attribute values of a user who is visiting the website by using the received one or more operation information.
  • the attribute value acquisition unit 234 normally acquires one or more static attribute values from the user information storage unit 211. Further, the attribute value acquisition unit 234 usually acquires one or more dynamic attribute values by using one or more received operation information.
  • the attribute value acquisition unit 234 acquires one or more dynamic attribute values, which are dynamically changing attribute values, using one or two or more operation information received by the operation information receiving unit 222.
  • the attribute value acquisition unit 234 is one or more attribute values of a user who is visiting the website, and is compared with one or more attribute values of another user who is a user other than the user, and satisfies a predetermined condition. Acquires one or more characteristic attribute values.
  • satisfying the predetermined conditions means that (1) the proportion of people who have the same attribute value as the user's attribute value is below the threshold value or smaller than the threshold value, and (2) other people who have the same attribute value. There is no such thing, (3) the number of people having the same attribute value as the user's attribute value is less than or equal to the threshold value or less than the threshold value, and (4) the same attribute value as the pre-stored attribute value is possessed.
  • the attribute value acquisition unit 234 acquires and acquires one or more operation information accumulated before login and one or more operation information accumulated after login from the operation information storage unit 213 for one user terminal 1. Acquire one or more attribute values using two or more operation information.
  • the score calculation means 2341 uses one or more attribute values of the user and two or more of the one or more operation information received by the operation information receiving unit 222 from the user terminal 1 of the user for one user. , Calculate the score of one user.
  • the score calculation means 2341 is, for example, one or more of the staying time, the number of purchases, the purchase amount, the total purchase amount, the number of PVs, the average staying time, and the number of visits of the website or web page acquired by the attribute value acquisition unit 234.
  • the score is calculated by an increasing function with information as a parameter.
  • the score is, for example, information indicating the importance of the user from the viewpoint of the administrator.
  • the number of PVs is the number of page views.
  • the thumbnail image acquisition unit 235 acquires a thumbnail image for each user by using the attribute value of 1 or more in the attribute value acquisition unit 234.
  • the thumbnail image acquisition unit 235 acquires, for example, an image corresponding to the gender and age of the user from the storage unit 21.
  • the thumbnail image acquisition unit 235 acquires a thumbnail image for each user by using one or more dynamic attribute values.
  • the thumbnail image acquisition unit 235 acquires a thumbnail image using, for example, attribute values such as the user's gender, age, and score.
  • the thumbnail image acquisition unit 235 acquires, for example, an original image corresponding to the gender and age of the user from the storage unit 21, generates a score image according to the user's score, pastes the score image into the original image, and thumbnails the image. To generate.
  • the original image corresponding to the gender and the age is stored in the storage unit 21.
  • the determination unit 236 determines whether or not the one or more attribute values acquired by the attribute value acquisition unit 234 satisfy the conditions of the dynamic processing information. In the determination unit 236, one or more attribute values acquired by the attribute value acquisition unit 234 satisfy the condition that any one or more dynamic processing information of the dynamic processing information storage unit 212 has. Judge whether or not.
  • condition processing execution unit 237 executes the processing identified by the processing identifier paired with the condition corresponding to the judgment result.
  • the instruction user processing unit 238 performs one process on the user terminal 1 of the user corresponding to the selection instruction.
  • One process may be a predetermined process or a process corresponding to the process identifier of the selection instruction.
  • One process may include a plurality of processes.
  • One process is, for example, advertisement distribution, recommendation to encourage purchase, sending of discount coupon, and the like.
  • the transmission unit 24 can transmit various information.
  • the various types of information are, for example, processing results and user terminal identifiers.
  • the processing result transmission unit 241 transmits the processing result related to the processing result in the response unit 232 to the user terminal 1.
  • the user terminal transmitting unit 242 sets the user terminal identifier, which is an identifier corresponding to the user terminal 1, to the user terminal 1. Send to.
  • the output unit 25 outputs various information.
  • the output is usually transmission to an external device (management terminal 3), but display on a display, projection using a projector, printing by a printer, sound output, storage on a recording medium, etc. It may be considered that the concept includes the delivery of the processing result to the processing device of the above and other programs.
  • the output unit 25 may transmit various information to the management terminal 3 corresponding to the second server device described later.
  • the attribute value output unit 251 outputs one or more attribute values acquired by the attribute value acquisition unit 234.
  • the attribute value output unit 251 outputs one or more dynamic attribute values acquired by the attribute value acquisition unit 234. It is preferable that the attribute value output unit 251 outputs one or more attribute values including the score.
  • the thumbnail image output unit 252 outputs the thumbnail image acquired by the thumbnail image acquisition unit 235.
  • the various types of information are stored in the management storage unit 31 that constitutes the management terminal 3.
  • the various types of information are, for example, an administrator identifier that identifies an administrator.
  • the management reception unit 32 receives various instructions, information, and the like.
  • the various instructions, information, and the like are, for example, selection instructions and dynamic processing information.
  • the input means for various instructions and information may be anything, such as a touch panel, keyboard, mouse, or menu screen.
  • the management reception unit 32 can be realized by a device driver for input means such as a touch panel or a keyboard, control software for a menu screen, or the like.
  • the management processing unit 33 performs various processes.
  • the various processes include a process of forming a data structure for transmitting instructions and information received by the management reception unit 32, a process of forming a data structure for outputting the received information by the management reception unit 35, and the like.
  • the management transmission unit 34 transmits various instructions, information, and the like.
  • the management transmission unit 34 normally transmits various instructions, information, and the like to the server device 2.
  • the various instructions, information, and the like are, for example, selection instructions and dynamic processing information.
  • the management receiving unit 35 receives various information.
  • the management receiving unit 35 usually receives various information from the server device 2.
  • the various information is, for example, one or more attribute values and thumbnail images for each user.
  • the management output unit 36 outputs various information.
  • the various information is one or more attribute values and thumbnail images for each user.
  • the user storage unit 11, the storage unit 21, the user information storage unit 211, the dynamic processing information storage unit 212, the operation information storage unit 213, and the management storage unit 31 are preferably non-volatile recording media, but are volatile. It can also be realized with a recording medium.
  • the process of storing information in the user storage unit 11 or the like does not matter.
  • the information may be stored in the user storage unit 11 or the like via the recording medium, or the information transmitted via the communication line or the like may be stored in the user storage unit 11 or the like.
  • the information input via the input device may be stored in the user storage unit 11 or the like.
  • the 238 and the management processing unit 33 can usually be realized from an MPU, a memory, or the like.
  • the processing procedure of the user processing unit 13 and the like is usually realized by software, and the software is recorded in a recording medium such as ROM. However, it may be realized by hardware (dedicated circuit).
  • the user transmission unit 14, the transmission unit 24, the processing result transmission unit 241, the user terminal transmission unit 242, the output unit 25, the attribute value output unit 251, the thumbnail image output unit 252, and the management transmission unit 34 are usually wireless or wired. It is realized by communication means.
  • the user receiving unit 15, the receiving unit 22, the login instruction receiving unit 221, the operation information receiving unit 222, the selection instruction receiving unit 223, and the management receiving unit 35 are usually realized by wireless or wired communication means.
  • the user output unit 16 and the management output unit 36 may or may not include output devices such as displays and speakers.
  • the user output unit 16 and the like can be realized by the driver software of the output device, the driver software of the output device, the output device, and the like.
  • Step S401 The user reception unit 12 determines whether or not the login instruction has been accepted. If the login instruction is accepted, the process goes to step S402, and if the login instruction is not accepted, the process goes to step S405.
  • Step S402 The user processing unit 13 configures a login instruction to be transmitted from the login instruction received in step S401.
  • the user transmission unit 14 transmits the login instruction to the server device 2.
  • Step S403 The user receiving unit 15 determines whether or not the result of the login process has been received. If the result of the login process is received, the process goes to step S404, and if the result of the login process is not received, the process returns to step S403.
  • Step S404 The user output unit 16 outputs the result of the login process received in step S403. Return to step S401.
  • Step S405 The user reception unit 12 determines whether or not an operation has been received from the user. If the operation is accepted, the process goes to step S406, and if the operation is not accepted, the process goes to step S410.
  • Step S406 The user processing unit 13 configures operation information based on the operation received in step S405.
  • Step S407 The user transmission unit 14 transmits the operation information configured in step S406 to the server device 2.
  • Step S408 The user receiving unit 15 determines whether or not the processing result corresponding to the operation information has been received from the server device 2. If the processing result is received, the process proceeds to step S409, and if the processing result is not received, the process returns to step S408.
  • Step S409 The user output unit 16 outputs the processing result received in step S408. Return to step S401.
  • Step S410 The user receiving unit 15 determines whether or not information has been received from the server device 2. If the information is received, the process goes to step S411, and if the information is not received, the process returns to step S401. It should be noted that such information is usually the result of processing in the server device 2, and is, for example, a coupon, an advertisement, a message, or the like.
  • Step S411 The user processing unit 13 configures information to be output using the information received in step S410.
  • the user output unit 16 outputs the information. Return to step S401.
  • Step S501 The login instruction receiving unit 221 determines whether or not the login instruction has been received from the user terminal 1. If the login instruction is received, the process goes to step S502, and if the login instruction is not received, the process goes to step S504.
  • Step S502 The login processing unit 231 executes a login process for the user of the user terminal 1 in response to the received login instruction.
  • the execution of the login process usually includes transmission of the result of the login process to the user terminal 1.
  • Step S503 The processing unit 23 associates the user terminal identifier with the user identifier. For example, the processing unit 23 stores the user terminal identifier and the user identifier of the login instruction as a pair in the storage unit 21. Return to step S501.
  • Step S504 The operation information receiving unit 222 determines whether or not the operation information or the like has been received from the user terminal 1 of the user who is a visitor to the website. If the operation information or the like is received, the process goes to step S505, and if the operation information or the like is not received, the process goes to step S511.
  • the operation information and the like are, for example, operation information and a user terminal identifier. Further, the operation information and the like are, for example, operation information and a user identifier. Further, the operation information and the like include, for example, one or more attribute values of the user.
  • Step S505 The response unit 232 performs processing according to the operation information received in step S504.
  • the user terminal transmitting unit 242 confirms that it is before receiving the login instruction from the user terminal 1, and when the operation information receiving unit 222 receives the operation information, the identifier corresponding to the user terminal 1.
  • the user terminal identifier is may be transmitted to the user terminal 1.
  • Step S506 The processing result transmission unit 241 transmits the processing result related to the processing result in the response unit 232 to the user terminal 1.
  • Step S507 The operation information storage unit 233 stores the operation information received in step S504 in the operation information storage unit 213 in association with the user terminal identifier or the user identifier received in pairs with the login instruction or the operation information. ..
  • Step S508 It is determined whether or not the user terminal identifier corresponding to the user terminal 1 that has transmitted the operation information is new (whether or not the user is the first visit). If it is new, it goes to step S509, and if it is not new, it returns to step S501.
  • Step S509 The processing unit 23 stores the user terminal identifier in the storage unit 21 by pairing the received operation information.
  • Step S510 The user terminal transmission unit 242 transmits the user terminal identifier to the user terminal 1 that has transmitted the operation information and the like. Return to step S501.
  • Step S511 The selection instruction receiving unit 223 determines whether or not the selection instruction has been received from the management terminal 3. If the selection instruction is received, the process goes to step S512, and if the selection instruction is not received, the process goes to step S514.
  • Step S512 The instruction user processing unit 238 acquires the user identifier or the user terminal identifier corresponding to the selection instruction.
  • the user identifier or user terminal identifier corresponding to the selection instruction may be a user identifier or a user terminal identifier included in the selection instruction, or may be a user identifier or a user terminal identifier paired with the ID included in the selection instruction.
  • the user identifier or user terminal identifier paired with the ID is managed by, for example, the storage unit 21.
  • Step S513 The instruction user processing unit 238 performs one process on the user terminal 1 identified by the user identifier or the user terminal identifier. Return to step S501.
  • the processing unit 23 performs administrator notification processing.
  • the administrator notification process is a process of outputting information of a currently visiting user to the management terminal 3. A specific example of the administrator notification process will be described with reference to the flowchart of FIG.
  • Step S575 The condition processing execution unit 237 performs automatic processing. Return to step S501.
  • the automatic processing is to automatically perform the processing corresponding to the condition on the user terminal 1 of the visiting user who matches the condition. A specific example of the automatic processing will be described with reference to the flowchart of FIG.
  • step S514 a specific example of the administrator notification process in step S514 will be described with reference to the flowchart of FIG.
  • Step S601 The processing unit 23 substitutes 1 for the counter i.
  • Step S602 The processing unit 23 determines whether or not the i-th visitor exists on the web page. If the i-th visitor exists, the process proceeds to step S603, and if the i-th visitor does not exist, the process returns to the higher-level processing.
  • Step S603 The attribute value acquisition unit 234 acquires the user identifier of the i-th visitor. Then, the attribute value acquisition unit 234 acquires one or more static attribute values paired with the user identifier from the user information storage unit 211.
  • Step S604 The attribute value acquisition unit 234 acquires one or more operation information paired with the user identifier of the i-th visitor from the operation information storage unit 213. Then, the attribute value acquisition unit 234 acquires one or more dynamic attribute values by using the one or more operation information.
  • the attribute value acquisition unit 234 may acquire one or more dynamic attribute values by using one or more attribute values of the user stored in the storage unit 21. It is preferable that the attribute value acquisition unit 234 stores the newly acquired dynamic attribute value of one or more in the user information storage unit 211 in association with the user identifier.
  • Step S605 The score calculation means 2341 scores using two or more attribute values out of one or more static attribute values acquired in step S603 and one or more dynamic attribute values acquired in step S604. Is calculated. It should be noted that this score may also be considered as a kind of dynamic attribute value.
  • Step S606 The thumbnail image acquisition unit 235 acquires the original image corresponding to one or more static attribute values acquired in step S603 from the storage unit 21.
  • Step S607 The thumbnail image acquisition unit 235 generates a score image using the score calculated in step S605.
  • Step S608 The thumbnail image acquisition unit 235 generates a thumbnail image using the original image acquired in step S606 and the score image generated in step S607.
  • the attribute value output unit 251 transmits one or more attribute values (static attribute value and dynamic attribute value) acquired by the attribute value acquisition unit 234 to the management terminal 3.
  • Step S610 The thumbnail image generated in step S608 is transmitted to the management terminal 3.
  • Step S611 The processing unit 23 increments the counter i by 1. Return to step S602.
  • step S515 a specific example of the automatic processing in step S515 will be described with reference to the flowchart of FIG.
  • Step S701 The processing unit 23 substitutes 1 for the counter i.
  • Step S702 The processing unit 23 determines whether or not the i-th visitor exists on the web page. If the i-th visitor exists, the process proceeds to step S703, and if the i-th visitor does not exist, the process returns to the higher-level processing. Whether or not the i-th visitor exists can be determined by inspecting the information in the storage unit 21. That is, the operation information of the visitor user, the user identifier, or the user terminal identifier is constantly updated and stored in the storage unit 21.
  • Step S703 The attribute value acquisition unit 234 acquires the user identifier of the i-th visitor. Then, the attribute value acquisition unit 234 acquires one or more static attribute values and / and one or more dynamic attribute values that are paired with the user identifier. Here, it is preferable to use the attribute value acquired at the time of processing the administrator notification of FIG.
  • Step S704 The processing unit 23 substitutes 1 for the counter j.
  • Step S705 The determination unit 236 determines whether or not the j-th dynamic processing information exists in the dynamic processing information storage unit 212. If the j-th dynamic processing information exists, the process goes to step S706, and if the j-th dynamic processing information does not exist, the process goes to step S711.
  • Step S706 The determination unit 236 determines whether or not the one or more attribute values acquired in step S703 meet the conditions of the j-th dynamic processing information.
  • Step S707 If the judgment result of the judgment unit 236 is a judgment result that the condition is met, the process goes to step S707, and if the judgment result is that the condition is not met, the process goes to step S710.
  • Step S708 The condition processing execution unit 237 acquires the processing identifier of the jth dynamic processing information.
  • Step S709 The condition processing execution unit 237 executes the processing identified by the processing identifier acquired in step S707.
  • Step S710 The processing unit 23 increments the counter j by 1. Return to step S605.
  • Step S711 The processing unit 23 increments the counter i by 1. Return to step S602.
  • Step S801 The management receiving unit 35 determines whether or not the user information has been received from the server device 2. If the information is received, the process goes to step S802, and if the information is not received, the process goes to step S803.
  • the user information includes one or more attribute values. Further, it is preferable that the user information includes a thumbnail image.
  • Step S802 The management output unit 36 outputs the information received in step S801. Return to step S801.
  • Step S803 The management reception unit 32 determines whether or not the selection instruction has been accepted. If the selection instruction is accepted, the process goes to step S804, and if the selection instruction is not accepted, the process goes to step S805.
  • Step S804 The management processing unit 33 has a data structure for transmitting the selection instruction received in step S803.
  • the management processing unit 33 constitutes, for example, a selection instruction including a user identifier of the selected user. Then, the management transmission unit 34 transmits the selection instruction to the server device 2. Return to step S801.
  • Step S805 The management reception unit 32 determines whether or not the dynamic processing information has been received. If the dynamic processing information is accepted, the process goes to step S806, and if the dynamic processing information is not accepted, the process returns to step S801.
  • Step S806 The management processing unit 33 changes the structure of the information for transmitting the dynamic processing information received in step S805. Then, the management transmission unit 34 transmits the dynamic processing information to the server device 2. Return to step S801. Here, the transmitted dynamic processing information is accumulated in the server device 2.
  • FIG. 1 The conceptual diagram of the information system A is FIG.
  • the server device 2 is, for example, a server in which a web page of an EC site for purchasing a product is stored. Then, it is assumed that the user visits the EC site and performs operations such as browsing the product and purchasing the product.
  • the user information storage unit 211 of the server device 2 stores a user information management table having the structure shown in FIG. Records having "user identifier”, “name”, “email address”, “static attribute value”, and “dynamic attribute value” are stored in the user information management table.
  • the "static attribute value” has “gender”, “age”, “unmarried / married”, “hometown”, “member” and the like.
  • "Unmarried / married” is information indicating whether the person is unmarried or married.
  • “Member” indicates whether or not the member is registered in this system (whether or not the member is).
  • the user corresponding to the member value "1" is a member, and the user corresponding to the member value "0" is not a member.
  • the “dynamic attribute value” has a “real-time dynamic attribute value” and a “history information utilization dynamic attribute value”.
  • the "real-time dynamic attribute value” has “stay time” and "number of pages viewed” here.
  • “Stay time” is the time spent on the site during the current visit.
  • "Number of pages viewed” is the number of pages viewed within the site during the current visit.
  • the “history information utilization dynamic attribute value” has, here, “visits”, “purchases”, “total purchase amount”, “average number of PVs", “score”, and the like.
  • “Number of visits” is the number of times the user has visited this EC site so far.
  • the “number of purchases” is the number of times a user has purchased a product on this EC site so far.
  • the “total purchase price” is the total price of the product purchased by the user on this EC site so far.
  • the "average number of PVs” is the average number of pages viewed per visit by the user.
  • the “score” is the user's score.
  • the storage unit 21 stores an arithmetic expression for calculating the score by an increasing function having one or more attribute values of "number of visits", “number of purchases", “total purchase amount”, and "average number of PVs" as parameters.
  • this calculation formula is a calculation formula for calculating a score using information of one or more of static attribute values.
  • the calculation formula may be a calculation formula calculated so as to increase the value of the score.
  • the calculation formula is, for example, multiplying the score value by 1.1 in the case of the member "1” and not changing the score value in the case of the member "0".
  • the dynamic processing information storage unit 212 stores a dynamic processing information management table having the structure shown in FIG. Records having "ID”, “condition”, “processing timing”, and “processing identifier” are stored in the dynamic processing information management table.
  • the "ID” is information for identifying a record.
  • the “condition” is a condition for executing the process.
  • “Processing timing” is information that specifies the timing for executing processing. When the processing timing is "-”, it means that the process is executed when the conditions are satisfied.
  • the “processing timing” may be only once a day, only once for one access, and the like.
  • the "processing identifier” is information for executing the processing corresponding to the condition, and here, is information for specifying the processing.
  • the message “I will deliver the coupon to be delivered” is sent, indicating that the coupon “Coupon 1" is to be sent to the user terminal 1 after login. It is assumed that the “coupon 1" is stored in the storage unit 21.
  • the original image corresponding to the condition of the attribute value of the user is stored in the storage unit 21.
  • the original score image, which is the source of the score image is stored in the storage unit 21.
  • the operation information management table having the structure shown in FIG. 11 is stored in the operation information storage unit 213.
  • one or more operation information is stored for each user.
  • records having "ID”, “user identifier”, “date and time”, “operation type identifier”, and “operation information” are stored.
  • the “ID” is information for identifying a record.
  • the “date and time” is generally the date and time when the operation was performed or the date and time when the operation information was received.
  • the "operation type identifier” is information indicating the type of operation information.
  • the operation type identifier "a” indicates that the operation information is information that identifies the user's operation.
  • the operation type identifier "b” indicates that the operation information is information that identifies the process executed by the server device 2.
  • the user reception unit 12 of the user terminal 1 receives a login instruction or operation information. Then, the user processing unit 13 constitutes the information to be transmitted. Next, the user transmission unit 14 transmits the configured information to the server device 2.
  • the receiving unit 22 of the server device 2 receives the login instruction or the operation information. Then, the login processing unit 231 or the response unit 232 performs processing according to the received login instruction or operation information. Then, the processing result transmission unit 241 transmits the processing result regarding the processing result in the response unit 232 to the user terminal 1.
  • the user can log in to this EC site, browse the product information of this EC site, and purchase the product.
  • the operation information storage unit 233 of the server device 2 stores the received operation information in the operation information management table.
  • the operation information storage unit 233 stores the operation information in the operation information management table in association with the user terminal identifier (here, Cookie ID). Further, after login, the operation information is stored in the operation information management table in association with the user terminal identifier and the corresponding user identifier.
  • the server device 2 can seamlessly use the operation information before login and the operation information after login even when the visitor of this EC site logs in during the visit.
  • the processing unit 23 performs the administrator notification processing as follows. That is, the processing unit 23 acquires the attribute values of each of the 11 visitors from the user information management table (FIG. 9). Then, the score calculation means 2341 calculates the score by using one or more attribute values used for score calculation among the one or more attribute values, and updates the score of each user in the user information management table.
  • the thumbnail image acquisition unit 235 acquires the original image corresponding to the acquired one or more static attribute values from the storage unit 21. Further, the thumbnail image acquisition unit 235 generates a score image using the calculated score. Next, the thumbnail image acquisition unit 235 generates a thumbnail image using the original image and the score image.
  • the output unit 25 transmits the thumbnail image acquired by the thumbnail image acquisition unit 235 and one or more attribute values acquired by the attribute value acquisition unit 234 to the management terminal 3 for each user.
  • the management receiving unit 35 of the management terminal 3 receives a thumbnail image and one or more attribute values for each visitor from the server device 2. Then, the management output unit 36 outputs the received information.
  • An example of such an output is shown in FIG. 1201 is a thumbnail image, 1202 is a score image in the thumbnail image, and 1203 is a user attribute value.
  • the attribute values and the like of the four visiting users are output, but as shown in 1204, there are currently 11 visitors. The administrator can also browse the attribute values of other visitors by scrolling the screen.
  • condition processing execution unit 237 performs automatic processing as follows. That is, the determination unit 236 determines for each of the 11 visitors whether or not the conditions paired with the matching processing timing among the conditions of each dynamic processing information in FIG. 10 are met. When the judgment result of the judgment unit 236 is a judgment result that the condition is satisfied, the condition processing execution unit 237 acquires the processing identifier paired with the condition. Then, the condition processing execution unit 237 executes the processing identified by the acquired processing identifier.
  • the user receiving unit 15 of the user terminal 1 of Ota B man receives the message "We will deliver a coupon to support you" from the server device 2. Further, the user receiving unit 15 receives a coupon called “coupon 1". Then, the user output unit 16 of the user terminal 1 of Ota B man outputs the message "I will deliver the coupon to support you" and "Coupon 1". And, Ota B man can enjoy shopping at a good price by using coupon 1. It is preferable that the output of the message or coupon is output in the web page being browsed by the user.
  • the management reception unit 32 receives the selection instruction of Csuke Tanaka. Then, it is assumed that the administrator has input or selected the action "send (coupon 2);” to be performed on Tanaka Csuke. Then, the management processing unit 33 makes the data structure "send (coupons 2 and 3);” for transmitting the accepted selection instruction.
  • the second argument of send is the user identifier "3" of Csuke Tanaka.
  • the management transmission unit 34 transmits the selection instruction "send (coupons 2 and 3);” to the server device 2.
  • the selection instruction receiving unit 223 of the server device 2 receives the selection instruction "send (coupons 2, 3);" from the management terminal 3.
  • the instruction user processing unit 238 acquires the user identifier "3" corresponding to the selection instruction.
  • the instruction user processing unit 238 transmits the coupon 2 to the user terminal 1 of Tanaka Csuke identified by the user identifier "3". It is assumed that the coupon 2 is stored in the storage unit 21, for example. However, the coupon 2 may be included in the selection instruction.
  • the user receiving unit 15 of the user terminal 1 via Tanaka C receives the coupon 2 from the server device 2. Then, the user processing unit 13 constitutes a coupon 2 that is output using the received coupon 2. The user output unit 16 outputs the coupon 2.
  • the management processing unit 33 acquires the user identifier "1" of the user "Ao Yamada”.
  • the management processing unit 33 acquires the operation information paired with the user identifier "1" from the operation information management table of FIG.
  • the management processing unit 33 configures an operation information presentation screen by using one or more acquired operation information.
  • the management output unit 36 outputs an operation information presentation screen. An example of such an operation information presentation screen is shown in FIG. In FIG. 13, operation information is output according to the type of operation information (operation information identifier), which makes it easier for the administrator to determine an action to be performed on the user.
  • the situation of visitors to the website can be grasped in real time.
  • the administrator of the site operation can take a more appropriate action for the user.
  • a specific action when the attribute value of the visitor to the website satisfies a specific condition, a specific action can be automatically executed. As a result, appropriate actions can be automatically taken for the user.
  • the administrator of the site operation can take an appropriate action for a specific user.
  • the attribute value of the user can be appropriately acquired by using both the operation information before login and the operation information after login.
  • the user's score can be output.
  • the administrator of the site operation can take an action against an appropriate user.
  • the thumbnail image acquired by using the attribute value of the user can be output.
  • a site operator can easily take action against an appropriate user.
  • the server device 2 and the management terminal 3 may be integrated.
  • the output unit 25 of the server device 2 usually displays various information.
  • the information system may be configured to include one or two or more user terminals 1, one or two or more second server devices 4, a server device 5, and one or more management terminals 3. .. It is assumed that the information system in such a case is the information system B. Further, the second server device 4 is a device that communicates with the user terminal 1 and transmits operation information to the server device 5. Further, the second server device 4 is, for example, a server of an existing EC site. Then, information such as operation information transmitted from the second server device 4 is received by the server device 5, and the server device 5 acquires one or more dynamic attribute values of the user using the operation information.
  • the server device 5 transmits one or more attribute values and thumbnail images of the user to the management terminal 3 corresponding to the second server device 4. Further, the second server device 4 takes an action on the user terminal 1 based on the instruction from the server device 5, or the server device 5 directly takes an action on the user terminal 1.
  • the action is, for example, the above-mentioned coupon sending, advertisement sending, message sending, and the like.
  • FIG. 14 shows a conceptual diagram of the information system B in such a case.
  • the information system B includes one or more user terminals 1, one or more second server devices 4, a server device 5, and one or more management terminals 3.
  • the second server device 4 and the server device 5 are, for example, an ASP server, a cloud server, and the like. However, the type of the second server device 4 and the server device 5 does not matter.
  • FIG. 15 shows a block diagram of the information system B in such a case.
  • the second server device 4 includes a second storage unit 41, a second reception unit 42, a second processing unit 43, and a second transmission unit 44.
  • the second storage unit 41 stores one or more static attribute values for each user.
  • the second receiving unit 42 includes a login instruction receiving unit 221 and a second operation information receiving unit 422.
  • the second operation information receiving unit 422 receives the operation information from the user terminal 1.
  • the second processing unit 43 includes a login processing unit 231 and a response unit 232.
  • the second transmission unit 44 includes a processing result transmission unit 241, a user terminal transmission unit 242, and a second operation information transmission unit 443.
  • the second operation information transmission unit 443 transmits the operation information to the server device 5.
  • the server device 5 includes a storage unit 21, a reception unit 52, a processing unit 53, a transmission unit 54, and an output unit 25.
  • the receiving unit 52 includes an operation information receiving unit 522 for receiving operation information from the second server device 4, and a selection instruction receiving unit 223.
  • the processing unit 53 includes an operation information storage unit 233, an attribute value acquisition unit 234, a thumbnail image acquisition unit 235, a determination unit 236, a condition processing execution unit 237, and an instruction user processing unit 238.
  • the attribute value acquisition unit 234 includes score calculation means 2341.
  • the transmission unit 54 includes a user terminal transmission unit 242.
  • the second storage unit 41 is preferably a non-volatile recording medium, but can also be realized by a volatile recording medium.
  • the process in which the information is stored in the second storage unit 41 does not matter.
  • the information may be stored in the second storage unit 41 via the recording medium, or the information transmitted via the communication line or the like may be stored in the second storage unit 41.
  • the information input via the input device may be stored in the second storage unit 41.
  • the second receiving unit 42 and the receiving unit 52 are usually realized by wireless or wired communication means.
  • the second processing unit 43 and the processing unit 53 can usually be realized from an MPU, a memory, or the like.
  • the processing procedure of the second receiving unit 42 and the like is usually realized by software, and the software is recorded in a recording medium such as ROM. However, it may be realized by hardware (dedicated circuit).
  • the second transmission unit 44 and the transmission unit 54 are usually realized by wireless or wired communication means.
  • the processing in the present embodiment may be realized by software. Then, this software may be distributed by software download or the like. Further, this software may be recorded on a recording medium such as a CD-ROM and disseminated. It should be noted that this also applies to other embodiments herein.
  • the software that realizes the server device in this embodiment is the following program. That is, this program includes, for example, a user information storage unit in which a recording medium accessible to a computer is information about a user and stores two or more user information which is information having one or more attribute values.
  • An operation information receiving unit that receives operation information related to the operation of the user's website from the user terminal of the user who is a visitor to the website, and a user who is visiting the website using the operation information.
  • This is a program for functioning as an attribute value acquisition unit that acquires one or more attribute values and an attribute value output unit that outputs one or more attribute values acquired by the attribute value acquisition unit.
  • the computer further functions as a response unit that performs processing according to the operation information and a processing result transmission unit that transmits the processing result related to the processing result in the response unit to the user terminal. Is preferable.
  • the software that realizes the management terminal in this embodiment is the following program. That is, in this program, for example, a management receiving unit that receives one or more attribute values for each user from a server device, a management output unit 36 that outputs one or more attribute values for each user, and the above 1 It is a program for functioning as a management reception unit for receiving a selection instruction of one user among the above users and a management transmission unit for transmitting the selection instruction of the one user to the server device.
  • a template having one or more learning parameters used in the learning process of machine learning and one or more prediction parameters used in the prediction process is stored, and the one or more learning parameters are used.
  • An information system including an information processing device that performs learning processing, acquires a learning device, acquires prediction processing using the learning device and one or more prediction parameters, and outputs a prediction result will be described.
  • the learner is used for prediction processing.
  • the prediction process is a process in which one or more explanatory variables are input and one or more objective variables are output.
  • the learning device may be referred to as a classifier, a learning model, or a classification model.
  • an information system including an information processing device that performs learning processing and prediction processing will be described using two or more templates for each administrator. Since the administrator uses the information processing device, it may be called a user. However, the administrator is usually different from the user corresponding to the user information which is an example of the data to be processed by the information processing device (for example, the user who purchases the product on the EC site).
  • an information system including an information processing device that performs learning processing and prediction processing will be described using a default template when used by an administrator who does not have a template.
  • an information system including an information processing device that performs learning processing and prediction processing by using one template selected by the administrator from two or more templates will be described.
  • an information system including a template having parameters used in post-processing and an information processing device for performing post-processing using the prediction results and parameters which are the results of the prediction processing will be described.
  • an information system including an information processing device for classifying users who purchase products (for example, determining whether or not they are royal customers) will be described.
  • an information system including an information processing device for forecasting product demand will be described.
  • FIG. 16 is a conceptual diagram of the information system C in the present embodiment.
  • the information system C includes an information processing device 6 and one or more terminal devices 7.
  • the information processing device 6 is a device that performs learning processing and prediction processing of machine learning.
  • the information processing device 6 is, for example, an ASP server, a cloud server, or the like. However, the type of the information processing device 6 does not matter.
  • the terminal device 7 is a device used by the administrator.
  • the terminal device 7 is, for example, a so-called personal computer, a tablet terminal, a smartphone, or the like, and the type thereof does not matter.
  • the information processing device 6 and the terminal device 7 can communicate with each other via a network such as the Internet or a LAN.
  • FIG. 17 is a block diagram of the information system C in the present embodiment.
  • FIG. 18 is a block diagram of the information processing apparatus 6.
  • the information processing device 6 includes a storage unit 61, a reception unit 62, a processing unit 63, and an output unit 64.
  • the storage unit 61 includes a teacher source data storage unit 611, a target data storage unit 612, and a template storage unit 613.
  • the reception unit 62 includes an identifier reception unit 621.
  • the processing unit 63 includes a learning unit 631, a prediction unit 632, and a post-processing unit 633.
  • the learning unit 631 includes, for example, a teacher data acquisition unit 6311, a learning unit 6312, an evaluation unit 6313, and a selection unit 6314.
  • the learning unit 631 may be, for example, only the teacher data acquisition means 6311 and the learning means 6312.
  • the terminal device 7 includes a terminal storage unit 71, a terminal reception unit 72, a terminal processing unit 73, a terminal transmission unit 74, a terminal reception unit 75, and a terminal output unit 76.
  • the various types of information include, for example, teacher source data described later, teacher data described later, target data described later, templates described later, and various programs.
  • the various programs are, for example, programs that perform learning processing and prediction processing of machine learning, which will be described later.
  • the various programs are, for example, programs that perform statistical processing described later.
  • the various programs are, for example, programs that perform post-processing described later.
  • the teacher source data described later, the target data described later, the template described later, and the like exist in the storage unit 61, but they may exist in an external device (not shown).
  • the storage unit 61 does not include the teacher source data storage unit 611, the target data storage unit 612, and the template storage unit 613.
  • the reception unit 62 receives the teacher source data described later, the target data described later, the template described later, and the like from an external device (not shown).
  • the teacher source data storage unit 611 stores one or more teacher source data.
  • the teacher source data is information that is the source of the teacher data used by the learning unit 631.
  • the teacher source data may be the same as the teacher data.
  • the teacher source data storage unit 611 may be the user information storage unit 211 described above. That is, the teacher source data is, for example, user information. It is preferable that the teacher source data has time information indicating the time. It is preferable that the teacher source data includes dynamic attribute values.
  • the dynamic attribute value here is, for example, information regarding an operation when a user purchases a product.
  • the dynamic attribute value includes, for example, the above-mentioned real-time dynamic attribute value and the above-mentioned history information utilization dynamic attribute value.
  • the real-time dynamic attribute value is, for example, the staying time of the web page currently viewed by the user, the number of web pages viewed during the current stay, and the like.
  • the dynamic attribute value using historical information is a dynamic attribute value acquired by using the history of operation information at the time of a past visit.
  • the historical information usage dynamic attribute value is, for example, the number of purchases, the purchase price, the total purchase price, the average staying time, the average number of PVs, the number of visits, the score described later, the average interval of sessions, the average purchase unit price, and the like.
  • the teacher data may be the teacher source data or a part of the teacher source data.
  • the teacher data may include explanatory variables that can be obtained from the teacher source data by a predetermined operation.
  • the calculation is, for example, statistical processing of two or more pieces of information and processing of one piece of information.
  • the teacher data includes one or more explanatory variables and one or more objective variables.
  • the target data storage unit 612 stores the prediction target data.
  • the forecast target data is the forecast target data.
  • the prediction target data is information used by the prediction unit 632 for prediction processing.
  • the prediction target data is a set of one or more explanatory variables. It can be said that the prediction target data is a set of one or more features.
  • the prediction target data includes one or more explanatory variables.
  • the explanatory variables included in the prediction target data are usually the same as the explanatory variables included in the teacher data.
  • the prediction target data is, for example, all or part of the above-mentioned user information.
  • the prediction target data is, for example, all or part of the user information of the user who has caused a specific action.
  • the specific action may be, for example, initial installation, initial login, initial purchase, login, predetermined operation on the EC site, or may be in a specific state such as rank on the site. ..
  • the initial installation is, for example, the installation of the EC application for the first time.
  • the first login is the first login after the user registration is completed.
  • the first purchase is, for example, the first purchase of some product on an EC site.
  • One or two or more templates are stored in the template storage unit 613.
  • the template contains one or more learning parameters and one or more prediction parameters.
  • the template may include one or more post-processing parameters.
  • the template may include one or more output parameters.
  • the template is, for example, a file.
  • the template may be a database table, a record in a database table, or the like.
  • the data structure of the template does not matter.
  • the learning parameter is data related to the learning process of machine learning.
  • Learning parameters are usually data used for the learning process of machine learning.
  • the learning parameters determine, for example, the data passed to the machine learning learning process module, the data to acquire the data passed to the machine learning learning process module, and when to execute the machine learning learning process module. It is the data for.
  • one or more learning parameters include one or more types of parameters among a learning timing parameter, a teacher source data selection parameter, a teacher data configuration parameter, and a learner configuration parameter.
  • (1-1) Learning timing parameters include one or more types of parameters among a learning timing parameter, a teacher source data selection parameter, a teacher data configuration parameter, and a learner configuration parameter.
  • the learning timing parameter is data that specifies the timing that constitutes the learning device.
  • the learning timing parameter is, for example, information that specifies a period when the learning process is periodically performed.
  • the learning timing parameter is, for example, data indicating whether or not to perform (real-time) each time the information to be learned is received.
  • the learning timing parameter is, for example, information indicating whether or not the learning process is performed according to the instruction of the administrator. (1-2) Parameters for selecting teacher source data
  • the teacher source data selection parameter is the data for selecting the teacher source data for creating the teacher data.
  • the parameter for selecting the teacher source data is, for example, data for constructing a search formula for the teacher source data.
  • the parameters for selecting the teacher source data are, for example, a learning period parameter and a learning target parameter. (1-2-1) Learning period parameter
  • the learning period parameter is information that identifies the time corresponding to the teacher source data.
  • the learning period parameter is, for example, information that specifies the period of the teacher source data to be selected.
  • the learning period parameter is, for example, time information possessed by the teacher source data, and is time information for determining matching time information. (1-2-2) Learning target parameters
  • the learning target parameter is data for specifying the learning target.
  • the learning target parameter is data for determining specific data possessed by the teacher source data.
  • the learning target parameters are, for example, a teacher user selection parameter, a teacher item selection parameter, and a data source parameter.
  • the teacher source data is information about the user (for example, the user information described above)
  • the teacher user selection parameter is information for selecting the user
  • the static attribute value of the user for example, gender, age
  • Information of one or more of the user's dynamic attribute values eg, number of purchases, score.
  • the teacher source data is information about the item (eg, product sales information)
  • the teacher item selection parameter is the information for selecting the item and the static attribute value of the item (eg, price, product type).
  • Information of one or more of the dynamic attribute values of the item for example, the number of sales, the total sales amount).
  • the data source parameter is, for example, information that identifies whether the teacher source data is the teacher source data obtained from the web or the teacher source data obtained from the application. (1-3) Parameters for teacher data composition
  • the teacher data composition parameter is data for acquiring an explanatory variable or an objective variable that constitutes the teacher data.
  • the parameters for teacher data composition are, for example, an aggregation period parameter and a degree of smoothing.
  • the aggregation period parameter is, for example, data that specifies the period of the teacher source data that is the target when calculating the explanatory variable that is the result of statistically processing the specific information constituting the teacher source data.
  • the aggregation period parameter is, for example, data that specifies the period of the teacher source data that is the target when calculating the objective variable that is the result of statistically processing the specific information constituting the teacher source data.
  • the aggregation period parameter is, for example, information that specifies the aggregation period of the dynamic attribute value.
  • the degree of smoothing is data indicating the degree of smoothing when statistically processing the time-series data constituting the teacher source data.
  • the degree of smoothing is, for example, 0 to 5, and a larger numerical value means that the degree of smoothing is increased. (1-4) Parameters for learning device configuration
  • the learning device configuration parameter is the data for acquiring the learning device.
  • the learning device configuration parameters are, for example, one or more learning algorithm parameters, learning device evaluation parameters, and learning accuracy parameters.
  • the parameters for evaluating the learner may be called method parameters.
  • the learning algorithm parameter is data that specifies the algorithm for acquiring the learning device.
  • the learning algorithm parameter is, for example, an identifier that identifies the algorithm for acquiring the learner.
  • the learning device evaluation parameter is, for example, an identifier of an algorithm for evaluating the learning device.
  • the parameter for evaluation of the learner is, for example, a value (natural number) indicating K in the K division method.
  • the learning device evaluation parameter is, for example, an identifier of the learning device evaluation algorithm. It was
  • the learning accuracy parameter is data indicating a threshold value of accuracy adopted as a learning device.
  • the learning accuracy parameter is, for example, information about one or more of the matching rate, the recall rate, the F value, and the accuracy, and is, for example, "the matching rate, the recall rate, the F value, or the accuracy is equal to or more than the threshold value" and "the matching rate". , The recall, the F-number, or the accuracy is greater than the threshold.
  • Prediction parameters are data related to machine learning prediction processing. Predictive parameters are usually data used for predictive processing in machine learning.
  • the one or more prediction parameters are, for example, a prediction timing parameter, a prediction period parameter, and a prediction target parameter. (2-1) Prediction timing parameters
  • the prediction timing parameter is data that specifies the timing for performing the prediction process.
  • the prediction timing parameter is, for example, the frequency of performing the prediction process (repetition frequency) and the date and time of performing the prediction process.
  • the prediction timing parameter is, for example, data indicating whether or not to perform (real-time) each time the information of the prediction target is received.
  • the prediction timing parameter is, for example, information indicating whether or not the prediction process is performed according to the instruction of the administrator. (2-2) Forecast period parameter
  • the forecast period parameter is data that specifies the period for which the forecast process is performed.
  • the prediction period parameter is, for example, the inference target period.
  • Forecast period parameters are, for example, information that identifies the future time for forecasting demand (for example, until one month later, until July 1, 2020), and information that specifies when to determine whether or not to become a royal customer. (For example, until the end of this year, September 30, 2020). (2-3) Prediction target parameters
  • the prediction target parameter is data for specifying the prediction target.
  • the prediction target parameters are, for example, prediction user selection parameters and prediction item selection parameters. (2-3-1) Predictive user selection parameters
  • Predictive user selection parameters are information for selecting a user.
  • the predictive user selection parameter is, for example, information constituting a search condition for user information.
  • the predictive user selection parameter is, for example, one or more types of information among a user's static attribute value (for example, gender, age) and a user's dynamic attribute value (for example, number of purchases, score).
  • the predictive user selection parameters are, for example, "user who installed the application”, "user who logged in for the first time”, “user who purchased the product for the first time”, and the like. (2-3-2) Predicted item selection parameters
  • Predicted item selection parameters are information for selecting items.
  • the predictive item selection parameter is, for example, information that constitutes a search condition for an item.
  • the predicted item selection parameter is, for example, one or more types of information of a static attribute value of an item (for example, price, product type) and a dynamic attribute value of an item (for example, number of sales, total sales amount).
  • the post-processing parameters are data for post-processing using the prediction result.
  • the post-processing parameter is, for example, data given to a module that performs post-processing.
  • the prediction result is the number of purchases and the output information is information indicating whether or not the user is a royal user
  • the post-processing parameter is, for example, a threshold value for determining that the user is a royal user.
  • the output parameter is the data used when outputting information.
  • the output parameters are, for example, an output type parameter, an output mode parameter, an output medium parameter, and an output destination parameter.
  • the output type parameter is data that specifies the type of information to be output.
  • the output type parameters are, for example, "prediction result” and "output information”.
  • the output type parameter is, for example, information indicating whether or not to output the prediction result and whether or not to output the output information.
  • the output mode parameter is data that specifies the output mode of information.
  • the output mode parameter is, for example, whether to output the information as a character string or to output a graph.
  • the output mode parameters are, for example, "graph”, "table”, and "character string”.
  • the output medium parameter is data that identifies the output medium of the information.
  • the output medium parameter is, for example, data that specifies whether to store, transmit, or display information.
  • the output medium parameters are, for example, "display”, “recording medium”, “communication”, and "mail”.
  • the output destination parameter is data that specifies the output destination.
  • the output destination parameters are, for example, a folder name of a disk, an e-mail address, a server name of a destination, and an IP address of the server of the destination.
  • the template storage unit 613 stores the template in association with the administrator identifier.
  • the administrator identifier is, for example, an ID, a name, an e-mail address, a telephone number, or the like. Templates are usually associated with a template identifier.
  • the template identifier is information that identifies the template.
  • the template identifier is, for example, an ID or a template name. It is preferable that the template identifier is information that specifies the target of the prediction process or information that specifies the prediction result.
  • the default template is stored in the template storage unit 613.
  • the template storage unit 613 stores two or more templates used for outputting prediction results of different targets in association with the template identifier.
  • the reception unit 62 receives various information, instructions, and the like.
  • the various information, instructions, and the like are, for example, an administrator identifier, a template identifier, an operation instruction, a template, and one or more parameters constituting the template.
  • the administrator identifier is information that identifies the administrator.
  • the template identifier is information that identifies the template.
  • the template identifier is, for example, an ID or a template name.
  • the operation instructions are, for example, learning instructions and prediction instructions.
  • the learning instruction is an instruction to execute the learning process.
  • the predictor finger is an instruction to execute the prediction process.
  • the operation instruction, learning instruction, and prediction instruction have, for example, an administrator identifier.
  • the operation instruction, the learning instruction, and the prediction instruction have, for example, a template identifier.
  • reception is usually reception from the terminal device 7.
  • the information processing apparatus 6 When the information processing apparatus 6 operates standalone, it is a concept including acceptance and acceptance of information read from a recording medium such as an optical disk, a magnetic disk, or a semiconductor memory.
  • the identifier reception unit 621 accepts the template identifier.
  • the identifier receiving unit 621 accepts, for example, a template identifier and an administrator identifier.
  • the processing unit 63 performs various processes.
  • the various processes are, for example, processes performed by the learning unit 631, the prediction unit 632, and the post-processing unit 633.
  • the learning unit 631 performs learning processing and acquires a learning device.
  • the learning process is a learning process using a machine learning algorithm.
  • the machine learning algorithm is, for example, deep learning, SVR, random forest, decision tree, or the like. However, the machine learning algorithm does not matter.
  • performing learning processing using a machine learning algorithm gives two or more teacher data to a function that realizes a machine learning algorithm (for example, a function of TinySVM, fastText, TensorFlow), and executes the learning process. That is.
  • the function may be called a method, a module, or the like.
  • the learning unit 631 acquires one or more learning parameters of the template to be used, and is one or more teacher source data corresponding to the one or more learning parameters, and one or more stored in the teacher source data storage unit 611.
  • the teacher data used for the learning process is acquired from each teacher source data of. Then, the learning unit 631 performs a learning process on the one or more teacher data and acquires a learning device.
  • the learning unit 631 acquires, for example, one or more teacher source data selection parameters from the template to be used. Next, the learning unit 631 selects one or more teacher source data from the teacher source data storage unit 611 using, for example, one or more acquired teacher source data selection parameters. Then, the learning unit 631 acquires, for example, one or more teacher data configuration parameters from the template to be used. Next, the learning unit 631 acquires teacher data from each of the acquired one or more teacher source data by using, for example, one or more acquired teacher data configuration parameters. Further, the learning unit 631 acquires, for example, one or more learning device configuration parameters from the template to be used.
  • the learning unit 631 performs one or two or more learning processes using, for example, one or two or more teacher data, and acquires one or two or more learners.
  • the learning unit 631 acquires one learning device to be used for the prediction process by using, for example, one or more learning devices and one or more learning device configuration parameters.
  • the learning unit 631 acquires, for example, a learning period parameter from a template to be used.
  • the learning unit 631 has one or more teacher source data having information corresponding to the period specified by the learning period parameter and having information matching the learning target parameter from the teacher source data storage unit 611. decide.
  • the learning unit 631 acquires one or more teacher data configuration parameters from the target template to be used.
  • the learning unit 631 uses one or more teacher data construction parameters for each of the determined one or more teacher source data, and one or more explanatory variables and one or more objective variables that constitute the teacher data. And get.
  • the learning unit 631 performs learning processing on two or more teacher data having one or more explanatory variables and one or more objective variables, and acquires one or more learners.
  • the learning unit 631 acquires one or more learning device configuration parameters from the target template to be used.
  • the learning unit 631 evaluates each of the acquired one or more learning devices using the acquired one or more learning device configuration parameters, and acquires one learning device.
  • the learning unit 631 performs learning processing using, for example, one or more learning parameters of the template corresponding to the administrator identifier received by the reception unit 62, and acquires a learning device.
  • the learning unit 631 performs learning processing using, for example, one or more learning parameters possessed by the template corresponding to the administrator identifier and the template identifier received by the reception unit 62, and acquires a learning device.
  • the learning unit 631 performs learning processing using, for example, one or more learning parameters of the template identified by the template identifier received by the identifier receiving unit 621, and acquires a learning device.
  • the learning unit 631 is one or more teacher source data corresponding to the template identified by the template identifier received by the identifier reception unit 621, and is from each one or more teacher source data stored in the teacher source data storage unit 611. , Acquire the teacher data used for the learning process, perform the learning process on the one or more teacher data, and acquire the learner.
  • the learning unit 631 performs learning processing using one or more learning parameters of the default template. , Get the learner.
  • the learning unit 631 reads the default template from the template storage unit 613 and uses it.
  • the learning unit 631 performs learning processing using, for example, one or more teacher data including one or more dynamic attribute values, and acquires a learning device.
  • the learning unit 631 performs learning processing using one or more teacher data including one or more static attribute values in addition to one or more dynamic attribute values, and acquires a learning device.
  • the teacher data acquisition means 6311 acquires one or two or more teacher data.
  • the teacher data acquisition means 6311 acquires one or more teacher data by using the teacher data of the teacher data storage unit 611.
  • the teacher data acquisition means 6311 acquires, for example, one or more learning parameters possessed by the template to be used, and acquires one or more teacher source data corresponding to the one or more learning parameters from the teacher source data storage unit 611.
  • the teacher data used for the learning process is acquired from each one or more teacher source data.
  • the teacher data acquisition means 6311 acquires, for example, one or more teacher data from one or more teacher source data having time information corresponding to the period specified by the learning period parameter.
  • the one or more teacher source data is the data of the teacher source data storage unit 611.
  • the teacher data acquisition means 6311 selects one or more teacher source data from the teacher source data storage unit 611 using, for example, one or more teacher source data selection parameters, and one or more from each of the one or more teacher source data. Acquire teacher data using the teacher data composition parameters of.
  • the teacher data acquisition means 6311 reads one or more teacher source data selection parameters and one or more teacher data configuration parameters from the template.
  • the teacher data acquisition means 6311 is, for example, from one or more teacher source data having time information corresponding to the period specified by the learning period parameter, one or more dynamic attribute values according to the aggregation period specified by the aggregation period parameter. Calculate the aggregated result of. Next, the teacher data acquisition means 6311 acquires one or more teacher data having one or more calculated aggregated results as explanatory variables and having a prediction target parameter as an objective variable.
  • the learning means 6312 performs learning processing on two or more teacher data acquired by the teacher data acquisition means 6311, and acquires a learning device.
  • the learning means 6312 performs learning processing of two or more different algorithms on two or more teacher data acquired by the teacher data acquisition means 6311, and acquires two or more learners.
  • the learning process of different algorithms is, for example, deep learning, random forest, SVM, SVR, decision tree, and the like.
  • the evaluation means 6313 evaluates each of one or more learning devices, and acquires the evaluation result for each learning device.
  • the evaluation is an evaluation regarding the accuracy of the learner, and is, for example, an evaluation using a K division method, a holdout method, a cross-validation, a mixed matrix, or the like.
  • the evaluation algorithm performed by the evaluation means 6313 may be fixed or may be dynamically changed according to the learning device configuration parameters.
  • the evaluation means 6313 evaluates each one or more learning devices based on the method parameters, and acquires the evaluation results for each of the two or more learning devices.
  • the selection means 6314 selects one learning device based on the evaluation result of one or two or more learning devices.
  • the selection means 6314 usually selects one learner with the best evaluation result (usually high accuracy). It should be noted that the selection means 6314 may not be able to acquire the learner.
  • the prediction unit 632 performs prediction processing on the prediction target data stored in the target data storage unit 612 using one or more prediction parameters and a learning device of the template to be used, and acquires the prediction result.
  • the acquisition of the prediction result may be obtained from the prediction processing performed when the query matches the preset query.
  • the learning device is information acquired by the learning unit 631.
  • the prediction process is a machine learning algorithm prediction process.
  • the machine learning algorithm is, for example, deep learning, SVR, random forest, decision tree, or the like. However, the machine learning algorithm does not matter.
  • performing prediction processing using a machine learning algorithm means that a learner and one or more prediction target data are added to a function that performs prediction processing using a machine learning algorithm (for example, a function of TinySVM, fastText, TensorFlow). Give and do.
  • the function may be called a method, a module, or the like.
  • the prediction unit 632 performs prediction processing using, for example, one or more prediction parameters included in the template corresponding to the administrator identifier received by the reception unit 62, and acquires the prediction result. For example, the prediction unit 632 acquires one or more prediction parameters of the template corresponding to the administrator identifier and the template identifier received by the reception unit 62, performs prediction processing using the one or more prediction parameters, and performs prediction processing. Get the prediction result.
  • the prediction unit 632 performs prediction processing using one or more prediction parameters of the default template. , Get the prediction result.
  • the prediction unit 632 is stored in the target data storage unit 612 using, for example, one or more prediction parameters of the template identified by the template identifier received by the identifier reception unit 621 and the learner acquired by the learning unit 631. Prediction processing is performed on the forecast target data, and the prediction result is acquired.
  • the prediction unit 632 acquires the prediction target data from the target data storage unit 612 in which one or more prediction target data, which is the prediction target data, is stored by using one or more prediction parameters of the template, and selects the prediction target data. Using the learner selected by the means 6314, the prediction process is performed on the prediction target data, and the prediction result is acquired.
  • the prediction unit 632 acquires, for example, the prediction result of the period according to the prediction period parameter of the template. For example, the prediction unit 632 acquires a learning device corresponding to the prediction period parameter of the template, performs prediction processing on the prediction target data using the learning device, and acquires the prediction result.
  • the prediction unit 632 performs prediction processing at a timing specified by a prediction timing parameter and acquires a prediction result.
  • the prediction unit 632 performs prediction processing using, for example, one or more prediction target data including one or more dynamic attribute values, and acquires a prediction result.
  • the prediction result is, for example, a user category, a user purchase count, a user purchase amount, and the like.
  • one or more dynamic attribute values are explanatory variables, and the prediction result is an objective variable.
  • the prediction unit 632 performs prediction processing using, for example, one or more prediction target data including one or more static attribute values, and acquires a prediction result.
  • one or more static attribute values are explanatory variables.
  • the prediction unit 632 calculates the aggregation result of one or more dynamic attribute values of the user corresponding to the prediction user selection parameter, and has the calculated one or more aggregation results as explanatory variables. Prediction processing is performed using and, and the prediction result is acquired.
  • the prediction unit 632 acquires one or more information about the sales result of the item corresponding to the prediction item selection parameter, and uses the explanatory variable group having the one or more information as the explanatory variables and the learning device to perform prediction processing. And get the forecast result which is the information about the future sales result.
  • the post-processing unit 633 processes the prediction result using one or more post-processing parameters and acquires output information.
  • the post-processing unit 633 acquires, for example, from a template that uses one or more post-processing parameters, and acquires output information using the one or more post-processing parameters and the prediction result.
  • the post-processing unit 633 determines, for example, whether or not it corresponds to a specific user by using the post-processing parameter, and acquires output information by using the determination result.
  • the specific user is, for example, a royal customer.
  • the post-processing unit 633 uses, for example, a prediction result that is the number of purchases or the purchase price acquired by the prediction unit 632 and a threshold value that is a post-processing parameter, and whether or not the prediction result is equal to or greater than the threshold value (or larger than the threshold value). If it is equal to or greater than the threshold value (or larger than the threshold value), the output information that makes the user a royal customer is acquired.
  • the output information is, for example, a user identifier, a user category identifier, or a user identifier of a user belonging to a specific category.
  • the user category identifier is information that identifies the user's category.
  • the output unit 64 outputs, for example, a prediction result.
  • the output unit 64 outputs, for example, output information.
  • the output unit 64 outputs output information, for example, in place of the prediction result or in addition to the prediction result.
  • the output unit 64 outputs information according to, for example, an output parameter.
  • the output unit 64 outputs, for example, the type of information indicated by the output type parameter (for example, one or more of the prediction result and the output information) in the mode indicated by the output mode parameter (for example, a character string, a graph, etc.).
  • the output is performed to the output destination indicated by the output destination parameter (for example, the folder A of the recording medium, the display, the external device X, etc.).
  • the output is usually a transmission to the terminal device 7.
  • the output means, for example, display on a display, projection using a projector, printing by a printer, sound output, transmission to an external device, storage on a recording medium, and the like. It is a concept that includes the delivery of the processing result to the processing device of the above and other programs.
  • the various types of information are stored in the terminal storage unit 71 that constitutes the terminal device 7.
  • the various types of information are, for example, an administrator identifier, a terminal identifier, and the like.
  • the terminal identifier is information that identifies the terminal device 7, and is, for example, a cookie ID, a session identifier, an IP address, a MAC address, or the like.
  • the terminal reception unit 72 receives input of instructions, information, etc. from the administrator.
  • the instruction, information, and the like are, for example, a template identifier, an operation instruction, a learning instruction, a prediction instruction, a template, and one or more parameters constituting the template.
  • the input means for instructions and information may be any means such as a touch panel, a keyboard, a mouse, and a menu screen.
  • the terminal processing unit 73 performs various processes.
  • the various processes include, for example, a process of changing the instructions and information received by the terminal receiving unit 72 into instructions and information of a structure to be transmitted, and a process of changing to a structure of outputting the information received by the terminal receiving unit 75. And so on.
  • the terminal transmission unit 74 transmits various information, instructions, and the like.
  • the various information, instructions, and the like are, for example, an administrator identifier, a template identifier, an operation instruction, a template, and one or more parameters constituting the template.
  • the terminal transmission unit 74 normally transmits information, instructions, and the like to the information processing device 6.
  • the terminal receiving unit 75 receives various information.
  • the various types of information are, for example, prediction results and output information.
  • the terminal output unit 76 outputs various information.
  • the various types of information are, for example, information that has been changed to a structure that is received by the terminal receiving unit 75 and output by the terminal processing unit 73, and is, for example, prediction results and output information.
  • the output is usually a display on a display, but is projected by a projector, printed by a printer, sound output, transmitted to an external device, stored in a recording medium, other processing devices and others. It may be considered that the concept includes the delivery of the processing result to the program or the like.
  • a non-volatile recording medium is suitable for the storage unit 61, the teacher source data storage unit 611, the target data storage unit 612, the template storage unit 613, and the terminal storage unit 71, but a volatile recording medium can also be used. ..
  • the process of storing information in the storage unit 61 or the like does not matter.
  • the information may be stored in the storage unit 61 or the like via the recording medium, or the information transmitted via the communication line or the like may be stored in the storage unit 61 or the like.
  • the information input via the input device may be stored in the storage unit 61 or the like.
  • the reception unit 62, the identifier reception unit 621, and the terminal reception unit 75 are usually realized by wireless or wired communication means.
  • the processing unit 63, the learning unit 631, the prediction unit 632, the post-processing unit 633, the teacher data acquisition unit 6311, the learning unit 6312, the evaluation unit 6313, the selection unit 6314, and the terminal processing unit 73 can be realized from a processor, a memory, or the like. ..
  • the processing procedure of the processing unit 63 and the like is usually realized by software, and the software is recorded in a recording medium such as ROM. However, it may be realized by hardware (dedicated circuit). It goes without saying that the processor is a CPU, MPU, GPU, or the like, and the type thereof does not matter.
  • the output unit 64 and the terminal transmission unit 74 are usually realized by wireless or wired communication means.
  • the terminal reception unit 72 can be realized by a device driver for input means such as a touch panel or a keyboard, control software for a menu screen, or the like.
  • the terminal output unit 76 may or may not include an output device such as a display or a speaker.
  • the terminal output unit 76 can be realized by the driver software of the output device, the driver software of the output device, the output device, or the like.
  • Step S1901 The reception unit 62 determines whether or not the template or the like has been received. If the template or the like is received, the process goes to step S1902, and if the template or the like is not received, the process goes to step S1903.
  • the template and the like are, for example, a template only, a template and an administrator identifier, a template and a template identifier, and a template and an administrator identifier and a template identifier.
  • the template or the like may be the entire template, or may be a part of the template to be updated or added.
  • Step S1902 The processing unit 63 stores the template received in step S1901 in the template storage unit 613 in association with the administrator identifier or the like.
  • the administrator identifier and the like are an administrator identifier, an administrator identifier and a template identifier, or a template identifier.
  • the processing unit 63 stores a part of the template in the template storage unit 613 in association with the administrator identifier or the like.
  • Step S1903 The processing unit 63 determines whether or not to perform the learning process. If the learning process is performed, the process goes to step S1904, and if the learning process is not performed, the process goes to step S1908. In the case of performing the learning process, for example, when a learning instruction is received from the terminal device 7, it is a case where a predetermined timing is reached.
  • the predetermined timing is, for example, a periodic timing and a predetermined timing.
  • the predetermined timing is, for example, when the date and time stored in the storage unit 61 is reached.
  • Step S1904 The learning unit 631 substitutes 1 for the counter i.
  • Step S1905 The learning unit 631 determines whether or not the i-th learning target exists. If the i-th learning target exists, the process goes to step S1906, and if the i-th learning target does not exist, the process returns to step S1901.
  • the i-th learning target is shown in, for example, the received learning instruction.
  • the learning target shown in the learning instruction is, for example, a target of learning processing using the template identified by the template identifier included in the learning instruction.
  • the learning target shown in the learning instruction is, for example, a target of learning processing using a template paired with the administrator identifier included in the learning instruction.
  • the i-th learning target is, for example, a target of learning processing using the i-th template of the template storage unit 613.
  • the i-th learning target is, for example, a target of learning processing corresponding to a template that satisfies the timing indicated by the learning timing parameter included in the template of the template storage unit 613.
  • Step S1906 The learning unit 631 performs a learning process using the i-th learning target. An example of the learning process will be described with reference to the flowchart of FIG.
  • Step S1907 The learning unit 631 increments the counter i by 1. Return to step S1905.
  • Step S1908 The processing unit 63 determines whether or not to perform the prediction processing. If the prediction process is performed, the process goes to step S1909, and if the prediction process is not performed, the process returns to step S1901.
  • a predetermined timing is, for example, a periodic timing and a predetermined timing.
  • the predetermined timing is, for example, when the date and time stored in the storage unit 61 is reached.
  • Information indicating that the predetermined timing matches, for example, the prediction target parameters included in the template for example, "user who installed the application", “user who logged in for the first time", "user who purchased the product for the first time”). (For example, "information that the application has been installed", “user's login ID and password", "purchase instruction of the product" is received.
  • Step S1909 The prediction unit 632 substitutes 1 for the counter i.
  • Step S1910 The prediction unit 632 determines whether or not the i-th prediction target exists. If the i-th prediction target exists, the process proceeds to step S1911, and if the i-th prediction target does not exist, the process returns to step S1901.
  • the i-th prediction target is, for example, a target of prediction processing using a template identified by a template identifier included in a prediction instruction.
  • the prediction target shown in the prediction instruction is, for example, a target of prediction processing using a template paired with the administrator identifier included in the prediction instruction.
  • the i-th prediction target is, for example, a target of prediction processing using the i-th template of the template storage unit 613.
  • the i-th prediction target is, for example, a target of prediction processing corresponding to a template that satisfies the timing indicated by the prediction timing parameter included in the template of the template storage unit 613.
  • Step S1911 The prediction unit 632 performs prediction processing for the i-th prediction target.
  • the prediction process will be described with reference to the flowchart of FIG.
  • Step S1912 The output unit 64 performs output processing. An example of output processing will be described with reference to the flowchart of FIG.
  • Step S1913 The prediction unit 632 increments the counter i by 1. Return to step S1910.
  • step S1906 an example of the learning process in step S1906 will be described with reference to the flowchart of FIG.
  • the learning unit 631 acquires the administrator identifier.
  • the learning unit 631 acquires, for example, the administrator identifier of the received instruction or the like.
  • Step S2002 The learning unit 631 acquires the template identifier. In some cases, the template identifier cannot be obtained. Further, the learning unit 631 acquires, for example, the template identifier of the received instruction or the like.
  • Step S2003 The learning unit 631 determines whether or not there is a template corresponding to the administrator identifier acquired in step S2001, the administrator identification acquired in step S2001, and the template identifier acquired in step S2002. If the template exists, the process goes to step S2004, and if the template does not exist, the process goes to step S2005.
  • Step S2004 The learning unit 631 acquires from the template storage unit 613 a template corresponding to the administrator identifier acquired in step S2001 or the administrator identification acquired in step S2001 and the template identifier acquired in step S2002.
  • Step S2005 The learning unit 631 acquires the default template from the template storage unit 613.
  • Step S2006 The learning unit 631 determines whether or not the learning timing parameter exists in the acquired template. If the learning timing parameter exists, the process goes to step S2007, and if the learning timing parameter does not exist, the process goes to step S2008.
  • Step S2007 The teacher data acquisition means 6311 determines whether or not the timing specified by the learning timing parameter in the template is satisfied (whether or not it is the learning timing). If the timing specified by the learning timing parameter is satisfied, the process proceeds to step S2008, and if not satisfied, the process returns to higher-level processing.
  • Step S2008 The teacher data acquisition means 6311 determines whether or not a parameter for selecting teacher source data exists in the acquired template. If the teacher source data selection parameter exists, the process goes to step S2009, and if it does not exist, the process goes to step S2010.
  • the teacher data acquisition means 6311 acquires the teacher source data corresponding to the teacher source data selection parameter in the template from the teacher source data storage unit 611.
  • Step S2010 The teacher data acquisition means 6311 acquires all the teacher source data of the candidates used in learning from the teacher source data storage unit 611.
  • Step S2011 The teacher data acquisition means 6311 performs a process of acquiring two or more teacher data using two or more teacher source data acquired in step S2009 or step S2010. An example of such a teacher data acquisition process will be described with reference to the flowchart of FIG.
  • Step S2012 The learning means 6312 determines whether or not the learning algorithm parameter exists in the template. If the learning algorithm parameter is present, the process goes to step S2013, and if the learning algorithm parameter is not present, the process goes to step S2020.
  • Step S2013 The learning means 6312 uses the two or more teacher data acquired in step S2011 to perform learning processing by one or two or more algorithms specified by the learning algorithm parameters, and learns one or two or more. Get a vessel.
  • Step S2014 The evaluation means 6313 determines whether or not the learning device evaluation parameter exists in the template. If the learning device evaluation parameter exists, the process goes to step S2015, and if the learning device evaluation parameter does not exist, the process goes to step S2016.
  • Step S2015 The evaluation means 6313 evaluates each of one or more acquired learning devices by an evaluation method according to the learning device evaluation parameters, and acquires the evaluation results.
  • Step S2016 The evaluation means 6313 evaluates each of the acquired 1 or 2 or more learning devices by the default evaluation method, and acquires the evaluation result.
  • Step S2017 The selection means 6314 determines whether or not the learning accuracy parameter exists in the template. If the learning accuracy parameter is present, the process goes to step S2018, and if the learning accuracy parameter is not present, the process goes to step S2019.
  • Step S2018 The selection means 6314 determines whether or not the accuracy of the learning device with the best evaluation result satisfies the learning accuracy parameter. If the learning accuracy parameter is satisfied, the process proceeds to step S2019, and if the learning accuracy parameter is not satisfied, the process returns to higher-level processing.
  • the learning unit 631 stores the learning device with the best evaluation result in the storage unit 61 in association with the administrator identifier, the template identifier, or the administrator identifier and the template identifier. Return to higher-level processing.
  • Step S2020 The learning means 6312 performs a default learning process using the two or more teacher data acquired in step S2011, and acquires a learning device.
  • Step S2021 The learning means 6312 stores the learning device acquired in step S2020 in the storage unit 61 in association with the administrator identifier, the template identifier, or the administrator identifier and the template identifier. Return to higher-level processing.
  • step S2011 an example of the teacher data acquisition process in step S2011 will be described with reference to the flowchart of FIG.
  • Step S2101 The teacher data acquisition means 6311 substitutes 1 for the counter i.
  • Step S2102 The teacher data acquisition means 6311 determines whether or not the i-th teacher source data exists. If the i-th teacher source data exists, the process proceeds to step S2103, and if the i-th teacher source data does not exist, the process returns to higher-level processing.
  • Step S2103 The teacher data acquisition means 6311 substitutes 1 for the counter j.
  • Step S2104 The teacher data acquisition means 6311 determines whether or not the j-th element constituting the teacher data exists. If the j-th element exists, the process goes to step S2105, and if the j-th element does not exist, the process goes to step S2112. The elements that make up the teacher data are predetermined.
  • Step S2105 The teacher data acquisition means 6311 determines whether or not the jth element corresponds to any of the teacher data configuration parameters in the template. If it corresponds to the parameter for teacher data composition, it goes to step S2106, and if it does not correspond to the parameter for teacher data composition, it goes to step S2107.
  • the teacher data acquisition means 6311 is information corresponding to the i-th teacher source data, acquires information corresponding to the j-th element, and uses the information and the teacher data configuration parameter to j. Get the second element.
  • the j-th element is, for example, information obtained by processing one element in the i-th teacher source data according to the parameters for teacher data composition, and two or more elements in the i-th teacher source data are the teacher data composition.
  • the teacher data acquisition means 6311 acquires the j-th element by using one or more information corresponding to the j-th element in the i-th teacher source data.
  • the j-th element is, for example, one element in the i-th teacher source data, information acquired by calculating two or more elements in the i-th teacher source data, and a plurality of teacher elements. This is information obtained by statistically processing (for example, calculating the average value and summing up) one element of the data.
  • Step S2108 The teacher data acquisition means 6311 determines whether the jth element constituting the teacher data is an explanatory variable or an objective variable. If it is an explanatory variable, it goes to step S2109, and if it is an objective variable, it goes to step S2110.
  • Step S2109 The teacher data acquisition means 6311 adds a flag to the jth element to the effect that it is an explanatory variable.
  • Step S2110 The teacher data acquisition means 6311 adds a flag to the jth element to the effect that it is an objective variable.
  • Step S2111 The teacher data acquisition means 6311 increments the counter j by 1. Return to step S2104.
  • the teacher data acquisition means 6311 constructs the i-th teacher data by using one or more elements of the explanatory variable and one or more elements of the objective variable.
  • Step S2113 The teacher data acquisition means 6311 increments the counter i by 1. Return to step S2102.
  • step S1911 an example of the prediction process in step S1911 will be described with reference to the flowchart of FIG. In the flowchart of FIG. 22, the description of the same steps as the flowchart of FIG. 20 will be omitted.
  • Step S2201 The prediction unit 632 determines whether or not the prediction timing parameter exists in the acquired template. If the prediction timing parameter exists, the process goes to step S2202, and if the prediction timing parameter does not exist, the process goes to step S2203.
  • Step S2202 The prediction unit 632 determines whether or not the timing satisfies the prediction timing parameter in the template. If the timing satisfies the prediction timing parameter, the process proceeds to step S2203, and if the timing does not satisfy the prediction timing parameter, the process returns to the higher-level processing.
  • Step S2203 The prediction unit 632 determines whether or not the prediction target parameter exists in the template. If the prediction target parameter exists, the process goes to step S2204, and if the prediction target parameter does not exist, the process goes to step S2205.
  • the prediction unit 632 acquires one or more prediction target data matching the prediction target parameters.
  • the prediction target data has one or more explanatory variables.
  • Step S2205 The prediction unit 632 acquires all target data.
  • Step S2206 The prediction unit 632 substitutes 1 for the counter i.
  • Step S2207 The prediction unit 632 determines whether or not the i-th prediction target data exists. If the i-th prediction target data exists, the process proceeds to step S2208, and if the i-th prediction target data does not exist, the process returns to higher-level processing.
  • Step S2208 The prediction unit 632 determines whether or not the prediction period parameter exists in the template. If the prediction period parameter exists, the process goes to step S2209, and if the prediction period parameter does not exist, the process goes to step S2210.
  • Step S2209 The prediction unit 632 acquires a learning device corresponding to the prediction period parameter.
  • the prediction unit 632 performs prediction processing using the i-th target data and the acquired learning device, and acquires a prediction result according to the prediction period parameter. It is assumed that the learning device corresponding to the prediction period parameter is acquired by the learning unit 631 and stored in the storage unit 61 in association with the prediction period parameter.
  • Step S2210) The prediction unit 632 acquires the default learner used for the prediction process.
  • the prediction unit 632 performs prediction processing using the i-th target data and the acquired learning device, and acquires the prediction result.
  • Step S2211 The prediction unit 632 stores the prediction result acquired in step S2209 or step S2210 in the storage unit 61 in association with the i-th prediction target data.
  • Step S2212 The post-processing unit 633 determines whether or not to perform post-processing. If post-processing is to be performed, the process goes to step S2213, and if no post-processing is to be performed, the process goes to step S2216.
  • Whether or not to perform post-processing may be determined in advance, and the post-processing unit 633 determines whether or not the post-processing parameter exists in the template, and only if it exists, post-processing You may decide to do.
  • Step S2213 The post-processing unit 633 determines whether or not the post-processing parameter exists in the template. If there is a post-processing parameter in the template, go to step S2214, and if there is no post-processing parameter in the template, go to step S2215.
  • the post-processing unit 633 performs post-processing on the prediction result using the post-processing parameters.
  • the post-processing unit 633 stores the output information acquired by the post-processing in the storage unit 61 in association with the i-th prediction target data.
  • Step S2215 The post-processing unit 633 performs default post-processing on the prediction result.
  • the post-processing unit 633 stores the output information acquired by the post-processing in the storage unit 61 in association with the i-th prediction target data.
  • Step S2216 The prediction unit 632 increments the counter i by 1. Return to step S2207.
  • Step S2301 The output unit 64 determines whether or not the output type parameter exists in the template. If the output type parameter exists, the process goes to step S2302, and if the output type parameter does not exist, the process goes to step S2304.
  • Step S2302 The output unit 64 acquires information of one or more types corresponding to the output type parameters in the template.
  • Step S2303 The output unit 64 acquires the default output target information.
  • Step S2304 The output unit 64 determines whether or not the output mode parameter exists in the template. If the output mode parameter is present, the process goes to step S2305, and if the output mode parameter is not present, the process goes to step S2306.
  • Step S2305 The output unit 64 is configured in the template as information in a mode in which the information acquired in step S2302 or step S2304 is output according to the output mode parameter.
  • Step S2306 The output unit 64 configures the acquired information as the information of the default output mode.
  • Step S2307 The output unit 64 determines whether or not the output destination parameter exists in the template. If the output destination parameter exists, the process goes to step S2308, and if the output destination parameter does not exist, the process goes to step S2309.
  • Step S2308 The output unit 64 outputs the acquired information first according to the output destination parameter.
  • Step S2309 The output unit 64 outputs the acquired information to the default output destination.
  • information may be output to a different medium according to the output medium parameter in the template.
  • Step S2401 The terminal reception unit 72 determines whether or not the template or the like has been accepted from the administrator. If the template or the like is accepted, the process goes to step S2402, and if the template or the like is not accepted, the process goes to step S2403.
  • Step S2402 The terminal processing unit 73 has a data structure for transmitting a template or the like.
  • the terminal transmission unit 74 transmits the configured template and the like to the information processing device 6.
  • the transmitted template is stored in the information processing device 6.
  • Step S2403 The terminal reception unit 72 determines whether or not the learning instruction has been received from the administrator. If the learning instruction is accepted, the process goes to step S2404, and if the learning instruction is not accepted, the process goes to step S2405.
  • Step S2404 The terminal processing unit 73 has a data structure for transmitting a learning instruction.
  • the terminal transmission unit 74 transmits the configured learning instruction to the information processing device 6. Return to step S2401. By transmitting the learning instruction, the learning device is configured and stored in the information processing device 6.
  • Step S2405 The terminal reception unit 72 determines whether or not the prediction instruction has been received from the administrator. If the prediction instruction is received, the process goes to step S2406, and if the prediction instruction is not received, the process returns to step S2401.
  • the terminal processing unit 73 has a data structure for transmitting a prediction instruction.
  • the terminal transmission unit 74 transmits the configured prediction instruction to the information processing device 6.
  • Step S2407 The terminal receiving unit 75 determines whether or not the output information or the like has been received from the information processing apparatus 6 in response to the transmission of the prediction instruction in step S2406. If the output information or the like is received, the process proceeds to step S2408, and if the output information or the like is not received, the process returns to step S2407.
  • Step S2408 The terminal processing unit 73 has a data structure for outputting output information and the like.
  • the terminal output unit 76 outputs the configured output information and the like. Return to step S2401.
  • the information processing apparatus 6 detects and stores the attribute values of a visitor to the website (hereinafter, appropriately referred to as “user”) in real time, and accumulates the attribute values of the server apparatus 2 of the first embodiment.
  • Use information As described above, the server device 2 receives and stores one or more operation information from the user terminal 1 of the user who is a visitor to the website.
  • the operation information may be primitive operation information (for example, "rightButtonON"), or information for which the meaning and significance of the operation can be determined (for example, "login”, “logout”, "purchased product A", “purchase product A”. Move page X "" XX page was displayed ”) is also acceptable.
  • the operation information accumulated by the server device 2 constitutes the teacher source data of the information processing device 6. That is, it is assumed that the teacher source data management table shown in FIG. 25 is now stored in the teacher source data storage unit 611 of the information processing apparatus 6.
  • the teacher source data management table may have the same structure as the operation information management table of FIG.
  • the “data source” in the teacher source data management table shown in FIG. 25 indicates the source in which the operation corresponding to the operation information is performed, and here, the information of “web” or “app” can be taken.
  • Web indicates that an operation has been performed on a web page.
  • “App” indicates that an operation has been performed on the application.
  • the information processing device 6 may have all or part of the functions of the server device 2. That is, according to the operation of the information system A described in the first embodiment, the teacher source data management table shown in FIG. 25 may be stored in the teacher source data storage unit 611 of the information processing apparatus 6. Further, the teacher source data management table of the teacher source data storage unit 611 of the information processing apparatus 6 may be received from the server device 2.
  • the template management table is a table that manages templates.
  • the template management table manages one or more records having "ID”, "administrator identifier", “template identifier”, and "template”.
  • the "ID” is information for identifying a record.
  • the "administrator identifier” is the identifier of the administrator who uses the template.
  • the “administrator identifier” may be the identifier of the administrator who holds the template.
  • the “template identifier” is the identifier of the template, and here is the name of the template.
  • the “template” here is the file name of the template. Further, it is assumed that the template file corresponding to the "template” is stored in the template storage unit 613.
  • Specific example 1 is a specific example in which templates are accumulated.
  • Specific example 2 is a specific example in which the information processing apparatus 6 classifies users.
  • Specific example 3 is a specific example in which the information processing apparatus 6 forecasts the demand for a product.
  • the administrator 1 identified by the administrator identifier "M01" inputs the information constituting the template to the screen of FIG. 27 displayed on the terminal device 7 as shown in FIG. 27. Yes.
  • the input screen and its method may be input using an editor such as yaml, or may be input hierarchically using the editor. In other words, it goes without saying that it does not matter how the information that constitutes the template is given.
  • the terminal reception unit 72 receives the template from the administrator 1. Then, the terminal processing unit 73 has a data structure for transmitting the template and the administrator identifier "M01". Next, the terminal transmission unit 74 transmits the configured template and the administrator identifier “M01” to the information processing device 6.
  • the reception unit 62 of the information processing device 6 receives the template and the like.
  • the processing unit 63 constructs a file using the received template information, and stores the file in the template storage unit 613.
  • FIG. 28 an example of the template of "tp_m01_1.txt” is shown in FIG. 28.
  • “ ⁇ learning timing parameter> reception of learning instruction” indicates that the learning unit 631 performs the learning process with the reception of the learning instruction as a trigger.
  • the learning process is performed using the teacher source data during the period of "2019/06/01 to 2019/12/31".
  • “-” Such as “ ⁇ teacher user selection parameter>-” indicates that there is no parameter specification, default data is used, default processing is performed, and the like.
  • “ ⁇ Data source parameter> web” indicates that the data source performs the learning process using the teacher source data of "web”.
  • “ ⁇ Aggregation period parameter> 30” is the information used as the feature (feature or feature quantity) of the teacher data, and each information as a result of statistical processing is after the first purchase (first time). It is shown that the information is the result of statistical processing using the teacher source data for 30 days (after the purchase of the product).
  • " ⁇ Number of days for totaling correct answer data> 60” indicates that the total number of days for calculating correct answer data (here, a predetermined "number of purchases”) is 60 days.
  • “ ⁇ correct answer data totaling days> 60” is information used when acquiring an objective variable, and is used not only for prediction processing but also for learning processing.
  • " ⁇ Forecast period parameter> 2020-01-01 to 2020-03-01” indicates that the date for performing the forecast process is between "2020-01-01 and 2020-03-01".
  • " ⁇ Threshold value for royal determination> 2" indicates that the user who is determined to be a royal customer is a user who has purchased two or more times.
  • " ⁇ Output type parameter> Royal customer user identifier” indicates that the information is to output the royal customer user identifier.
  • the output information is "royal customer user identifier (character string)” and "daily bar graph of the number of royal customer user identifiers”. Is shown.
  • " ⁇ Output medium parameter> HDD, mail” indicates that the output information is stored in the HDD and transmitted by mail.
  • " ⁇ HDD> / root / admin / x /” indicates the storage destination in the HDD.
  • " ⁇ Email> admin1@x.jp, admin2@x.jp” indicates the email address of the destination of the output information.
  • the reception unit 62 of the information processing device 6 receives an operation instruction having the administrator identifier "M01" and the template identifier "Royal customer classification" from the terminal device 7. It is assumed that the operation instruction includes a learning instruction and a prediction instruction.
  • the information processing device 6 operates as follows. That is, the learning unit 631 of the information processing apparatus 6 performs the learning process using the teacher source data of FIG. 25 based on the template of FIG. 28.
  • the processing unit 63 refers to the template " ⁇ learning timing parameter> reception of learning instruction" and determines that the learning process is performed. It is assumed that the reception of the operation instruction also corresponds to the reception of the learning instruction.
  • the teacher data acquisition means 6311 of the learning unit 631 refers to " ⁇ learning period parameter> 2019-06-01 to 2019-12-31” and " ⁇ data source parameter> web", and the date and time is "2019-06".
  • the teacher source data whose data source is "web” and which is the date and time within the period of "-01 to 2019-12-31" is acquired from the teacher source data management table (FIG. 25).
  • the teacher data acquisition means 6311 refers to the template " ⁇ aggregation period parameter> 30 days” and acquires the dynamic attribute value to be statistically processed. That is, the teacher data acquisition means 6311 statistically processes one or more acquired teacher source data for each user, and "visits" for 30 days after the first purchase and “total” for 30 days after the first purchase. "Purchase amount (yen)”, “Average number of PVs” for 30 days after the first purchase, “Session average interval” for 30 days after the first purchase, “Average stay time” for 30 days after the first purchase, After the first purchase Acquire explanatory variables such as "average purchase unit price” for 30 days. It should be noted that the process of acquiring the "number of visits" and the like for 30 days after the initial purchase from the teacher source data (operation information) as shown in FIG. 25 can be performed by a known technique.
  • the teacher data acquisition means 6311 acquires a predetermined static attribute value (for example, gender, age, member, etc.) constituting an explanatory variable for each user.
  • a predetermined static attribute value for example, gender, age, member, etc.
  • the teacher data acquisition means 6311 refers to " ⁇ correct answer data totaling days> 60" of the template, and acquires correct answer data for 60 days (here, "number of purchases") for each user. This "number of purchases" is an objective variable. Further, it is assumed that the teacher data acquisition means 6311 knows (programmed) in advance that the correct answer data is the "number of purchases”.
  • the teacher data acquisition means 6311 configures teacher data having two or more explanatory variables and one objective variable for each user.
  • An example of such teacher data for each user is shown in FIG. 29.
  • the teacher data has a "static attribute value” and a "dynamic attribute value”.
  • the "dynamic attribute value” is, here, the "history information utilization dynamic attribute value”.
  • FIG. 29 is a teacher data management table.
  • the learning means 6312 performs a learning process using the teacher data of a plurality of users to form a learning device.
  • the learning means 6312 associates the learning device with the template and stores it in the storage unit 61.
  • the prediction unit 632 refers to the template " ⁇ Prediction period parameter> 2020-01-01 to 2020-03-01", and each day of the period "2020-01-01 to 2020-03-01".
  • the number of purchases (cumulative) of a user is predicted for each user and each day, and the prediction result for each user and each day is acquired.
  • the post-processing unit 633 uses the number of purchases for each user and each day and " ⁇ threshold value for royal determination> 2" to determine whether or not each user is a "royal customer” for each day. Whether or not the number of purchases (cumulative) is predicted to be 2 or more) is determined. Then, the post-processing unit 633 refers to the template " ⁇ output type parameter> royal customer user identifier" and, on a daily basis. Get the user identifier that is a royal customer. The user identifier for each day is output information here. The post-processing unit 633 may acquire the user identifier of the user who becomes a new royal customer on a daily basis.
  • the output unit 64 refers to the template " ⁇ output mode parameter> character string, daily bar graph", and constitutes a character string of "daily user identifier" and a daily bar graph. Further, the output unit 64 refers to " ⁇ output medium parameter> HDD, mail" and " ⁇ output destination parameter> ⁇ HDD> / root / admin / x / ⁇ mail> admin1@x.jp, admin2@x.jp". Then, the configured character strings and daily bar graphs are stored in the HDD folder "/ root / admin / x /" and sent by e-mail to two people, "admin1@x.jp and admin2@x.jp".
  • the daily bar graph is, for example, a bar graph having the day on the horizontal axis and the number of royal customers on the vertical axis, and the technique for forming such a bar graph is a known technique. Further, the number of royal customers may be the cumulative number of royal customers or the number of new royal customers.
  • the " ⁇ teacher user selection parameter> product A" in the template of FIG. 30 indicates that the teacher source data indicating the sale of the product A is selected.
  • " ⁇ Learning period parameter> 2019-06-01 to 2019-12-31” indicates that the teacher source data for the period of "2019/06/01 to 2019/12/31” is selected.
  • the template " ⁇ aggregation period parameter> 60" is the information used as the features (features or feature quantities) of the teacher data, and each information as a result of statistical processing is after the initial purchase. It is shown that the information is the result of statistical processing using the teacher source data for 60 days (after the first purchase of the product).
  • “ ⁇ Smoothness> 5" indicates that the degree of smoothing in statistical processing is "5".
  • ⁇ Learning accuracy parameter> 0.7 indicates that the learning device is not used unless the accuracy when the learning device is evaluated is “0.7” or more.
  • ⁇ Forecast period parameter> 1 month later indicates that the demand forecast is to perform the demand forecast 1 month later.
  • ⁇ Output type parameter> prediction result indicates that the output information is a prediction result.
  • the forecast result here is a forecast result of demand, and is a forecast sales number of product A one month later.
  • the information processing apparatus 6 searches the template management table of FIG. 26 using the administrator identifier "M01" as a key, selects two templates, configures a screen for selecting one template from the two templates, and configures a terminal device. Send to 7.
  • the terminal device 7 receives and outputs a template selection screen.
  • a template selection screen An example of such a screen is shown in FIG. In FIG. 31, the menu items (3101, 3102) of the two templates of 3101 “Royal customer classification” and “Demand forecast” are displayed.
  • the terminal device 7 receives the operation instruction having the template identifier of the selected "demand forecast” and transmits the operation instruction to the information processing device 6.
  • the information processing apparatus 6 receives an operation instruction having a "demand forecast”. Next, the information processing apparatus 6 performs learning processing and prediction processing according to the template “demand forecast”.
  • the teacher data acquisition means 6311 of the learning unit 631 refers to the template "demand forecast”, and is the teacher source data for the period of "2019/06/01 to 2019/12/31", and is the "product A”.
  • the teacher source data indicating the purchase of is selected from the teacher source data management table of FIG.
  • the teacher data acquisition means 6311 refers to " ⁇ learning period parameter> 2019-06-01 to 2019-12-31" and " ⁇ aggregation period parameter> 60" of the template, and "2019-06-01 to”. For each day of the period indicated by "2019-12-31", each information of 1 or more for 60 days with the day as the last day is statistically processed, and 1 or more dynamic attribute values are acquired. That is, the teacher data acquisition means 6311 has a 60-day “visit count”, a 60-day “total purchase amount (yen)", and a 60-day "average PV" from one or more teacher source data acquired for each day.
  • the teacher data acquisition means 6311 refers to the template " ⁇ learning period parameter> 2019-06-01 to 2019-12-31" and " ⁇ predicted period parameter> one month later" to refer to the teacher source data management table. By using it, the number of sales of "Product A" for one day after one month is acquired for each day of the period indicated by "2019-06-01 to 2019-12-31". The number of such sales is an objective variable.
  • the teacher data acquisition means 6311 constitutes teacher data having two or more explanatory variables and one objective variable for each day.
  • the learning means 6312 performs a learning process using the teacher data of a plurality of days to form a learning device.
  • the learning means 6312 associates the learning device with the template and stores it in the storage unit 61.
  • the evaluation means 6313 refers to the template " ⁇ learning accuracy parameter> 0.7", and a value is entered in any of the evaluation-related parameters " ⁇ learning device evaluation parameter>” and " ⁇ learning accuracy parameter>”. Therefore, it is judged that the learner is evaluated using the given parameters.
  • the evaluation means 6313 evaluates the acquired learner by using, for example, the K division method which is the default evaluation method, and obtains an accuracy of "0.91".
  • the selection means 6314 refers to the template " ⁇ learning accuracy parameter> 0.7", and the accuracy "0.91" acquired by the evaluation means 6313 satisfies "0.7 or more”, so that this gear learning Judge that the vessel is usable.
  • the prediction unit 632 refers to the template " ⁇ aggregation period parameter> 60" and " ⁇ prediction period parameter> one month later", and acquires two or more explanatory variables from the teacher source data of the last 60 days up to yesterday. do.
  • the prediction unit 632 gives the accumulated learning device and the two or more explanatory variables to the module that performs the prediction processing, executes the module, and acquires the prediction result (for example, "1286").
  • the forecast result is a sales forecast for product A one month later.
  • the output unit 64 outputs the prediction result "1286" according to the template " ⁇ output type parameter> prediction result".
  • the output unit 64 has a forecast result "The number of sales forecasts for product A one month later is" 1286 ". Is transmitted to the terminal device 7.
  • the forecast result is "The number of sales forecasts for product A one month later is" 1286 ". Is received and output.
  • the learning process and the prediction process when the learning process and the prediction process are performed by the machine learning algorithm, the learning process and the prediction process that can be flexibly customized can be performed by using the template having various parameters to be used.
  • learning processing and prediction processing can be performed using different templates for each administrator who uses the information processing device.
  • the learning process and the prediction process can be easily performed by using the template selected by the user.
  • the learning period parameter used in the teacher data selection process which is the pre-process of the learning process, can be specified by the template.
  • the method parameters used in the post-processing of the learning process can be specified by the template.
  • the prediction period parameter used in the prediction processing can be specified by the template.
  • the timing for performing the prediction process can be specified by the template.
  • the post-processing parameters used in the post-processing after the prediction processing can be specified by the template.
  • a prediction result including a user's category it is possible to acquire a prediction result including a user's category.
  • a classification prediction of a user who purchases a product it is possible to obtain a classification prediction of a user who purchases a product.
  • the prediction result including the category of the user can be acquired by performing the learning process and the prediction process using one or more static attribute values of the user.
  • the template can be distributed. That is, for example, a template of one administrator may be used by another person to perform learning processing and prediction processing.
  • the processing in the present embodiment may be realized by software. Then, this software may be distributed by software download or the like. Further, this software may be recorded on a recording medium such as a CD-ROM and disseminated. It should be noted that this also applies to other embodiments herein.
  • the software that realizes the information processing device 6 in this embodiment is the following program. That is, this program has a template storage unit that stores a template having one or more learning parameters that are parameters related to machine learning learning processing and one or more prediction parameters that are parameters related to machine learning prediction processing, and teacher data.
  • the template has a computer that can access a teacher source data storage unit that stores one or more original teacher source data and a target data storage unit that stores prediction target data that is prediction target data.
  • Learning processing is performed from one or more teacher source data that acquire the one or more learning parameters, are one or more teacher source data corresponding to the one or more learning parameters, and are stored in the teacher source data storage unit.
  • prediction processing is performed on the prediction target data stored in the target data storage unit, and the prediction result is acquired. It is a program for functioning as a prediction unit and an output unit that outputs the prediction result.
  • FIG. 32 shows the appearance of a computer that executes the program described in the present specification to realize the information processing apparatus 6 and the like in various embodiments described above.
  • the above-described embodiment can be realized by computer hardware and a computer program executed on the computer hardware.
  • 32 is an overview view of the computer system 300
  • FIG. 33 is a block diagram of the system 300.
  • the computer system 300 includes a computer 301 including a CD-ROM drive, a keyboard 302, a mouse 303, and a monitor 304.
  • the computer 301 in addition to the CD-ROM drive 3012, the computer 301 includes an MPU 3013, a bus 3014 connected to the CD-ROM drive 3012 and the like, a ROM 3015 for storing a program such as a boot-up program, and an MPU 3013. It includes a RAM 3016 that is connected and for temporarily storing instructions of an application program and providing a temporary storage space, and a hard disk 3017 for storing an application program, a system program, and data.
  • the computer 301 may further include a network card that provides a connection to the LAN.
  • the program for causing the computer system 300 to execute the functions of the information processing apparatus 6 and the like according to the above-described embodiment may be stored in the CD-ROM 3101, inserted into the CD-ROM drive 3012, and further transferred to the hard disk 3017. .. Alternatively, the program may be transmitted to the computer 301 via a network (not shown) and stored in the hard disk 3017. The program is loaded into RAM 3016 at run time. The program may be loaded directly from the CD-ROM3101 or the network.
  • the program does not necessarily have to include an operating system (OS) or a third-party program that causes the computer 301 to execute the functions of the information processing apparatus 6 and the like according to the above-described embodiment.
  • the program only needs to include a part of instructions that call the appropriate function (module) in a controlled manner and obtain the desired result. It is well known how the computer system 300 works, and detailed description thereof will be omitted.
  • the processing performed by the hardware for example, the processing performed by the modem or the interface card in the transmission step (only performed by the hardware). Processing) is not included.
  • the number of computers that execute the above program may be singular or plural. That is, centralized processing may be performed, or distributed processing may be performed.
  • the two or more communication means existing in one device may be physically realized by one medium.
  • each process may be realized by centralized processing by a single device, or may be realized by distributed processing by a plurality of devices.
  • the information processing apparatus has the effect of being able to flexibly customize learning processing and prediction processing by using a template, and is useful as a server device or the like that performs various prediction processing. be.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

Le problème à résoudre par la présente invention est que, de manière classique, il est impossible d'exécuter un traitement d'apprentissage et un traitement de prédiction qui sont personnalisables de manière flexible. La solution selon l'invention porte sur un dispositif de traitement d'informations comprenant : une unité d'apprentissage qui acquiert des données d'entraînement destinées à être utilisées dans un traitement d'apprentissage à partir d'un ou de plusieurs éléments de données de source d'entraînement correspondant à un ou plusieurs paramètres d'apprentissage d'un modèle, le modèle comprenant un ou plusieurs paramètres d'apprentissage qui sont des paramètres liés au traitement d'apprentissage dans l'apprentissage machine et un ou plusieurs paramètres de prédiction qui sont des paramètres liés au traitement de prédiction dans l'apprentissage machine, qui exécute un traitement d'apprentissage sur un ou plusieurs éléments de données d'entraînement, et qui acquiert un outil d'apprentissage ; une unité de prédiction qui exécute un traitement de prédiction sur des données pour une prédiction à l'aide d'un ou de plusieurs paramètres de prédiction du modèle et de l'outil d'apprentissage acquis par l'unité d'apprentissage, et qui acquiert un résultat de prédiction ; et une unité de sortie qui émet le résultat de prédiction. Avec le dispositif de traitement d'informations, il est possible d'exécuter un traitement d'apprentissage et un traitement de prédiction qui sont personnalisables de manière flexible à l'aide du modèle.
PCT/JP2021/014700 2020-06-09 2021-04-07 Dispositif de traitement d'informations, procédé de traitement d'informations et support d'enregistrement WO2021250987A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2020-100335 2020-06-09
JP2020100335A JP6893010B1 (ja) 2020-06-09 2020-06-09 情報処理装置、情報処理方法、およびプログラム

Publications (1)

Publication Number Publication Date
WO2021250987A1 true WO2021250987A1 (fr) 2021-12-16

Family

ID=76464561

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2021/014700 WO2021250987A1 (fr) 2020-06-09 2021-04-07 Dispositif de traitement d'informations, procédé de traitement d'informations et support d'enregistrement

Country Status (2)

Country Link
JP (2) JP6893010B1 (fr)
WO (1) WO2021250987A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114925034A (zh) * 2022-05-05 2022-08-19 五八有限公司 异构数据处理方法、装置、电子设备及存储介质
JP7313615B1 (ja) * 2023-03-22 2023-07-25 コグニロボ株式会社 プログラム、製造計画装置、および製造計画方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001325582A (ja) * 2000-05-17 2001-11-22 Chugoku Electric Power Co Inc:The 時系列データ学習・予測装置
JP2009187058A (ja) * 2008-02-01 2009-08-20 Fujitsu Ltd 情報処理装置、情報処理方法及びコンピュータプログラム
WO2017159638A1 (fr) * 2016-03-14 2017-09-21 オムロン株式会社 Dispositif de génération de données de communication de capacité
WO2019032156A1 (fr) * 2017-08-11 2019-02-14 Google Llc Plateforme d'apprentissage machine sur dispositif
JP2020052514A (ja) * 2018-09-25 2020-04-02 日本電気株式会社 ＡＩ（ａｒｔｉｆｉｃｉａｌ ｉｎｔｅｌｌｉｇｅｎｃｅ）実行支援装置、方法、及びプログラム

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7102801B2 (ja) * 2018-03-14 2022-07-20 オムロン株式会社 制御システム、制御装置および表示装置
DE112018007550T5 (de) * 2018-06-05 2021-01-28 Mitsubishi Electric Corporation Lernvorrichtung, Inferenzvorrichtung, Verfahren und Programm
JP6596565B1 (ja) * 2018-12-27 2019-10-23 国立大学法人 東京大学 抽出システムおよびプログラム

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001325582A (ja) * 2000-05-17 2001-11-22 Chugoku Electric Power Co Inc:The 時系列データ学習・予測装置
JP2009187058A (ja) * 2008-02-01 2009-08-20 Fujitsu Ltd 情報処理装置、情報処理方法及びコンピュータプログラム
WO2017159638A1 (fr) * 2016-03-14 2017-09-21 オムロン株式会社 Dispositif de génération de données de communication de capacité
WO2019032156A1 (fr) * 2017-08-11 2019-02-14 Google Llc Plateforme d'apprentissage machine sur dispositif
JP2020052514A (ja) * 2018-09-25 2020-04-02 日本電気株式会社 ＡＩ（ａｒｔｉｆｉｃｉａｌ ｉｎｔｅｌｌｉｇｅｎｃｅ）実行支援装置、方法、及びプログラム

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114925034A (zh) * 2022-05-05 2022-08-19 五八有限公司 异构数据处理方法、装置、电子设备及存储介质
JP7313615B1 (ja) * 2023-03-22 2023-07-25 コグニロボ株式会社 プログラム、製造計画装置、および製造計画方法

Also Published As

Publication number Publication date
JP2021196657A (ja) 2021-12-27
JP6893010B1 (ja) 2021-06-23
JP2021197166A (ja) 2021-12-27

Similar Documents

Publication Publication Date Title
US10783563B2 (en) Methods and systems for modeling campaign goal adjustment
US10936963B2 (en) Systems and methods for content response prediction
Bhagat et al. Buy it again: Modeling repeat purchase recommendations
Chen et al. Fuzzy regression-based mathematical programming model for quality function deployment
Wang et al. Collaborative filtering with entropy‐driven user similarity in recommender systems
Patiniotakis et al. PuLSaR: preference-based cloud service selection for cloud service brokers
Cao et al. Customer demand prediction of service-oriented manufacturing incorporating customer satisfaction
WO2021250987A1 (fr) Dispositif de traitement d'informations, procédé de traitement d'informations et support d'enregistrement
JP2018077615A (ja) 広告画像生成装置、広告画像生成方法、広告画像生成装置用プログラム
JP2011141833A (ja) 株価予測装置、方法及びプログラム
US20160117703A1 (en) Large-Scale Customer-Product Relationship Mapping and Contact Scheduling
Wang et al. A personalized electronic movie recommendation system based on support vector machine and improved particle swarm optimization
CN113065882A (zh) 一种商品处理方法、装置及电子设备
US8843428B2 (en) Survey prioritization engine
JP2016062275A (ja) 抽出装置、抽出方法および抽出プログラム
Luo et al. Optimizing customer's selection for configurable product in B2C e-commerce application
Zou et al. Group interaction and evolution of customer reviews based on opinion dynamics towards product redesign
US20160085758A1 (en) Interest-based search optimization
CA2909957A1 (fr) Cartographie de la relation consommateur-produit grande echelle et planification de communication
JP6752330B1 (ja) 機械学習ベースのマッチング装置およびマッチング方法
JP6018541B2 (ja) 推薦ルール生成装置、推薦ルール生成方法および推薦ルール生成プログラム
JP2016164750A (ja) 選択行動モデル化装置、選択行動予測装置、方法、及びプログラム
JP6887648B1 (ja) ウェブページ処理装置、ウェブページ処理方法、およびプログラム
Lu et al. Srrl: select reliable friends for social recommendation with reinforcement learning
CN112334931A (zh) 信息处理设备、信息处理方法和程序

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21821596

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21821596

Country of ref document: EP

Kind code of ref document: A1