JP2014134940A - Information processing device, recommendation information generation method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information processing device capable of generating recommendation information regarding a price with respect to a service according one aspect.SOLUTION: A data correlation analysis section 13 computes correlation values of a plurality of parameters of a service A and a service B. On the basis of the computed correlation values, combinations (ra, rb) of correlation values ra of the service A and correlation values rb of the service B having one parameter in common are extracted. Among the extracted combinations (ra, rb), a combination (ra, rb) maximizing a sum ra+rb of the correlation values belonging to the combination is determined. On the basis of the combination (ra, rb) maximizing the sum, an estimate calculation section 14 generates recommendation information 6 of payment amounts regarding the service A and the service B.

Description

  The present invention relates to an information processing apparatus, a recommendation information generation method, and a program.

  In a marketplace that is a trading market on a network such as the Internet, there are a plurality of services having similar functions. Therefore, an apparatus for recommending a better and cheaper service to the user has been developed.

  For example, the recommendation device is usage history information indicating usage histories of services used by a plurality of users, a plurality of user attributes indicating the characteristics of the users, and a plurality of service attributes indicating the characteristics of the services selected by the user, , Based on the usage history information that associates the service attribute with the user attribute model that indicates the characteristics of the user model that is estimated to be desired for the service attribute. It is proposed to calculate a service attribute model indicating the characteristics of the service model estimated to be desired by the requesting user based on the recommended basic information calculated by the above and the user attribute of the requesting user.

  Also, the detailed data obtained from the telecommunications company is converted into a unique format, and the usage details are checked in comparison with the PBX communication record, the communication fee is calculated using the fee calculation table, and the billing amount is checked. The format-converted detail data is corrected by PBX communication records, and the information of the in-house server log is added, and processing such as tabulation, communication check, simulation, etc. is performed using the reinforcement detail history data accumulated over time. Proposed.

  In addition, a plurality of users who ask for storage locations of luggage, a plurality of warehouse companies (service provider companies) that ask for luggage to be stored, warehouse request information (service application) from users, and demand information (services) from warehouse companies Provided information) via a server on a network such as the Internet, and information on the package using the information from the demand / recruitment information processing means. It has been proposed to make a general configuration with a service provider that establishes a storage service.

Japanese Patent No. 4552230 JP 2001-297264 A JP 2002-092133 A

  The recommendation information provided to the user is created and displayed based on static information at a specific time, such as a purchase history. Also, the recommendation information provided to the user is created and displayed based on the service purchase tendency of other users who have purchased the same service.

  However, the recommendation information provided to the user is only based on the preference of the user and the purchase tendency of other users. Therefore, the recommendation information does not include a price evaluation for the service. This is because the contents of similar services on the market place are slightly different, and the charging method is not unified for time charging or user number charging. As a result, the user cannot obtain recommendation information about the price for the service. For this reason, the user must evaluate a price for a similar service on the marketplace based only on his / her judgment, and effectively cannot select a service based on the price evaluation appropriately. .

  An object of the present invention is, according to one aspect, to provide an information processing apparatus capable of generating recommendation information about a price for a service.

  According to one aspect, the information processing apparatus includes a collection unit, an analysis unit, and a calculation unit. The collection unit obtains usage records for the first service and the second service. The analysis unit obtains a correlation value between each of the plurality of parameters of each service for the first service and the second service based on the acquired usage record, and based on the obtained correlation value A combination of the correlation value of the first service and the correlation value of the second service, the combination of the correlation values having one parameter in common, and extracting the combination in the extracted combination Find the combination that maximizes the sum of the correlation values belonging to. The collection unit generates recommendation information of a payment amount that is a consideration for the service for the first service and the second service, based on the combination that maximizes the calculated sum.

  According to one aspect, the information processing apparatus can generate recommendation information about a price for a service.

It is a figure which shows an example of an information processing system. It is a figure which shows an example of information processing apparatus. It is a figure which shows an example of information processing apparatus. It is explanatory drawing of a correlation process. It is explanatory drawing of a correlation process. It is explanatory drawing of a correlation process. It is explanatory drawing of a correlation process. It is a figure which shows an example of recommendation information. It is a figure which shows an example of the hardware constitutions of information processing apparatus. It is a recommendation information generation process flow. It is a recommendation information generation process flow. It is a recommendation information generation process flow.

  FIG. 1 is a diagram illustrating an example of an information processing system.

  The information processing system is a computer network, for example, and includes a user management server 1, a plurality of terminals 2, a service management server 3, a plurality of service providing apparatuses 4, and a network 5. The network 5 is, for example, the Internet, and connects the user management server 1, the terminal 2, the service management server 3, and the service providing device 4 to each other.

  The user management server 1 and the plurality of terminals 2 are management computers belonging to one organization user such as a company. The user management server 1 is a web server, for example, and manages a plurality of terminals 2. Each of the plurality of terminals 2 is a client device belonging to, for example, an organization user, and is a computer used by a user who uses the service. A user including a user management server 1 that is a management computer and a plurality of terminals 2 that are computers used by the user is referred to as an organization user.

  Note that the user management server 1 may include a function of the user who uses the service, in other words, the terminal 2. Further, the user management server 1 and each of the plurality of terminals 2 may be connected by a network other than the network 5.

  The service is provided, for example, at a marketplace. The marketplace is a trading market for services provided on the network 5 such as the Internet, for example. Accordingly, the service is, for example, a service on the Internet, and the user is an Internet user. The user actually uses the service by the web client device used by the user.

  The service management server 3 is a web server, for example, and manages a plurality of services provided on the network 5. Each of the plurality of service providing apparatuses 4 is, for example, a client apparatus, and provides a service on the network 5 via the service management server 3. Each service providing device 4 may be a web server, for example, and may provide a service on the network 5 without going through the service management server 3.

  The services provided by each service providing apparatus 4 are similar, but are different from each other, and the price of the service is also different. For example, a certain service providing apparatus 4 provides a service A, and a certain service providing apparatus 4 provides a service B. Accordingly, service contents differ among the plurality of service providing apparatuses 4 and the accounting method is not unified. For example, time billing, number-of-users billing, or the like is adopted between the plurality of service providing apparatuses 4, and billing parameters are different.

  2 and 3 are diagrams illustrating an example of the information processing apparatus. FIG. 2 mainly shows the relationship among the user management server 1, terminal 2, service management server 3, and service providing apparatus 4. FIG. 3 mainly shows the configurations of the user management server 1 and the service management server 3.

  As shown in FIGS. 2 and 3, the user management server 1 includes an event collection unit 11, a usage record storage unit 12, a data correlation analysis unit 13, and an approximate calculation unit 14. As shown in FIG. 3, the service management server 3 includes a management processing unit 31 and a user information storage unit 32.

  In FIG. 2, the service management server 3 transmits a service catalog, which is a list of services provided by the plurality of service providing apparatuses 4, to the terminal 2 used by the user. The service management server 3 generates a service catalog by registering services provided from a plurality of service providing apparatuses 4, for example.

  The terminal 2 that has received the service catalog displays the received service catalog. When the user who has viewed the displayed service catalog selects a service from the service catalog, the terminal 2 transmits an application for use of the selected service to the service management server 3.

  The service management server 3 transmits the service use application received from the terminal 2 to the service providing apparatus 4 that provides the service. Upon receiving the service use application, the service providing apparatus 4 examines the received service use application and permits the use of the service. As a result, the terminal 2 can start using the service provided by the service providing apparatus 4. The terminal 2 is provided with the selected service from the service providing apparatus 4 or from the service management server 3.

  In the service management server 3, the management processing unit 31 manages the service catalog, service use application, and service use permission, thereby obtaining management data regarding service provision and storing it in the user information storage unit 32. . The management data is acquired, for example, every time the service is provided, in other words, triggered by the use of the service using the terminal 2 by the user. The management processing unit 31 is, for example, a Systemwalker Service Catalog Manager (CTMG). The user information storage unit 32 is, for example, a user information database.

  The management data is acquired by the user management server 1 and used as a usage record. The management data is, for example, usage time, the number of registered users, the number of events, and data unique to the service for the provided service. The usage time, the number of registered users, the number of events, and the data specific to the service are parameters for use of the service. The value of the usage time, the number of registered users, the value of the number of events, and the value of the data specific to the service are The value of the parameter. The parameter value is sample data. The usage time, the number of registered users, the number of events, and data unique to the service will be described later with reference to FIG.

  For example, as illustrated in FIG. 2, when the service B is charged for the usage time, the management processing unit 31 acquires management data at the timing when the user starts using the service B from the terminal 2. Specifically, the management processing unit 31 acquires the number of registered users, which is the number of users in the organization user to which the terminal 2 belongs, at the timing when the user starts using the service B. In addition, the management processing unit 31 counts the usage time of the service by the terminal 2 at the timing when the user starts using the service B, adds “+1” to the value at that time, and stores the data unique to the service. Get the value. Then, the management processing unit 31 holds the value of the parameter, the value of the usage time, the value of the number of registered users, the value of the number of events, and the value of the data unique to the service, as sample data.

  The service management server 3 may be the same server as the user management server 1. In other words, the user management server 1 may include the management processing unit 31 and the user information storage unit 32. The service management server 3 may be a server belonging to an organization user managed by the user management server 1 and managed by the user management server 1.

  The user management server 1 is a server that manages organization users, collects management data from the service management server 3, and manages a plurality of terminals 2 inside the organization user using the collected management data. The user management server 1 is a recommendation information generation device that generates recommendation information 6.

  In the user management server 1, as shown in FIG. 3, the event collection unit 11 is an example of a collection unit, and a plurality of predetermined services, for example, service A and service B, for which usage records are acquired. Get usage records. The usage record is acquired from the service management server 3, for example. The event collection unit 11 may collect usage records from a plurality of terminals 2, or may collect usage records by including functions similar to those of the service management server 3.

  Whether or not a plurality of services are similar can be known in advance. In other words, the plurality of services for which usage records are acquired are similar services and are determined in advance. Service A is a first service, and service B is a second service. Note that the plurality of predetermined services for which usage records are to be acquired are not limited to two, and may be three or more services.

  The event collection unit 11 acquires from the service management server 3 usage records for a plurality of predetermined services, for example, service A and service B, for which usage records are to be acquired, at a predetermined timing. For example, when the service A or service B is used by a user, the event collection unit 11 accesses the service management server 3 and requests a usage record, thereby acquiring usage records for the service A and the service B. . The event collection unit 11 accesses the service management server 3 at a predetermined time interval such as every week or every month, and requests usage records, thereby using the services A and B. A record may be acquired. Each usage record for each service includes a plurality of sample data.

  The event collection unit 11 stores the usage records for the acquired service A and service B in the usage record storage unit 12. The usage record storage unit 12 is, for example, a usage record database. An example of the usage record is shown in FIG.

  FIG. 4 is an explanatory diagram of the correlation process, and shows an example of a usage record for each user. FIG. 4A shows a usage record in which the terminal 2 used by the user # 1, in other words, the terminal # 1 uses the service A. FIG. 4B shows a usage record in which the terminal 2 used by the user # 2, in other words, the terminal # 2 uses the service B.

  The usage record includes, for example, values of a plurality of parameters for the service A. As parameters, as shown in FIGS. 4A and 4B, for example, use time T, number of registered users U, number of events E, and service-specific parameter S are used. The parameter value is recorded every month, for example. Therefore, the usage record actually includes, for example, the value t1 as the usage time T and the value u1 as the number of registered users U for X month, as shown in FIG.

  The usage time T is the total usage time of the service, specifically, the total time from the usage start time of the service to the usage stop time. The registered user number U is, for example, the total number of users registered to use the service in the organization user. The number of events E is, for example, the total number of times of logging in and logging out of the service. The service specific parameter S is a parameter specific to the service, and is, for example, the number of files, disk capacity, and the like. The number of files and the disk capacity are the number of files and disk capacity used in using the service, respectively.

  The event collection unit 11 updates the total of each month for the value of each parameter each time the usage records for the service A and the service B are acquired, and calculates and updates the average value of the total of each month. As a result, the usage record is updated to the latest value.

  The event collection unit 11 creates, for example, usage records shown in FIGS. 4A and 4B for each terminal 2. In the example of FIGS. 4A and 4B, one user to be used is determined in advance for each terminal 2, and one service to be used is determined in advance. A single terminal 2 may be used by a plurality of users, or a plurality of services may be used from a single terminal 2. In these cases, the usage records shown in FIGS. 4A and 4B may be created, for example, for the same terminal 2 with the same user ID and the same service.

  Here, the management data acquired by the service management server 3 has the same configuration as the usage records shown in FIGS. 4 (A) and 4 (B). As described above, the service management server 3 can know the start and end of the use of the service from the terminal 2. Therefore, the service management server 3 can acquire management data having the same configuration as the usage records shown in FIGS. 4 (A) and 4 (B).

  For example, the service management server 3 may acquire only the registered user number U among the plurality of parameters in the usage record, and the user management server 1 may acquire other parameters. In this case, the other parameters are acquired by the user management server 1 communicating with the terminal 2 using the service.

  The data correlation analysis unit 13 is an example of an analysis unit, and for the service A and the service B based on the acquired usage record, the correlation value between each of the plurality of parameters of each service, in other words, the correlation Find a number. The parameter is, for example, a service specific parameter that is a parameter specific to each of usage time, number of registered users, number of events, service A and service B. The data correlation analysis unit 13 stores the obtained correlation value in the usage record storage unit 12. An example of the calculation of the correlation value is shown in FIG.

  FIG. 5 is an explanatory diagram of the correlation process, and is an example of the correlation distribution of each service.

  FIG. 5A is a graph of the correlation distribution for the service A, for example, and shows the correlation between the usage time Ta and the registered user number Ua. In FIG. 5A, the horizontal axis indicates the usage time Ta, and the vertical axis indicates the number of registered users Ua.

  FIG. 5B is a graph of the correlation distribution for the service B, for example, and shows the correlation between the usage time Tb and the number of events Eb. In FIG. 5B, the horizontal axis indicates the usage time Tb, and the vertical axis indicates the number of events Eb.

  For example, in FIG. 5A, sample data included in the usage record is plotted on a graph. In FIG. 5A, the plotted sample data is represented by dots. As sample data included in the usage record, for example, the usage time Ta and the number of registered users Ua in each month are used. The number of dots to be plotted, in other words, the number of sample data is sufficient to obtain the correlation value, for example, 15 to 20. The number of dots plotted is not limited to this.

  For example, as the usage record of the service A of the terminal # 1 shown in FIG. 4A, for the X month, the value t1 is recorded as the usage time T and the value u1 is recorded as the number of registered users U. In this case, in FIG. 5A, dots corresponding to the month X are plotted at the intersection of the vertical line from the usage time t1 on the horizontal axis and the vertical line from the registered user number u1 on the vertical axis.

  Actually, the data correlation analysis unit 13 does not plot the sample data on the graph as shown in FIGS. 5 (A) and 5 (B). The data correlation analysis unit 13 calculates a correlation value by calculation using sample data.

  The data correlation analysis unit 13 obtains a correlation value ra1 between the usage time Ta and the registered user number Ua based on the graph shown in FIG. The data correlation analysis unit 13 also sequentially creates a combination of two parameters for other parameters of the service A, and obtains a correlation value between the combined two parameters.

  Further, the data correlation analysis unit 13 obtains a correlation value rb2 between the usage time Tb and the number of events Eb based on the graph shown in FIG. The data correlation analysis unit 13 also sequentially creates a combination of two parameters for other parameters of the service B, and obtains a correlation value between the combined two parameters.

  The data correlation analysis unit 13 extracts a combination of correlation values corresponding to a predetermined condition based on the obtained correlation value. The combination of correlation values is a combination of the correlation value ra of service A and the correlation value rb of service B in two different services, and is a combination of two correlation values ra and rb having one parameter in common. It is a combination (ra, rb). The data correlation analysis unit 13 stores the extracted combination of correlation values in the usage record storage unit 12.

  The correlation value ra belonging to the combination (ra, rb) is any one of one correlation value ra having a value equal to or greater than a predetermined value among the correlation values of the service A. The correlation value rb belonging to the combination (ra, rb) is any one of the correlation values rb having a value equal to or greater than a predetermined value among the correlation values of the service B.

  The correlation value is a value between “−1” and “1”. In general, when the absolute value of the correlation value is “0.4” or more, it can be said that “there is a considerable correlation”. Therefore, the data correlation analysis unit 13 determines that there is a high correlation with an absolute value of the correlation value of “0.4” or more, and extracts it as one of the correlation values belonging to the combination (ra, rb). . As a result, parameters having a certain degree of correlation can be extracted without omission, and parameters having little correlation can be eliminated. As a result, an appropriate number of combinations (ra, rb) are extracted. be able to. An example of the extracted combination (ra, rb) is shown in FIG.

  FIG. 6 is an explanatory diagram of correlation processing, and shows an example of correlation data of a service. FIG. 6A shows a correlation value having an absolute value of “0.4” or more for the service A. FIG. 6B shows a correlation value with an absolute value of “0.4” or more for the service B.

  For example, for service A, as shown in FIG. 6A, the correlation value between the usage time Ta and the registered user number Ua is ra1, and the correlation value between the usage time Ta and the event number Ea is ra2. The correlation value ra1 and the correlation value ra2 are “0.4” or more. For service B, as shown in FIG. 6B, the correlation value between the usage time Tb and the registered user number Ub is rb1, and the correlation value between the usage time Tb and the event number Eb is rb2. The correlation value rb1 and the correlation value rb2 are “0.4” or more.

  In this case, the correlation value ra1 in the service A has a usage time T as a common parameter between the correlation value rb1 and the correlation value rb2 in the service B. Further, the correlation value ra2 in the service A has a usage time T as a common parameter between the correlation value rb1 and the correlation value rb2 in the service B. Therefore, the data correlation analysis unit 13 extracts four combinations as combinations of correlation values. Specifically, the combination (ra1, rb1), the combination (ra1, rb2), the combination (ra2, rb1), and the combination (ra2, rb2) are extracted.

  When the absolute value of the correlation value is “0.7” or more, it can be generally said that “there is a strong correlation”. Therefore, a correlation value having an absolute value of “0.7” or more may be extracted as one of correlation values belonging to the combination (ra, rb) by determining that there is an extremely high correlation. In practice, since the correlation between service parameters is almost a positive correlation, a negative correlation value may be ignored.

  In the extracted combination (ra, rb), the data correlation analysis unit 13 obtains the sum of correlation values belonging to the combination (ra, rb), in other words, the combination that maximizes ra + rb. For example, as described above, when four combinations (ra1, rb1), combination (ra1, rb2), combination (ra2, rb1), and combination (ra2, rb2) are extracted, the sum of correlation values ra1 + rb1 , Ra1 + rb2, ra2 + rb1, and ra2 + rb2 are calculated. Based on the comparison of the calculated four sums, for example, a combination (ra1, rb2) is obtained as a combination of correlation values that maximizes the sum of correlation values.

  The data correlation analysis unit 13 obtains a sample related to the correlation value ra of the service A for each of the correlation value ra of the service A and the correlation value rb of the service B belonging to the combination (ra, rb) that maximizes the calculated sum ra + rb. Data x (a) and sample data x (b) related to the correlation value rb of service B are extracted as common sample data x. The data correlation analysis unit 13 stores the common sample data x, in other words, the extracted sample data x (a) and the sample data x (b) in the usage record storage unit 12, for example.

  For example, when the correlation value combination that maximizes the sum of correlation values is (ra1, rb2), the sample data x (a) related to the correlation value ra of service A is used in FIG. Data, in other words, a plurality of sample data of the usage time Ta of the service A and a plurality of sample data of the registered user number Ua of the service A are stored. Further, as the sample data x (b) related to the correlation value rb of the service B, the data used in FIG. 5B, in other words, a plurality of sample data of the service B usage time Tb, and the service B A plurality of sample data with the number of events Eb is stored. Therefore, in practice, the common sample data x, in other words, the extracted sample data x (a) and sample data x (b) are each a set of two sample data related to the correlation value ra of the service A. , And a set of two sample data related to the correlation value rb of service B.

  The approximate calculation unit 14 is an example of a calculation unit, and generates recommendation information 6 of the payment amount, which is a consideration for the service A and the service B, based on the combination (ra, rb) that maximizes the calculated sum ra + rb. To do.

  Specifically, the approximate calculation unit 14 calculates the payment amount M for the service A and the service B based on the extracted common sample data x. The payment amount M is calculated by, for example, M = T × Pt + U × Pu + E × Pe + S × Ps +. Here, T is the usage time, Pt is the usage time unit price, U is the number of registered users, Pu is the number of registered user units, E is the number of events, Pe is the unit number of events, S is a service specific parameter, and Ps is a service specific parameter unit price. It is. An example of the calculation of the payment amount M is shown in FIG.

  FIG. 7 is an explanatory diagram of the correlation process, and shows an example of calculation of the payment amount M, in other words, an example of a graph for determining a recommendation of the payment amount M.

  For example, as described above, when the combination of the correlation values that maximizes the sum of the correlation values is (ra1, rb2), the sample data x (a) related to the correlation value ra of the service A is the usage of the service A. It includes a plurality of sample data of time Ta and a plurality of sample data of the number of registered users Ua of service A. In other words, the sample data x (a) is a set of sample data. Accordingly, E = 0 and S = 0 in the calculation formula for the payment amount M described above. As a result, as shown in FIG. 7, a straight line or a curve of the payment amount M for the service A is obtained.

  The sample data x (b) related to the correlation value rb of the service B includes a plurality of sample data of the service B usage time Tb and a plurality of sample data of the service B event number Eb. In other words, the sample data x (b) is a set of sample data. Accordingly, U = 0 and S = 0 in the calculation formula for the payment amount M described above. As a result, as shown in FIG. 7, a straight line or a curve of the payment amount M for the service B is obtained.

  Therefore, FIG. 7 is a graph for determining the recommendation information 6 of the payment amount M when there is sample data having a high correlation value in common for the billing target of the service. Based on the graph of FIG. 7, the rough calculation unit 14 determines to recommend the service B when the user usage time is from t1 to t2, and when the user usage time is from t2 to t3. Decides to recommend service A.

  In the graph for determining the recommendation of the payment amount M, the horizontal axis is not limited to the use time, but may be various other parameters.

  The approximate calculation unit 14 generates recommendation information 6 on the payment amount M based on the calculated payment amount M and the user status included in the acquired usage record. The user status is, for example, actual values of parameters such as usage time, number of registered users, number of events, and service-specific parameters. For example, in the example of FIG. 7, the user status is usage time. For example, when the usage time of the user who uses the service B is from t2 to t3, the recommendation information 6 recommending the service A is generated based on the decision to recommend the service A. An example of the recommendation information 6 in this case is shown in FIG.

  FIG. 8 is a diagram illustrating an example of recommendation information.

  The recommendation information 6 includes, for example, an address such as “Mr. xx” and a text such as “Subscribed service B...”. The recommendation information 6 is not limited to the format shown in FIG. 8, and may take other various formats. The text is not limited to the text shown in FIG. 8, and may include other various texts.

  The address is the identification information of the user or the identification information of the terminal 2, and can be acquired from the user information held by the user management server 1. The address may be omitted. The main text includes a recommendation about the price, "If you use service A, you can save xxxx yen per month." Accordingly, the user “xx” can change the service to be used.

  In addition, the main text includes information that “a similar organization tends to use service A rather than service B as a result of analysis” in connection with a recommendation about price. Thereby, the user “xx” can know the tendency of the users who are using the same service. In addition, you may make it notify the tendency of the user who is using the own service to the service provision apparatus 4. FIG.

  Thereafter, as shown in FIG. 2, the approximate calculation unit 14 transmits the generated recommendation information 6 to the terminal 2. For example, as shown in FIG. 2, the recommendation information 6 is transmitted to the terminal 2 that uses the service B when the recommendation information 6 recommends the use of the service A instead of the service B. In this case, the rough calculation unit 14 extracts the terminal 2 having the usage record using the service B by referring to the usage record stored in the usage record storage unit 12, and recommends information 6 to the extracted terminal 2. Send.

  Note that the recommendation information 6 may be transmitted to the terminal 2 that has transmitted the usage application that triggered the acquisition of the usage record. Further, the recommendation information 6 may be transmitted to all of the plurality of terminals 2. The recommendation information 6 may be transmitted to the terminal 2 used by the administrator of the user management server 1.

  As a result, the user can obtain appropriate recommendation information 6 based on the optimum correlation for a plurality of services having different charging parameters. The user can easily compare a plurality of similar services based on the recommendation information 6 and can select the cheapest service.

  Note that, for example, in addition to the service A and the service B, a service C exists as a plurality of services for which usage records are acquired. In this case, the combination (ra, rb) of the correlation value ra of the service A and the correlation value rb of the service B, the combination (rb, rc) of the correlation value rb of the service B and the correlation value rc of the service C, and the service A combination (rc, ra) of the correlation value rc of C and the correlation value ra of service A is extracted. Then, recommendation information 6 based on the combination (ra, rb), recommendation information 6 based on the combination (rb, rc), and recommendation information 6 based on the combination (rc, ra) are generated. Thereby, the judgment material about selection of the service A, the service B, and the service C can be provided to a user. The number of services for which usage records are acquired may be four or more.

  FIG. 9 is a diagram illustrating an example of a hardware configuration of the information processing apparatus.

  In the user management server 1, a CPU 101, a memory 102, a display control unit 103, a hard disk device (HDD) 105, an input device 106, a drive device 107, and a communication control unit 109 are connected to each other via a bus 110. A display device 104 is connected to the display control unit 103.

  The CPU 101 controls the user management server 1 that is a computer in accordance with an operating system (OS) that resides on the memory 102 that is a main memory. The CPU 101 executes a correlation processing program on the memory 102. Thereby, the event collection unit 11, the data correlation analysis unit 13, and the approximate calculation unit 14 are realized. For example, the correlation processing program is stored in a recording medium 108 such as a CD-ROM or DVD, and is input from the recording medium 108 to a hard disk device (HDD) 105 via a drive device 107 such as a CD-ROM drive or DVD drive. The data is loaded from the hard disk device 105 to the memory 102.

  The input device 106 is a keyboard, for example, and may include a mouse or the like. The display control unit 103 controls the display device 104 according to a request from the CPU 101 and displays a screen on the display device 104. The display device 104 is an output device, for example, a display. The output device may include an output device such as a printer.

  The communication control unit 109 is, for example, a transmission / reception device, connected to the network 5, and connected to another computer, for example, the service management server 3 via the network 5. As a result, the user management server 1 communicates with the service management server 3.

  The usage record storage unit 12 is provided in the hard disk device 105, for example. In other words, the usage record is stored in the hard disk device 105, for example. The usage record is collected from the service management server 3 via the network 5 by the event collection unit 11 and input to the hard disk device 105, and loaded from the hard disk device 105 to the memory 102, for example, the data correlation analysis unit 13 and the approximate calculation unit. 14 is processed.

  The service management server 3 also has the hardware configuration shown in FIG.

  FIG. 10 is a recommendation information generation process flow. FIG. 10 shows an outline of processing for generating recommendation information in the user management server.

  The event collection unit 11 acquires the usage record of the user as shown in FIG. 4 from the service management server 3 by using the service using the terminal 2 by the user as a trigger, and stores the acquired usage record in the usage record storage unit 12. Save (step S11).

  The data correlation analysis unit 13 obtains correlation values ra and rb as shown in FIG. 5 based on each sample data included in the usage record stored in the usage record storage unit 12 (step S12), and the obtained correlation value. Are used to compare the correlations, thereby extracting a combination (ra, rb) of correlation values having a high correlation as shown in FIG. 6 (step S13).

  The approximate calculation unit 14 uses the combination of correlation values (ra, rb) extracted by the data correlation analysis unit 13 and based on the user status, for example, the usage time, the price as shown in FIG. 7 and FIG. Recommendation, in other words, the recommendation information 6 is determined (step S14), and the determined recommendation information 6 is transmitted to the user terminal 2 (step S15).

  FIG. 11 is a recommendation information generation process flow. FIG. 11 shows details of the recommendation information generation process in the information processing system. In other words, FIG. 11 is a concrete example of the processing flow of FIG.

  When the user who viewed the service catalog received from the service management server 3 selects a service from the service catalog, the terminal 2 transmits a use application for the selected service to the service providing apparatus 4 via the service management server 3. The service providing apparatus 4 that has received the service use application permits the use of the service for the received service use application. As a result, the terminal 2 starts using the service provided by the service providing apparatus 4.

  In the service management server 3, the management processing unit 31 manages the service catalog, service use application, and service use permission, thereby obtaining management data regarding service provision and storing it in the user information storage unit 32. .

  As shown in FIG. 4, the event collection unit 11 acquires usage records of all users in the user management server 1 as a trigger when the user uses the service using the terminal 2, and is used by the entire user management server 1. The calculation formula of the payment amount M for each of all the services to be performed is acquired, and the acquired usage record and calculation formula are stored in the usage record storage unit 12 (step S21). The usage record is acquired from the service management server 3, and the calculation formula is acquired from the user management server 1. The calculation formula is stored in a predetermined storage area of the user management server 1. The calculation formula includes the unit price of each parameter.

  For example, when the service provision ends, the service management server 3 transmits a charge amount notification for the provided service to the terminal 2 used by the user.

  The terminal 2 receives the charge amount notification from the service management server 3 and transmits the received charge amount notification to the user management server 1, for example.

  The event collection unit 11 determines whether or not the notification of the charge amount from the terminal 2 has been received, in other words, whether or not the use of the service provided in response to the triggering application for use has ended ( Step S22). When the notification of the charge amount is not received (step S22, No), the event collection unit 11 repeats step S21. If the notification of the charge amount is not received, the event collection unit 11 may repeat step S22.

  When the notification of the charge amount is received (step S22, Yes), the data correlation analysis unit 13 reads the usage records of all the users stored in the usage record storage unit 12, and records the usage records of all the read users. The correlation distribution based on each of the sample data is created, and the correlation value r is obtained as shown in FIG. 5 (step S23).

  Thereafter, as shown in FIG. 6, the data correlation analysis unit 13 determines whether or not sample data having a correlation exists in common for each service (step S <b> 24). Step S21 will be described later with reference to FIG. If the sample data does not exist in common (No at Step S24), the data correlation analysis unit 13 repeats Step S21.

  When sample data exists in common (step S24, Yes), the approximate calculation unit 14 obtains a calculation formula for the payment amount M for each of all services from the usage record storage unit 12 (step S25). Then, the approximate calculation unit 14 uses the common sample data x, the calculation formula of the payment amount M for each of the services related to the common sample data x, and the user status such as the number of users and the usage time. As shown in FIGS. 7 and 8, a service to be recommended to the user is determined, and recommendation information 6 is generated (step S26). Thereafter, the approximate calculation unit 14 transmits the generated service recommendation, in other words, the recommendation information 6 to, for example, the terminal 2 related to the recommendation information 6, in other words, the user (step S27).

  FIG. 12 is a recommendation information generation processing flow executed by the data correlation analysis unit, and shows details of step S24.

  The data correlation analysis unit 13 selects the correlation value ra of one service A as the target of optimal correlation processing from the correlation value data of two services to be processed, for example, service A and service B, respectively. The correlation value rb of one service B is read (step S31). Here, the correlation value ra and the correlation value rb have one sample data in common among the two sample data that is the basis of each correlation value. In other words, the correlation value ra and the correlation value rb have one parameter in common among the two parameters that are the basis of each correlation value. This is referred to as “condition 1” in the optimum correlation process.

  The data correlation analysis unit 13 determines whether “ra ≧ 0.4 and rb ≧ 0.4” holds for the read correlation value ra and correlation value rb (step S32). When “ra ≧ 0.4 and rb ≧ 0.4” is satisfied (step S32, Yes), two correlation values ra and rb and the basis of the correlation value having one sample data in common The two obtained sample data are stored in the array tmp (step S33). The two sample data that is the basis of the correlation value is the common sample data x. The array tmp is a storage area having a structure corresponding to the table of FIG. When “ra ≧ 0.4 and rb ≧ 0.4” does not hold (No in step S32), the data correlation analysis unit 13 omits step S33.

  Thereafter, the data correlation analysis unit 13 determines whether or not all combinations of the correlation value ra and the correlation value rb satisfying the condition 1 have been compared with the correlation value “0.4” that is the threshold value in step S32. (Step S34). When the comparison has not been completed for all the combinations (step S34, No), the data correlation analysis unit 13 sets the two correlation values ra and rb within the range satisfying the condition 1, for example, the storage position in the array tmp. (Step S35), and then repeat step S32.

  If all the combinations have been compared (step S34, Yes), the data correlation analysis unit 13 uses the data stored in the array tmp in step S33, in other words, the two correlation values ra and rb. It is determined whether or not there is one or more sets of common sample data x (step S36).

  When one or more sets exist (step S36, Yes), the data correlation analysis unit 13 sets the combination (ra, rb) of the sum ra + rb of the correlation value ra and the correlation value rb among the data stored in the array tmp. rb) is obtained, the obtained combination (ra, rb) and the common sample data x are stored in the usage record storage unit 12 (step S37), and then step S25 of FIG. 11 is executed.

  When one or more sets do not exist (step S36, No), step S21 in FIG. 11 is executed.

DESCRIPTION OF SYMBOLS 1 User management server 2 Terminal 3 Service management server 4 Service provision apparatus 5 Network 6 Recommendation information 11 Event collection part 12 Usage record storage part 13 Data correlation analysis part 14 Approximation calculation part 31 Management processing part 32 User information storage part

Claims (8)

A collection unit for obtaining usage records of the first service and the second service;
Based on the acquired usage record, for the first service and the second service, a correlation value between each of a plurality of parameters of each service is obtained, and based on the obtained correlation value, the first service A correlation value of one service and a correlation value of the second service, a combination of correlation values having one parameter in common is extracted, and the correlation value belonging to the combination in the extracted combination An analysis unit for finding a combination that maximizes the sum of
And a calculation unit that generates recommendation information of payment amount as a consideration of the service for the first service and the second service based on the combination that maximizes the calculated sum. apparatus. The information according to claim 1, wherein the parameter is a service specific parameter that is a parameter specific to each of a usage time, the number of registered users, the number of events, the first service, and the second service. Processing equipment. The correlation values belonging to the combination are one correlation value having a value equal to or greater than a predetermined value among the correlation values of the first service and the predetermined value among the correlation values of the second service. The information processing apparatus according to claim 1, wherein the correlation value is a single correlation value having a value greater than or equal to a given value. The usage records for the first service and the second service each include a plurality of sample data;
Sample data related to the correlation value of the first service for each of the correlation value of the first service and the correlation value of the second service belonging to the combination in which the sum obtained by the analysis unit is the maximum. And the sample data related to the correlation value of the second service are extracted as common sample data. The calculation unit calculates the payment amount for the first service and the second service based on the extracted common sample data, and includes the calculated payment amount and the acquired usage record The information processing apparatus according to claim 4, wherein the recommendation information for the payment amount is generated based on a status of a user who receives the payment. The collection unit obtains usage records for the first service and the second service, which are predetermined services for which the usage records are acquired, at a predetermined timing. The information processing apparatus according to claim 1. The collection unit obtains usage records for the first service and the second service,
The analysis unit obtains a correlation value between each of the plurality of parameters of each service for the first service and the second service based on the acquired usage record, and based on the obtained correlation value A combination of the correlation value of the first service and the correlation value of the second service, the combination of the correlation values having one parameter in common, and extracting the combination in the extracted combination Find the combination that maximizes the sum of correlation values belonging to
The calculation unit generates recommendation information of a payment amount that is a consideration of the service for the first service and the second service, based on the combination that maximizes the calculated sum. Method. A program for generating recommendation information,
On the computer,
Obtain usage records for the first service and the second service,
Based on the obtained usage record, for the first service and the second service, a correlation value between each of a plurality of parameters of each service is obtained,
Based on the obtained correlation value, a combination of the correlation value of the first service and the correlation value of the second service, and extracting a combination of correlation values having one parameter in common,
In the extracted combination, find a combination that maximizes the sum of correlation values belonging to the combination;
A program for generating recommendation information of a payment amount that is a consideration of the service for the first service and the second service on the basis of a combination that maximizes the calculated sum. .

Images