JP2010282561A - Information processing apparatus, information processing system, information processing method, and information processing program - Google Patents

Abstract

It is an object of the present invention to easily extract information for grasping a user's behavioral orientation from behavior history data about a user's information collection behavior or purchasing behavior.
An analysis server acquires action history data of an analysis target user for a predetermined period, and divides the acquired user action history data for each predetermined unit period. Each divided action history data is classified into a plurality of action history data sets based on a predetermined classification criterion, and the number of histories belonging to each classified action history data set is totaled. Furthermore, each behavior history when each behavior history data set is developed in a graph using as parameters the number of histories belonging to each behavior history data set and the ranking information ranking the behavior history data sets according to the number of histories. By approximating the distribution tendency of the data set using a predetermined function, an index for specifying the user's action orientation is calculated for each predetermined unit period.
[Selection] Figure 1

Description

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

  In recent years, with the advance of communication technology using mobile devices such as mobile terminals and the spread of communication infrastructure such as networks, users can easily collect information and purchase goods.

  And, it is important to classify user behavior into goal-oriented and diversity-oriented, and to capture user-oriented behavior, such as marketing analysis, advertisement distribution design, information navigation system, etc. Development of technology aimed at analyzing user behavior such as purchasing and purchasing behavior is in progress.

  For example, Non-Patent Document 1 proposes a technique for acquiring behavior history data such as information collection behavior and purchasing behavior of a user through a questionnaire and classifying the behavior of the user into purpose-oriented and desire-oriented. Non-Patent Document 2 also proposes a technique for extracting purpose directivity based on generality of keywords and genres used when a user searches for information.

  In addition, as a related technique, a technique for analyzing a distribution trend of macro data such as market data or purchase data in a specific site and using it for a market change prediction or a demand prediction has been proposed. For example, Patent Document 1 proposes a technique for evaluating a risk of a financial portfolio by providing a threshold value for an index of a correlation matrix between financial products. Patent Document 2 discloses a technique for modeling a user by using a power (冪) index of a shopping place and the number of times of shopping as an index of ease of movement of a consumer when setting a trade area of a store. Has also been proposed.

JP 2002-183429 A JP 2003-281348 A

Kuniichi Amagasa, Fumitoshi Kato, Daisuke Okabe, “A Study on Situation Information Collection and its Possibility of Use”, Information Processing Society of Japan, May 15, 2008, 2008-UBI-18 Matsuyuki Tomoyuki et al., “Implementation of Information Filtering System Considering Purpose”, IEICE, December 2008, WI2-2008-60

  However, since the technique proposed in Non-Patent Document 1 acquires data related to a user through a questionnaire, it is not only troublesome but also lacks real-time properties in analyzing user behavior. In addition, since the technique proposed in Non-Patent Document 2 is intended for diversity of keywords and genres, a machine for extracting important sentences and keywords from the preparation of prerequisite knowledge such as a genre estimation dictionary and collected information Costs are required to prepare the learning mechanism.

  In addition, it has been found that WEB access history data relating to user information collection behavior, purchase behavior, etc. shows a power distribution trend. Therefore, for example, in order to analyze the distribution tendency of WEB access history data, it may be possible to set the average value or variance value of the distribution data as a threshold value. However, unlike the normal distribution, the average value is used for analyzing the power distribution. Values and variance values do not have much meaning as threshold values. Therefore, the threshold setting is not always easy depending on the distribution tendency of the analysis target data.

  The present invention has been made in view of the above, and it is possible to easily extract information for capturing the user's behavior orientation from the behavior history data about the user's information collecting behavior and purchasing behavior. An object is to provide a processing device, an information processing system, an information processing method, and an information processing program.

  In order to solve the above-described problems and achieve the object, the present invention is an information processing apparatus that performs information processing using user behavior history data, and collects a plurality of behavior history data from a plurality of user terminals. Action history data collecting unit, a plurality of behavior history data collected by the behavior history data collecting unit for each user, and a plurality of behavior history data stored in the behavior history storage unit. From among them, the behavior history data of the analysis target user is acquired for a predetermined period, the acquired behavior history data of the user is divided for each predetermined unit period, and a plurality of divided behavior history data are based on a predetermined classification criterion. Each of the behavior history data sets, and a data totaling unit that counts the number of histories belonging to each classified behavior history data set, and the data totaling unit In addition, the total information storage unit that stores the history number and the rank information that ranks the action history data set according to the history number in association with each other, and the graph that uses the history number and the rank information as parameters. An index for identifying the user's behavior orientation is calculated for each predetermined unit period by approximating the distribution tendency of each behavior history data set when each history data set is developed using a predetermined function. And an exponent calculation result output unit that outputs an exponent calculated by the exponent calculation unit for each predetermined unit period as an exponent calculation result to an analyzer terminal that analyzes a user's behavioral orientation using the exponent And have.

  An information processing apparatus, an information processing system, an information processing method, and an information processing program according to the present invention easily extract information for capturing a user's behavior orientation from behavior history data about a user's information collection behavior or purchase behavior. it can.

FIG. 1 is a diagram illustrating a system configuration according to the first embodiment. FIG. 2 is a diagram illustrating the configuration of the analysis server according to the first embodiment. FIG. 3 is a diagram illustrating a configuration example of an action log according to the first embodiment. FIG. 4 is a diagram illustrating a configuration example of a purchase log according to the first embodiment. FIG. 5 is a diagram illustrating a configuration example of a ranking tabulation result according to the first embodiment. FIG. 6 is a diagram for explaining an index calculation procedure according to the first embodiment. FIG. 7 is a diagram illustrating a configuration example of an exponent calculation result according to the first embodiment. FIG. 8 is a diagram for explaining the processing flow of the analysis server according to the first embodiment. FIG. 9 is a diagram for explaining the effect according to the first embodiment. FIG. 10 is a diagram for explaining the effect according to the first embodiment. FIG. 11 is a diagram illustrating a computer that executes an information processing program. FIG. 12 is a diagram for explaining the relationship between the user's tendency to access the WEB and the user's purpose of action.

  Hereinafter, embodiments of an information processing apparatus, an information processing system, an information processing method, and an information processing program according to the present invention will be described in detail with reference to the drawings.

[Relationship between user's access tendency to WEB and user's purpose of action]
Below, the relationship between a user's access tendency to WEB and a user's action purpose is demonstrated. For all users to be observed, data such as WEB access destination (for example, URL or domain of the site), access date and time, and browsing time are recorded. FIG. 12 is a graph in which the horizontal axis is ranked (access number rank) and the vertical axis is the access number for all observation target users in descending order of the number of accesses. In addition, FIG. 12 is a figure for demonstrating the relationship between a user's access tendency to WEB and a user's action purpose.

  As shown in FIG. 12, the distribution of the number of access to the user's WEB tends to be a power distribution (Yusuke Ichikawa et al., “Web log trend analysis focusing on user's action purpose”, Shinsei Institute 2009, WI2-2009-9). The power distribution means a distribution according to a power law in which a certain observation amount is proportional to the power of a parameter.

Then, as shown in FIG. 12A, when attention is paid to an exponent “k” that should be an exponent in the approximate expression f (x) = ax ^k of the power distribution of the number of accesses, the exponent is a small negative value. In this case (when the slope of the graph is small), it has been found that the user's behavior is likely to be performing a behavior specialized in achieving a specific purpose.

  Also, as shown in FIG. 12B, when the power index “k” shows a large negative value (when the slope of the graph is small), the user's actions are highly diverse and the purpose itself It has been found that there is a high possibility that the behavior (diversity-oriented behavior) whose purpose is to look for is.

  Therefore, in the following, an embodiment will be described in which, for example, information for capturing the user's behavioral orientation is easily extracted using the fact that the distribution tendency of the user's WEB access history becomes a power distribution tendency.

  In a first embodiment described below, an embodiment of a system having an analysis server functioning as an information processing apparatus will be described as an example of the information processing apparatus according to the present invention. The present invention is not limited to the following Example 1.

  In the following first embodiment, the outline and features of the analysis server according to the first embodiment, the configuration and processing of the analysis server will be described in order, and finally the effects of the first embodiment will be described.

[Outline and Features (Example 1)]
The analysis server according to the first embodiment is summarized to execute information processing using user behavior history data. The analysis server according to the first embodiment is characterized in that attention is paid to the fact that the distribution tendency of the WEB access history regarding the information collection behavior and purchase behavior of the user is a power distribution.

  Specifically, the analysis server acquires the action history data of the analysis target user for a predetermined period, and divides the acquired user action history data for each predetermined unit period. Each divided action history data is classified into a plurality of action history data sets based on a predetermined classification criterion, and the number of histories belonging to each classified action history data set is totaled.

  Further, the analysis server expands each behavior history data set in a graph using the number of histories belonging to each behavior history data set and the rank information ranking the behavior history data sets according to the number of histories as parameters. By approximating the distribution tendency of each action history data set using a predetermined function, an index for specifying the user's action orientation is calculated for each predetermined unit period. The analysis server according to the first embodiment will be specifically described below.

[Configuration of Example 1]
FIG. 1 is a diagram illustrating a system configuration according to the first embodiment. FIG. 2 is a diagram illustrating the configuration of the analysis server according to the first embodiment. FIG. 2 shows only processing function units necessary for explaining the configuration of the analysis server according to the first embodiment.

  As illustrated in FIG. 1, the analysis server 300 according to the first embodiment performs electronic communication between an analysis target user terminal group including a plurality of user terminals 100 that perform information collection behavior and purchase behavior by WEB access, and the user terminal 100. It is connected to an EC server group composed of a plurality of EC (electronic commerce) servers 200 for performing commercial transactions via the Internet 1. The analysis server 300 is connected to a state in which various information can be exchanged with the analyst terminal 400 that analyzes the behavior of the user of the analysis target user terminal group.

  2, the analysis server 300 includes an action log acquisition unit 310, an action log storage unit 320, a purchase log acquisition unit 330, a purchase log storage unit 340, a ranking tabulation unit 350, a ranking tabulation result storage unit 360, A power index calculation unit 370, a power index calculation result output unit 380, and a target period log extraction unit 390 are included.

  The action log acquisition unit 310 acquires a WEB access history including a user ID and an access destination URL from each user terminal 100 as an action log of a user to be analyzed. Then, the behavior log acquisition unit 310 records the acquired WEB access history in the behavior log storage unit 320.

  The behavior log storage unit 320 stores the WEB access history recorded by the behavior log acquisition unit 310 for each user. For example, as illustrated in FIG. 3, the action log storage unit 320 includes a recording date and time (Time stamp) when the WEB access history is recorded by the action log acquisition unit 310, a user ID (User ID) of the user who performed the access, and The URL of the access destination is stored in association with it. FIG. 3 is a diagram illustrating a configuration example of an action log according to the first embodiment.

  The purchase log acquisition unit 330 receives, as a purchase log of a user to be analyzed, a purchase ID including the user ID, the ID of the product purchased by the user, and the price of the product from each EC server 200 that performs electronic commerce with the user terminal 100. Get history. Then, the purchase log acquisition unit 330 records the acquired purchase history in the purchase log storage unit 340.

  The purchase log storage unit 340 stores the purchase history recorded by the purchase log acquisition unit 330 for each user. For example, as illustrated in FIG. 4, the purchase log storage unit 340 includes a recording date and time (Time stamp) when a purchase history is recorded by the purchase log acquisition unit 330, a user ID (User ID) of a user who purchased the product, a user Stores the product ID (Item ID) and the price of the product purchased in association with each other. FIG. 4 is a diagram illustrating a configuration example of a purchase log according to the first embodiment.

  The ranking totalization unit 350 acquires the WEB access history from the action log storage unit 320 for each user, for example, totals the number of accesses of each site, and ranks the sites in order of the number of accesses. Do it every time.

  Specifically, the ranking tabulation unit 350 receives a power index calculation request of the analysis target user from the analyst terminal 400. The power index calculation request includes a user ID to be analyzed, a total unit period, and a total period. For example, the total unit period represents a unit for totalizing the web access history of the analysis target user, such as a daily unit, a week unit, or a month unit. The aggregation period represents a period during which ranking aggregation is performed for each aggregation unit period, such as aggregation for one day for one week and aggregation for one month for one year.

  The ranking totaling unit 350 acquires the web access history of the analysis target user from the behavior log storage unit 320 for the totaling period, using the user ID and the totaling period included in the index calculation request that should be received from the analyst terminal 400 as keys. The WEB access history is divided for each aggregation unit period included in the power index calculation request. Each WEB access history divided for each aggregation unit period is classified into a plurality of WEB access history sets based on whether or not the same site has been accessed, and belongs to each classified WEB access history set. Total the number of histories.

  For example, the WEB access history of the analysis target user is acquired for one year (12 months) received as a totaling period, and the acquired WEB access history is divided every one month received as a totaling unit period. Each WEB access history divided every month is classified into a plurality of WEB access history sets based on whether or not the WEB access history is a history of accessing the same site.

  If there is a power index calculation request from the analyst terminal 400 on January 1, the ranking totaling unit 350 calculates the total period for one year from the analysis target user's WEB access history, that is, January. WEB access history from 1st to 31st December is acquired. Then, the acquired WEB access history for one year is January 1 to January 31, February 1 to February 28 (or 29), March 1 to March 31,... In this way, it is divided every month, which is the total unit period.

  Next, the ranking tabulation unit 350 performs the counting unit period from January 1 to January 31, February 1 to February 28 (or 29), March 1 to March 31, and so on. Each divided WEB access history is classified into a plurality of WEB access history sets based on whether or not the WEB access history is a history of accessing the same site. Simply put, WEB access history from January 1st to January 31st is classified for each site. Thereby, a set of a plurality of WEB access histories classified for each site is generated.

  Then, the ranking totaling unit 350 totals the number of histories belonging to each WEB access history set. Simply put, for example, by classifying the WEB access history from January 1st to January 31st for each site, set 1 accessing site A, set 2 accessing site B, and accessing site C When 100 sets such as the set 3 to be generated are generated, the number of histories included in the set is aggregated for each of the sets 1 to 100.

  When the aggregation up to the number of histories is completed, the ranking aggregation unit 350 records the ranking aggregation result of the analysis target user in the ranking aggregation result storage unit 360. Specifically, as shown in FIG. 5, for each user ID (User ID: 12345678), a site name (Site) is associated with a ranking (Ranking, “1”) indicating a ranking with a large number of totalized histories. Name, “www.example.com”) and the total number of histories (Access Count, “1234”) are associated with each other and recorded in the ranking aggregation result storage unit 360. FIG. 5 is a diagram illustrating a configuration example of the ranking count result according to the first embodiment.

  The ranking tabulation result storage unit 360 stores the ranking tabulation results recorded by the ranking tabulation unit 350 for each user. For example, as shown in FIG. 5, ranking (Ranking) indicating the order of the number of accesses, the site name (Site Name, “www.example.com”) of the access destination, and the number of accesses (Access Count, “1234”) ) In association with each other (for example, User ID: 12345678). In the ranking column, numbers “1, 2, 3,...” Are inserted in advance from the top column, indicating that the number of accesses is higher in order from ranking 1.

  When the power index calculation unit 370 acquires the ranking aggregation result of the analysis target user from the ranking aggregation result storage unit 360 and expands it into a graph with the number of accesses as the vertical axis and the ranking as the horizontal axis (plot the ranking aggregation result) The index for identifying the behavioral orientation of the WEB access history of the user to be analyzed is calculated for each total unit period by approximating the distribution tendency of the ranking tabulation results at the time of

  Specifically, for example, as shown in FIG. 6, the power index calculation unit 370 uses the analysis target user (for example, user ID) included in the power index calculation request as a key to calculate the ranking aggregation result of the analysis target user. Obtained from the ranking tabulation result storage unit 360. Then, the exponent calculation unit 370 plots each data of the ranking aggregation result obtained from the ranking aggregation result storage unit 360 on a graph with the number of accesses as the vertical axis (logarithmic axis) and the ranking as the horizontal axis (logarithmic axis). (Refer to (1) in FIG. 6).

When the plotting of each data of the ranking aggregation result is completed, the power index calculation unit 370 displays the data distribution plotted on the graph with the number of accesses as the vertical axis and the ranking as the horizontal axis, on the vertical axis and the horizontal axis that are logarithmic axes. Sampling is performed at equal intervals (see (2) in FIG. 6). Then, the exponent calculation unit 370 obtains an approximate expression (f (x) = ax ^k ) of the sampled data by using, for example, the least square method (see (3) in FIG. 6), and totals the totals. The exponent of the unit period is calculated (variable (k) in the approximate expression). FIG. 6 is a diagram for explaining an index calculation procedure according to the first embodiment.

  For example, in the example of the ranking totalization unit 350 described above, the WEB access history from January 1 to January 31 is classified for each site, thereby accessing the set 1 accessing the A site and the B site. Ranking aggregation results for 100 sets such as set 2 and set 3 accessing the C site are plotted on a graph and sampled at equal intervals on the logarithmic axis.

Then, an approximate expression (f (x) = ax ^k ) of the sampled data is obtained, and a power index of the WEB access history regarding the total unit period from January 1 to January 31 is calculated. In the same procedure, by calculating an approximate expression for the total unit period from February 1 to February 28 (or 29), March 1 to March 31, and so on, it is included in the total period “1 year” The power index for all the aggregate unit periods “one month” is derived, and the derived power index is output to the power index calculation result output unit 380 as the power index calculation result.

  The power exponent calculation unit 370 is not limited to the case where the power exponent is calculated by the above-described method. For example, the power exponent calculation unit 370 divides a section having a small value that is out of the tendency of the data distribution plotted on the graph. The power exponent can be calculated using a method such as applying the least square method.

  As described above, it is not limited to the case where the result of ranking aggregation of the number of accesses aggregated in units of sites is adopted as a power index calculation target. For example, domain unit, service unit, product genre unit, brand unit, The results of ranking the number of accesses, such as genre units, can also be used as power index calculation targets.

  The power index calculation result output unit 380 outputs the power index calculation result received from the power index calculation unit 370 to the analyzer terminal 400. Specifically, for example, as shown in FIG. 7, an index that should be calculated every month that is a total unit period is output to the analyzer terminal 400 over one year that is a total period. FIG. 7 is a diagram illustrating a configuration example of an exponent calculation result according to the first embodiment.

  When the target period log extraction unit 390 receives an input of the log extraction specification range from the analyst terminal 400, the action log storage unit 320 and the purchase log storage unit 340 store the action log and purchase log for the period specified as the log extraction specification range. Respectively. Then, the target period log extraction unit 390 outputs the extracted action log and purchase log to the analyzer terminal 400.

[Process of Example 1]
FIG. 8 is a diagram for explaining the processing flow of the analysis server according to the first embodiment. As shown in the figure, when the ranking totaling unit 350 receives a power index calculation request of the analysis target user from the analyst terminal 400 (step S1), the ranking totaling unit 350 executes ranking totalization for each site (step S2).

  Specifically, the ranking tabulation unit 350 acquires and acquires the WEB access history of the analysis target user for the tabulation period using the user ID and the tabulation period included in the index calculation request that should be received from the analyzer terminal 400 as keys. The WEB access history is divided for each aggregation unit period included in the power index calculation request. Each WEB access history divided for each aggregation unit period is classified into a plurality of WEB access history sets based on whether or not the same site has been accessed, and belongs to each classified WEB access history set. Total the number of histories. When the aggregation up to the number of histories is completed, the ranking aggregation unit 350 records the ranking aggregation result (see, for example, FIG. 5) of the analysis target user in the ranking aggregation result storage unit 360.

  When the ranking aggregation by the ranking aggregation unit 350 is completed, the power index calculation unit 370 acquires the ranking aggregation result of the analysis target user from the ranking aggregation result storage unit 360, and calculates a power index for each aggregation unit period (step S3). ).

  Specifically, the power index calculation unit 370 acquires the analysis target user's ranking total result from the ranking total result storage unit 360 using the analysis target user (for example, user ID) included in the power index calculation request as a key. . Then, the exponent calculation unit 370 plots each data of the ranking aggregation result obtained from the ranking aggregation result storage unit 360 on a graph with the number of accesses as the vertical axis (logarithmic axis) and the ranking as the horizontal axis (logarithmic axis). (For example, see (1) in FIG. 6).

When the plotting of each data of the ranking summary result is completed, the power index calculation unit 370 samples the data distribution plotted on the graph at equal intervals on the vertical axis and the horizontal axis that are logarithmic axes (for example, ( 2)). Then, the exponent calculation unit 370 obtains an approximate expression (f (x) = ax ^k ) of the sampled data by using, for example, the least square method (see (3) in FIG. 6), and totals the totals. The exponent of the unit period is calculated (variable (k) in the approximate expression).

  The power index calculation result output unit 380 outputs the power index calculation result calculated by the power index calculation unit 370 to the analyzer terminal 400 (step S4).

  When the target period log extraction unit 390 receives an input of the log extraction specification range from the analyst terminal 400, the action log storage unit 320 and the purchase log storage unit 340 store the action log and purchase log for the period specified as the log extraction specification range. And the extracted action log and purchase log are output to the analyst terminal 400 (step S5).

[Effects of Example 1]
As described above, according to the first embodiment, the analysis server 300 captures the user behavior orientation indicated by the WEB access history from the distribution tendency of the WEB access history as information for specifying the user behavior orientation. Calculate the exponent of power.

  That is, by performing a simple process of calculating a power index using the distribution tendency of the user's WEB access history becoming a power distribution tendency, for example, data related to the user can be obtained by a questionnaire as in the past. It is possible to more easily extract information for capturing the user's behavior orientation from the user's behavior history without taking time and cost than acquiring. The index that should be provided to the analyst terminal 400 as information for capturing the user's behavioral orientation is used as described below, for example.

  For example, as illustrated in FIG. 9, when the power index changes greatly from July 1 to July 31 to August 1 to August 31, It can be analyzed that there is a high possibility that the period from Sunday to July 31 to August 1 to August 31 is a change point at which a certain purpose of the analysis target user is achieved. And the analyzer terminal 400 acquires the action log and purchase log of the period from July 1 to July 31 to August 1 to August 31 from the analysis server 300, so that the behavior of the user is further increased. The change factors can be analyzed in detail. FIG. 9 is a diagram for explaining the effect according to the first embodiment.

  As shown in FIG. 10, when the power index is changing, the behavior of the target unit of the analysis target user can be extracted by cutting out the power index calculation result on the analyzer terminal 400 side in FIG. . Further, if the index calculation result on the side of the analyzer terminal 400 is cut out in FIG. 2B, it is possible to extract only behaviors with strong diversity orientation. Furthermore, if the calculation result of the exponent on the side of the analyzer terminal 400 is cut out in FIG. 3 (3), it is possible to extract only actions with strong target orientation. FIG. 10 is a diagram for explaining the effect according to the first embodiment.

  Then, for example, a purchase log related to the diversity-oriented behavior extracted in (2) of FIG. 10 is acquired, and a purchase log related to the goal-oriented behavior extracted in (3) of FIG. 10 is acquired. Thus, analyzing both purchase logs individually increases the possibility of more clearly analyzing the difference between purchasing factors when diversity orientation is strong and purchasing factors when goal orientation is strong.

  Whether the purpose of the user's behavior that appears in the user's WEB access history is clear (whether the purpose is clear) or not (clearly) It can also be used as an indicator of

  In the first embodiment, the case where the WEB access history set of the analysis target user is classified based on the site name of the site accessed by the user has been described. However, the present invention is not limited to this, and the user's WEB access history may be classified based on the domain name of the site, the genre name of the site, or the brand name of the product sold on the site. By doing in this way, it becomes possible to analyze a user's action orientation from various angles in the analyzer terminal 400.

  Hereinafter, Example 2 will be described as another embodiment of the information processing apparatus, the information processing system, the information processing method, and the information processing program according to the present invention.

(1) Device Configuration, etc. Each component of the analysis server 300 shown in FIG. 2 is functionally conceptual, and does not necessarily need to be physically configured as illustrated. The specific form of integration is not limited to that shown in the figure. For example, the ranking totaling unit 350, the exponent calculation unit 370, and the exponent calculation result output unit 380 shown in FIG. 2 are configured functionally or physically distributed and integrated.

  As described above, all or a part of the analysis server 300 can be configured to be functionally or physically distributed / integrated in an arbitrary unit according to various loads or usage conditions. Furthermore, each processing function (see FIG. 8 and the like) performed by the analysis server 300 is realized by a CPU and a program that is analyzed and executed by the CPU, or by wired logic. It can be realized as hardware.

(2) Information processing program Various processes (see FIG. 8 and the like) of the analysis server 300 described in the first embodiment are executed by a computer system such as a personal computer or a workstation. Can be realized.

  Therefore, in the following, an example of a computer that executes an information processing program that realizes a function similar to the function of the analysis server 300 described in the above embodiment will be described with reference to FIG. FIG. 11 is a diagram illustrating a computer that executes an information processing program.

  As shown in the figure, the computer 500 as the analysis server 300 is configured by connecting a communication control unit 510, an HDD 520, a RAM 530, and a CPU 540 via a bus 600.

  Here, the communication control unit 510 controls communication related to the exchange of various information. The HDD 520 stores information necessary for the CPU 540 to execute various processes. The RAM 530 temporarily stores various information. The CPU 540 executes various arithmetic processes.

  As shown in FIG. 11, the HDD 520 stores in advance an information processing program 521 that exhibits the same function as each processing unit of the analysis server 300 shown in the above embodiment, and information processing data 522. ing. Note that the information processing program 521 can be appropriately distributed and stored in a storage unit of another computer that is communicably connected via a network.

  The CPU 540 reads out the information processing program 521 from the HDD 520 and expands it in the RAM 530, whereby the information processing program 521 functions as an information processing process 531 as shown in FIG. That is, the information processing process 531 reads out the information processing data 522 and the like from the HDD 520, expands the data in the area allocated to itself in the RAM 530, and executes various processes based on the expanded data and the like.

  Note that the information processing process 531 corresponds to, for example, processing executed in the ranking tabulation unit 350, the power index calculation unit 370, and the like of the analysis server 300 illustrated in FIG.

  The information processing program 521 is not necessarily stored in the HDD 520 from the beginning. For example, a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, an IC set in the computer 500 is not necessary. Each program is stored in a “portable physical medium” such as a card, and “another computer (or server)” connected to the computer 500 via a public line, the Internet, a LAN, a WAN, or the like. The computer 500 may read out and execute each program from these.

(3) Information Processing Method For example, the following information processing method is realized by the analysis server 300 described in the above embodiment.

  That is, a behavior history data collecting step for collecting a plurality of behavior history data from a plurality of user terminals, and a behavior history storage unit for storing a plurality of behavior history data collected by the behavior history data collection step for each user, for a predetermined period The behavior history data of the user to be analyzed stored in is read and divided every predetermined unit period, and each divided behavior history data is classified into a plurality of behavior history data sets based on a predetermined classification criterion, A data aggregation step for aggregating the number of histories belonging to each classified action history data set (see, for example, step S2 in FIG. 8), the number of histories totaled by the data aggregation step, and the action history data set based on the number of histories The history number and rank information are read from the total information storage section that stores the rank order information in association with each other. By approximating the distribution tendency of each behavior history data set when each behavior history data set is expanded to a graph with the number of records and rank information as parameters, using a predetermined function, An index calculating step for calculating an index for identifying each predetermined unit period (see, for example, step S3 in FIG. 8), an analyzer terminal for analyzing the user's behavioral orientation is subjected to a predetermined unit period by the index calculating step. An information processing method including an index calculation result output unit that outputs an index calculated for each as an index calculation result (see, for example, step S4 in FIG. 8) is realized.

  As described above, the information processing apparatus, the information processing system, the information processing method, and the information processing program according to the present invention are useful when information processing using user behavior history data is performed. It is suitable for easily extracting information for grasping the user's behavioral orientation from behavior history data about behavior and purchasing behavior.

DESCRIPTION OF SYMBOLS 1 Internet 100 User terminal 200 EC server 300 Analysis server 310 Action log acquisition part 320 Action log storage part 330 Purchase log acquisition part 340 Purchase log storage part 350 Ranking total part 360 Ranking total result storage part 370 Power index calculation part 380 Power index calculation Result output unit 390 Target period log extraction unit 400 Analyst terminal 500 Computer 510 Communication control unit 520 HDD (Hard Disk Drive)
521 Information processing program 522 Information processing data 530 RAM (Random Access Memory)
531 Information Processing Process 540 CPU (Central Processing Unit)
600 buses

Claims (8)

An information processing apparatus that executes information processing using user behavior history data,
An action history data collection unit for collecting a plurality of action history data from a plurality of user terminals;
A behavior history storage unit that stores a plurality of behavior history data collected by the behavior history data collection unit for each user;
Among the plurality of behavior history data stored in the behavior history storage unit, the behavior history data of the analysis target user is acquired for a predetermined period, and the acquired behavior history data of the user is divided for each predetermined unit period, A data aggregation unit that classifies each divided behavior history data into a plurality of behavior history data sets based on a predetermined classification criterion, and totals the number of histories belonging to each classified behavior history data set;
A total information storage unit that stores the number of histories totaled by the data totaling unit and the rank information that ranks action history data sets according to the number of histories in association with each other;
By approximating the distribution tendency of each behavior history data set when each behavior history data set is expanded to a graph using the number of histories and the rank information as parameters using a predetermined function, An index calculating unit that calculates an index for specifying the orientation for each predetermined unit period;
An analyzer terminal that analyzes a user's behavioral orientation using the index includes an index calculation result output unit that outputs an index calculated for each predetermined unit period by the index calculation unit as an index calculation result. A characteristic information processing apparatus.   The data totaling unit accepts the site name, domain name, genre name or brand name of the site accessed by the user from the analyst terminal as the predetermined classification standard, and the divided action history data is the site name, The information processing according to claim 1, wherein the information is classified into a plurality of action history data sets based on a domain name, a genre name, or a brand name, and the number of histories belonging to each classified action history data set is totaled. apparatus. An information processing system having an information processing apparatus that executes information processing using user behavior history data acquired from a user terminal,
The information processing apparatus includes:
An action history data collection unit for collecting a plurality of action history data from a plurality of user terminals;
A behavior history storage unit that stores a plurality of behavior history data collected by the behavior history data collection unit for each user;
Among the plurality of behavior history data stored in the behavior history storage unit, the behavior history data of the analysis target user is acquired for a predetermined period, and the acquired behavior history data of the user is divided for each predetermined unit period, A data aggregation unit that classifies each divided behavior history data into a plurality of behavior history data sets based on a predetermined classification criterion, and totals the number of histories belonging to each classified behavior history data set;
A total information storage unit that stores the number of histories totaled by the data totaling unit and the rank information that ranks action history data sets according to the number of histories in association with each other;
By approximating the distribution tendency of each behavior history data set when each behavior history data set is expanded to a graph using the number of histories and the rank information as parameters using a predetermined function, An index calculating unit that calculates an index for specifying the orientation for each predetermined unit period;
An analyzer terminal that analyzes a user's behavioral orientation using the index includes an index calculation result output unit that outputs an index calculated for each predetermined unit period by the index calculation unit as an index calculation result. A featured information processing system.   The data aggregation unit receives, as the predetermined classification standard, a site name, domain name, genre name or brand name of a site accessed by a user from the analyst terminal, and the divided action history data is converted to the site name, domain 4. The information processing system according to claim 3, wherein the information is classified into a plurality of action history data sets based on a name, a genre name, or a brand name, and the number of histories belonging to each classified action history data set is totaled. . An information processing method executed by an information processing apparatus that performs information processing using user behavior history data,
An action history data collection step for collecting a plurality of action history data from a plurality of user terminals;
From the behavior history storage unit that stores, for each user, a plurality of behavior history data collected by the behavior history data collection step, the behavior history data of the analysis target user stored within a predetermined period is read for each predetermined unit period A data aggregation step for classifying the divided behavior history data into a plurality of behavior history data sets based on a predetermined classification criterion, and totaling the number of histories belonging to each classified behavior history data set;
The history number and the rank information are read from the total information storage unit that stores the history number totaled in the data aggregation step and the rank information that ranks the action history data set according to the history number in association with each other. By approximating the distribution tendency of each behavior history data set when each behavior history data set is expanded to a graph using the read history number and rank information as parameters, using a predetermined function An index calculating step for calculating an index for identifying the user's action orientation for each predetermined unit period;
An analyzer terminal that analyzes a user's behavior orientation using the index includes an index calculation result output unit that outputs the index calculated for each predetermined unit period by the index calculation step as an index calculation result. An information processing method characterized by the above.   In the data aggregation step, as the predetermined classification standard, the site name, domain name, genre name or brand name of the site accessed by the user is received from the analyst terminal, and the divided action history data is received as the site name, domain 6. The information processing method according to claim 5, wherein the information is classified into a plurality of action history data sets based on a name, a genre name, or a brand name, and the number of histories belonging to each classified action history data set is totaled. . An information processing program for causing a computer as an information processing apparatus to perform information processing using user behavior history data,
Action history data collection procedure for collecting a plurality of action history data from a plurality of user terminals,
From the behavior history storage unit that stores a plurality of behavior history data collected by the behavior history data collection procedure for each user, the behavior history data of the analysis target user stored within a predetermined period is read for each predetermined unit period A data summarizing procedure for classifying each of the divided behavior history data into a plurality of behavior history data sets based on a predetermined classification criterion, and totaling the number of histories belonging to each classified behavior history data set;
The history number and the ranking information are read from the counting information storage unit that stores the history number totaled by the data counting procedure and the rank information that ranks the action history data set according to the history number in association with each other. By approximating the distribution tendency of each behavior history data set when each behavior history data set is expanded to a graph using the read history number and rank information as parameters, using a predetermined function , An index calculation procedure for calculating an index for identifying the user's action orientation for each predetermined unit period;
An index calculation result output unit that outputs an index calculated for each predetermined unit period by the index calculation procedure as an index calculation result to an analyst terminal that analyzes a user's behavioral orientation using the index. An information processing program that is executed.   In the data aggregation procedure, as the predetermined classification standard, the site name, domain name, genre name or brand name of the site accessed by the user is accepted from the analyst terminal, and the divided action history data is received as the site name, domain 8. The information processing program according to claim 7, wherein the information is classified into a plurality of action history data sets based on a name, a genre name, or a brand name, and the number of histories belonging to each classified action history data set is totaled. .

Images