Behavior analysis device and behavior analysis method
The behavior analysis device addresses the challenge of analyzing complex user data by prioritizing and visually highlighting important actions, enhancing the efficiency and clarity of user behavior analysis.
Patent Information
- Application Number
- PCT/JP2025/017344
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-05-13
- Publication Date
- 2026-11-19
AI Technical Summary
Conventional information collection systems provide analysts with vast amounts of time-series data for user behavior analysis, leading to increased analysis time due to the complexity and volume of data from multiple services used by a user.
A behavior analysis device that acquires log data, determines the importance of user actions based on log data, and displays user actions with varying prominence based on their importance, using a customer journey map or timeline for efficient analysis.
Enables analysts to efficiently analyze user behavior by prioritizing and visually highlighting important actions, reducing analysis time and improving the clarity of user interaction insights.
Smart Images

Figure JP2025017344_19112026_PF_FP_ABST
Abstract
Description
Behavior Analysis Device and Behavior Analysis Method
[0001] The present disclosure relates to a behavior analysis device and a behavior analysis method.
[0002] Patent Document 1 discloses an information collection system that collects data related to blogs of the same user. The information collection system includes an information collection device and a plurality of server devices. The information collection device collects data of the same user from the plurality of server devices. Each of the plurality of server devices stores data having different data items for the same user. The information collection device extracts the update date and time, which is a common attribute, from the plurality of collected data. The information collection device sorts the plurality of collected data based on the extracted update date and time. The information collection device sorts all the data collected from the plurality of server devices and provides the sorted data to an analyst. The sorted plurality of data is time-series blog data written by the same user.
[0003] Japanese Unexamined Patent Application Publication No. 2014-127097
[0004] Meanwhile, in the planning of marketing strategies, there are cases where it is desired to focus on a single user and analyze the behavior of the user. When a conventional information collection device is used for this analysis, a plurality of time-series data indicating the user's behavior is provided to the analyst. For example, when a user uses a wide variety of services, a plurality of time-series data indicating the user's behavior for each service is provided to the analyst.
[0005] However, there has been a problem that the plurality of time-series data provided to the analyst becomes enormous, which increases the time required for the analyst to perform the analysis.
[0006] The present disclosure aims to provide a behavior analysis device that enables an analyst to efficiently analyze a user's behavior.
[0007] An action analysis device according to one aspect of the present disclosure includes: an acquisition unit that acquires log data recording a plurality of actions performed by each of a plurality of users; a determination unit that determines the importance of each of the plurality of actions based on the log data, wherein the log data includes a plurality of unit data corresponding one-to-one with the plurality of actions performed by each of the plurality of users, the plurality of actions are classified into a plurality of types, and each of the plurality of unit data includes identification information of the user who performed the corresponding action and information identifying the type of the corresponding action; and a display control unit that displays an object on a display device based on at least one unit data indicating an action performed by a user selected from among the plurality of users, wherein the display control unit changes the display mode of the object based on the determined importance.
[0008] An action analysis method according to one aspect of the present disclosure acquires log data recording multiple actions performed by each of a plurality of users, the log data includes a plurality of unit data corresponding one-to-one with the plurality of actions performed by each of the plurality of users, the plurality of actions are classified into a plurality of types, each of the plurality of unit data includes identification information of the user who performed the corresponding action and information identifying the type of the corresponding action, the importance of each of the plurality of types is determined based on the log data, and an object based on at least one unit data representing an action performed by a user selected from among the plurality of users is displayed on a display device. The display mode of the object is changed based on the determined importance.
[0009] According to one aspect of this disclosure, analysts can efficiently analyze user behavior.
[0010] This is a block diagram showing an example of the overall configuration of the data processing system 100 according to the embodiment. This is a block diagram showing an example of the configuration of the behavioral analysis device 2. This is a schematic diagram showing an example of service data SD, intermediate data XD, and analysis data ZD. This is a schematic diagram showing an example of data items for intermediate data XD. This is a schematic diagram showing an example of personal data PD. This is a schematic diagram showing an example of the display mode of the customer journey map MP. This is a schematic diagram showing an example of the display mode of the customer journey map MP. This is a schematic diagram showing an example of the display mode of the timeline TL. This is a schematic diagram showing an example of the display mode of the timeline TL. This is a schematic diagram showing an example of the display mode of the timeline TL. This is a flowchart showing an example of the operation of the processing device 21 of the behavioral analysis device 2.
[0011] [Embodiment] [System Configuration] Figure 1 is a block diagram showing an example of the overall configuration of the data processing system 100 according to the embodiment. The data processing system 100 is a system for analyzing the behavior of multiple users U[1], ... U[q], ... U[Q]. Q is an integer of 2 or more. Analyst B uses the data processing system 100 to analyze the behavior of multiple users U[1], ... U[q], ... U[Q]. The results of the behavioral analysis are used for marketing or service improvement.
[0012] The data processing system 100 comprises Q terminal devices 1[1] to 1[Q], a behavior analysis device 2, a data processing server 3, and a first service server A1, a second service server A2, ... a yth service server Ay, where y is an integer of 2 or more. Each element of the data processing system 100 is connected via a network NW.
[0013] Q terminal devices 1[1] to 1[Q] are used by Q users U[1] to U[Q]. In the following description, terminal devices 1[1] to 1[Q] will be referred to as terminal device 1 when not distinguished, and users U[1] to U[Q] will be referred to as user U when not distinguished. Each of the multiple users U is assigned a user ID that uniquely identifies user U. Multiple user IDs correspond one-to-one with multiple users U.
[0014] Multiple terminal devices 1 include, for example, smartphones or personal computers. Each of the multiple users U uses multiple services. The first service server A1, the second service server A2, ... the yth service server Ay each provide different services to the multiple users U. Examples of services provided to the multiple users U include mobile phone services, electronic payment services, video streaming services, call center response services, store visit reservation services, and information provision services on websites.
[0015] Each of the first service server A1, second service server A2, ... yth service server Ay provides the various services described above to multiple users U. When the first service server A1 to the yth service server Ay are not distinguished, they are referred to as service server A. Each of the first service server A1, second service server A2, ... yth service server Ay records logs related to the services it provides as service data SD. More specifically, the first service server A1 records logs related to the first service as first service data SD1. The second service server A2 records logs related to the second service as second service data SD2. The yth service server Ay records logs related to the yth service as yth service data SDy. Furthermore, each of the first service server A1, second service server A2, ... yth service server Ay transmits the service data SD to the data processing server 3 at predetermined timings. The predetermined timing could be, for example, every certain amount of time, or it could be when the log in the service data SD has increased by a certain amount since the previous service data SD was sent.
[0016] The first service data SD1, the second service data SD2, ... the yth service data SDy each have different data items. In the following explanation, when service data SD1 to SDy are not distinguished, they will be referred to as service data SD. A data item is an element that makes up data. When the data items of two data are different, it means that the elements that make up one data are different from the elements that make up the other data. When the data items of two data are common to each other, it means that the elements that make up one data are the same from the elements that make up the other data.
[0017] The data processing server 3 comprises a processing unit 31 and a storage device 32. The processing unit 31 of the data processing server 3 converts multiple service data SDs, each with different data items, into multiple intermediate data XDs, each with common data items. The multiple intermediate data XD1, XD2, ... XDy correspond one-to-one with the multiple service data SD1, SD2, ... SDy. The multiple intermediate data XD1, XD2, ... XDy are stored in the storage device 32. In the following description, if the intermediate data XD1 to XDy are not distinguished, they will be referred to as intermediate data XD. The processing unit 31 of the data processing server 3 uses the multiple intermediate data XDs to create analysis data ZD in response to the request of analyst B (the first request, described later).
[0018] The behavioral analysis device 2 is an information processing device used by analyst B to analyze the behavior of user U. The behavioral analysis device 2 is, for example, a desktop computer or a notebook computer. Analyst B uses the behavioral analysis device 2 to send a first request to the data processing server 3. The first request is a request to the data processing server 3 to create and send analysis data ZD.
[0019] The first request includes target conditions and extraction conditions. The target conditions specify which of the multiple intermediate data XD will be used for extraction. The target conditions can also be described as specifying which service logs from multiple services will be included in the analysis data ZD. The extraction conditions specify what kind of logs (unit data) will be extracted from one or more intermediate data XD that satisfy the target conditions. For example, the extraction conditions specify the user ID and time period to be extracted.
[0020] In this embodiment, intermediate data XD to be included in the analysis data ZD can be specified according to the target conditions, so analyst B can obtain analysis data ZD that does not contain noise (logs containing content unrelated to the analysis target). Furthermore, since the analysis data ZD is extracted from the intermediate data XD specified according to the target conditions, the processing load for generating the analysis data ZD is reduced compared to the case where the analysis data ZD is extracted from all intermediate data XD.
[0021] Analyst B obtains analysis data ZD from data processing server 3 and performs various analyses using analysis data ZD. In this embodiment, analyst B selects one user U from among multiple users U and performs an individual analysis of the behavior of the selected user U (hereinafter referred to as selected user U[q], where q is any integer satisfying 1 ≤ q ≤ Q). In the individual analysis, rather than the behavior of users U as a group (statistical value), the behavior of individual users U is examined in detail to verify what factors motivate their behavior or what factors act as barriers to their behavior. For the individual analysis, the behavior analysis device 2 extracts only the log of selected user U[q] as individual data PD from analysis data ZD which contains logs of multiple users U. The behavior analysis device 2 displays the individual data PD in a format such as a customer journey map MP (see Figure 6A, etc.) or a timeline TL (see Figure 7A, etc.), which will be described later.
[0022] The data processing system 100 may also include multiple behavioral analysis devices 2. In this case, each of the multiple analysts B sends a first request to the data processing server 3 using their corresponding behavioral analysis device 2.
[0023] Figure 3 is a schematic diagram showing an example of service data SD, intermediate data XD, and analysis data ZD. Each row (each record) of service data SD, and the intermediate data XD, analysis data ZD, and personal data PD described below, will be referred to as "unit data" below. Unit data is generated, for example, each time an event occurs.
[0024] The first service data SD1 recorded by the first service server A1 includes, for example, the time when an action (event) occurred, the user ID, item 1-1, item 1-2, and item 1-3. Each row of the first service data SD1 is unit data indicating the action of user U on the first service that occurred in the first service.
[0025] Furthermore, the second service data SD2 recorded by the second service server A2 includes, for example, time, user ID, item 2-1, item 2-2, and item 2-3. Each row of the second service data SD2 is unit data indicating the user U's actions on the second service. Items 1-1, 1-2, and 1-3 are different from items 2-1, 2-2, and 2-3.
[0026] In Figure 3, for illustrative purposes, only hours, minutes, and seconds are shown as time; however, in reality, time includes year, month, and day in addition to hours, minutes, and seconds. Also, in Figure 3, the user ID is common to both the first service data SD1 and the second service data SD2. In other words, Figure 3 assumes that a common user ID is used for both the first and second services. However, if the user IDs for the first and second services can be converted to each other, the user IDs for the first and second services may be different.
[0027] The data processing server 3 obtains the first service data SD1 from the first service server A1 and converts the first service data SD1 into the first intermediate data XD1. The data processing server 3 also obtains the second service data SD2 from the second service server A2 and converts the second service data SD2 into the second intermediate data XD2. The first intermediate data XD1 and the second intermediate data XD2 each include the time, user ID, item T-1, item T-2, and item T-3, respectively.
[0028] When converting the first service data SD1 to the first intermediate data XD1, items 1-1 to 1-3 are converted to items T-1 to T-3. Similarly, when converting the second service data SD2 to the second intermediate data XD2, items 2-1 to 2-3 are converted to items T-1 to T-3.
[0029] For example, if item 1-1 corresponds to item T-1, the data "AAA" in item 1-1 is converted to the data "MMM" in item T-1. "MMM" is "AAA" converted to conform to the format of the intermediate data XD. Therefore, "AAA" and "MMM" essentially represent the same content. For example, items with the same format between the service data SD and the intermediate data XD do not need to be converted.
[0030] Figure 4 is a schematic diagram showing an example of data items in intermediate data XD. The multiple unit data that make up intermediate data XD each include time (date and time), user ID, channel (e.g., item T-1), target of action (e.g., item T-2), and action content (e.g., item T-3).
[0031] The time indicates the date and time when user U performed an action (an event occurred) on the service. The user ID is the identification information of user U who performed the action. The channel indicates the point of contact between user U and the service at the time of the action. That is, the channel indicates which of the multiple services, from the first service to the yth service, user U used. For example, if the service used by user U is a response service, the channel is the call center. If the service used by user U is a store visit reservation service, the channel is the name of the reserved store. If the service used by user U corresponds to an information provision service on a website, the channel is the website. The target of the action is the target of the action. For example, if user U purchases a product using an electronic payment service, the target of the action is the product name. The content of the action indicates the specific details of the action taken by user U. For example, if user U cancels a mobile phone service contract, the content of the action is contract cancellation.
[0032] The multiple data items included in the intermediate data XD correspond to the so-called 4W1H. The time of the action corresponds to "when". The user ID corresponds to "who". The channel corresponds to "where". The target of the action corresponds to "what". The content of the action corresponds to "how". In this way, by converting multiple service data SDs, which have different items from each other, into multiple intermediate data XDs that share common items, it becomes easier to combine and analyze logs from multiple different services.
[0033] Next, the analysis data ZD will be described. When the data processing server 3 receives the first request from the behavioral analysis device 2 (analyst B), it creates the analysis data ZD based on the first request. The data processing server 3 then transmits the created analysis data ZD to the behavioral analysis device 2. As described above, it includes the target conditions and the extraction conditions.
[0034] For example, suppose the target conditions are "First Intermediate Data XD1 and Second Intermediate Data XD2," and the extraction conditions are "User ID is 002 or 003" and "Time is between 12:15:00 and 12:59:59." In this case, the data processing server 3 extracts unit data from "First Intermediate Data XD1 and Second Intermediate Data XD2" where "User ID is 002 or 003" and "Time is between 12:15:00 and 12:59:59." One or more extracted unit data becomes the analysis data ZD.
[0035] Specifically, for example, in the unit data included in the first intermediate data XD1, the first row does not meet the extraction criteria because the user ID is 001. The second row meets the extraction criteria because the user ID is 002 and the time falls within the period from 12:15:00 to 12:59:59. Similarly, in the unit data included in the second intermediate data XD2, the first row does not meet the extraction criteria because the time is before 12:15:00. The second row meets the extraction criteria because the user ID is 003 and the time falls within the period from 12:15:00 to 12:59:59. The data processing server 3 creates analysis data ZD, which includes the unit data from the second row of the first intermediate data XD1 and the unit data from the second row of the second intermediate data XD2. Generally, the extracted unit data is arranged in chronological order. Furthermore, while the analysis data ZD shown in Figure 3 is composed of unit data from two users U for the sake of explanation, in reality, the analysis data ZD may contain unit data from dozens to tens of thousands of users U.
[0036] Here, we will explain the "type of action". In this embodiment, the user ID and period are specified as extraction conditions for the analysis data ZD. That is, "when" and "who" are specified from among the multiple data items included in the intermediate data XD. Therefore, the "type of action" is defined by the combination of the contents (values) of items T-1 to T-3 that correspond to "where (channel)", "what (target of action)", and "how (action content)". More specifically, for example, in the analysis data ZD shown in Figure 3, an action where item T-1 is "VVV", item T-2 is "WWW", and item T-3 is "XXX" is defined as "type α". Similarly, an action where item T-1 is "PPP", item T-2 is "QQQ", and item T-3 is "RRR" is defined as "type γ". Behavior where item T-1 is "AAB", item T-2 is "BBC", and item T-3 is "CCD" is defined as "Type β".
[0037] Next, we will explain personal data PD. Personal data PD is data extracted from analysis data ZD for a specific user U (selected user U[q]). In this embodiment, personal data PD is created by the behavioral analysis device 2 using analysis data ZD received from data processing server 3. When the user ID of selected user U[q] is specified by analyst B, the behavioral analysis device 2 extracts unit data containing the user ID of selected user U[q] from the unit data included in analysis data ZD.
[0038] Figure 5 is a schematic diagram showing an example of personal data PD. In the example in Figure 5, user U with user ID "003" is designated as selected user U[q]. Therefore, the personal data PD shown in Figure 5 includes the unit data from the first row and the unit data from the third row of the analysis data ZD shown in Figure 3. The first row of personal data PD corresponds to the first row of the analysis data ZD in Figure 3 and indicates behavior of "Type α". The second row of personal data PD corresponds to the third row of the analysis data ZD in Figure 3 and indicates behavior of "Type β". The third row of personal data PD is not shown in the analysis data ZD in Figure 3, but the values of items T-1 to T-3 match those of the second row of the analysis data ZD in Figure 3. Therefore, the third row of personal data PD indicates behavior of "Type γ".
[0039] Next, we will explain the display formats of personal data (PD). Personal data (PD) can be displayed in various formats, such as tables or text as shown in Figure 5, and used for behavioral analysis. Examples of visualization formats for personal data (PD) include customer journey maps (MP) and timelines (TL).
[0040] The Customer Journey Map (MP) visualizes the customer journey of a selected user U[q]. A customer journey refers to the series of processes from when a selected user U[q] becomes aware of a product or service, compares and considers it, and ultimately makes a purchase (or decides not to purchase it).
[0041] Figure 6A is a schematic diagram showing an example of how the customer journey map MP is displayed. The customer journey map MP1 shown in Figure 6A shows the actions taken by selected user U[q] when purchasing a new mobile phone. The customer journey map MP1 includes a plurality of points PT that represent the actions taken by selected user U[q], and lines LN1 that connect the plurality of points PT. The plurality of points PT correspond one-to-one with a plurality of unit data in the personal data PD. The points PT are displayed in a field having a vertical axis AX1 and a horizontal axis AX2. The vertical axis AX1 shows the name of the service on which selected user U[q] performed the action. The horizontal axis AX2 shows the time on which selected user U[q] performed the action. In addition, the content DT of the action taken by selected user U[q] is associated with each point PT.
[0042] In the example in Figure 6A, at time t1, user U[q] uses a web service to browse mobile phone models on an online shop, and at time t2, adds the mobile phone to be purchased to their cart. At time t3, user U[q] uses a user application service to begin the process of contracting a new mobile phone service. At time t4, user U[q] uses a web service to refer to the corporate website and confirm the contact information for the call center. At time t5, user U[q] uses the call center's response service to inquire about the next steps. At time t6, user U[q] uses the user application service to complete the process of contracting a new mobile phone service. At time t7, user U[q] uses the store visit reservation service to make a reservation to visit a physical store. At time t8, user U[q] receives the mobile phone purchased at the physical store.
[0043] Furthermore, a timeline TL, as shown in Figure 7A, may be used as a simpler way to visualize the actions of a selected user U[q] than a customer journey map MP.
[0044] FIG. 7A is a schematic diagram illustrating an example of a display mode of a timeline TL. A timeline TL1 shown in FIG. 7A includes a plurality of icons IC representing actions performed by a selected user U[q], and a line LN2 connecting the icons IC. The plurality of icons IC correspond to the plurality of actions performed by the selected user U[q]. In other words, each icon IC corresponds to unit data in personal data PD.
[0045] In FIG. 7A, the plurality of icons IC are displayed with the same color and the same size. In the example of FIG. 7A, the action type such as "type α" is written on the icon IC, but for example, at least one of a channel (item T-1), an action target (item T-2), or an action content (item T-3) may be written thereon. Specifically, for example, text indicating action content such as "model registration" and "add to cart" may be written on the icon IC. In this case, when an analyst B clicks the icon IC or moves a cursor over the icon IC, text indicating the remaining information, that is, the channel and the action target may be displayed. The line LN2 indicates the time (relative time) at which the selected user U[q] performed the corresponding action.
[0046] By using a customer journey map MP or the timeline TL, the analyst B can easily visually grasp a series of actions performed by the selected user U[q]. Here, in the present embodiment, the personal data PD is created using analysis data ZD obtained by integrating logs of a plurality of services. Therefore, as compared with personal data PD created using logs of one service, the number of pieces of unit data increases, and the time required for analysis may become longer. Therefore, the behavior analysis device 2 according to the present embodiment sets an importance level for each of the plurality of pieces of unit data included in the personal data PD. More specifically, the behavior analysis device 2 sets an importance level for the type of action recorded in the unit data. Then, the behavior analysis device 2 changes the display mode of each of the plurality of pieces of unit data on the basis of the importance level. Accordingly, the analyst B can quickly grasp unit data having a high importance level among the plurality of pieces of unit data, and can efficiently analyze the personal data PD.
[0047] [Configuration of Behavior Analysis Device 2] FIG. 2 is a block diagram showing an example of the configuration of the behavior analysis device 2. As shown in FIG. 2, the behavior analysis device 2 includes a processing device 21, a storage device 22, a communication device 23, an input device 24, and a display device 25, each connected to a bus 200 so as to be capable of mutually transmitting and receiving data.
[0048] The processing device 21 includes at least one CPU (Central Processing Unit). The CPU includes interfaces with peripheral devices, arithmetic units, registers, and the like. The at least one CPU is an example of at least one processor. A device including each of the processor and the CPU is an example of a computer. In the processing device 21, circuits such as a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), and an FPGA (Field Programmable Gate Array) may be used in place of the CPU, or in combination with the CPU.
[0049] The storage device 22 is a recording medium readable by the processing device 21. The storage device 22 includes a non-volatile memory and a volatile memory. The non-volatile memory is, for example, a ROM (Read Only Memory), an EPROM (Erasable Programmable Read Only Memory), and an EEPROM (Electrically Erasable Programmable Read Only Memory). The volatile memory is, for example, a RAM (Random Access Memory). The storage device 22 of the present embodiment stores a control program PG. The control program PG is a program for controlling the behavior analysis device 2.
[0050] The communication device 23 is a device that performs communication via a network NW. The communication device 23 communicates with a data processing server 3 under the control of the processing device 21. The input device 24 outputs a signal corresponding to the operation of an analyst B to the processing device 21. The input device 24 corresponds to a keyboard, a pointing device, and the like.
[0051] The display device 25 displays an image based on the image data output from the processing device 21. For example, the display device 25 receives image data generated based on personal data PD (image data showing a timeline TL or a customer journey map MP).
[0052] The processing unit 21 reads the control program PG from the storage device 22. By executing the control program PG, the processing unit 21 functions as a first receiving unit 211, an acquisition unit 212, a determination unit 213, a second receiving unit 214, and a display control unit 215.
[0053] The first reception unit 211 receives the target conditions and extraction conditions (specification of the content of the first request) from analyst B. As described above, the target conditions are conditions that specify which of the multiple intermediate data XD to target for extraction. The first reception unit 211 receives, for example, the specification of one or more services or one or more service servers A from analyst B as the target conditions. The extraction conditions are conditions that specify what kind of unit data to extract from one or more intermediate data XD that satisfy the target conditions. The first reception unit 211 receives, for example, the specification of the user IDs and period to be extracted as extraction conditions. In this embodiment, multiple (two or more) user IDs are specified as extraction conditions.
[0054] The acquisition unit 212 acquires analysis data ZD, which records multiple actions performed by each of the multiple users U. Analysis data ZD is an example of log data. The multiple users U are users U whose user ID is specified as an extraction condition and who have logs in the intermediate data XD specified as a target condition. The acquisition unit 212 sends a first request to the data processing server 3. The data processing server 3 creates analysis data ZD based on the first request and sends it to the data processing server 3. The acquisition unit 212 acquires the analysis data ZD by receiving it from the data processing server 3.
[0055] The analysis data ZD includes multiple unit data points that correspond one-to-one with multiple actions performed by each of multiple users U. These actions are categorized into multiple types. Each unit data point includes identification information for the user U who performed the corresponding action, and information identifying the type of the corresponding action. The user U identification information is the "User ID" shown in Figure 4. The information identifying the type of action is the "Channel (item T-1)", "Action Target (item T-2)", and "Action Content (item T-3)" shown in Figure 4.
[0056] The decision unit 213 determines the importance of each of the multiple types based on the analysis data ZD. In this embodiment, the decision unit 213 calculates a score for each of the multiple types of behavior using one of the methods 1 to 3 shown below. A higher score indicates a more important type of behavior.
[0057] [Method 1] Method 1 will be explained using the example of calculating the importance of the first type (T1) among several types. Method 1 is a method of calculating importance (score) based on the number of users U who performed the first type of behavior among the users U included in the analysis data ZD (population). Hereinafter, the multiple users U specified in the extraction conditions will be referred to as users U[1] to U[n]. n is an integer between 2 and Q. Any one of users U[1] to U[n] will be referred to as user U[i]. i is an integer between 1 and n.
[0058] First, the determination unit 213 calculates the following equation (1). The following equation (1) indicates whether user U[i] performed a first type of action. In other words, the following equation (1) indicates whether there is unit data indicating a first type of action in the unit data of user U[i] (unit data in which the value of user ID indicates the user ID of user U[i]). In the following equation (1), Ui is user U[i] and T1 indicates the first type. If there is unit data indicating a first type of action, the value of the following equation (1) is 1. If there is no unit data indicating a first type of action, the value of the following equation (1) is 0.
[0059]
[0060] Next, the determination unit 213 calculates the following equation (2). The following equation (2) shows how many users U[1] to U[n] performed the first type of action. Note that for scaling purposes, the number of users U who performed the first type of action is divided by n. The value of the following equation (2) becomes larger as the number of users U who performed the first type of action increases.
[0061]
[0062] The determination unit 213 performs the calculations of formulas (1) and (2) above for all types (T1 to Tm, where m is an integer of 1 or more, and the number of types of behavior included in the analysis data ZD) and calculates the importance of each type of behavior.
[0063] Thus, in Method 1, the decision unit 213 determines the importance of the first type of action based on the number of users U who performed the first type of action among a plurality of users U. The decision unit 213 assigns a higher importance to the first type of action the more users U who performed the first type of action are. This makes it possible to identify major actions performed by many users U as actions of high importance. For example, the decision unit 213 may assign a higher importance to the first type of action the fewer users U who performed the first type of action are. This makes it possible to identify minor actions performed by a small number of users U as actions of high importance.
[0064] [Method 2] Method 2 will be explained using the example of calculating the importance of the second type (T2) among several types. Method 2 is a method of calculating importance (score) based on the frequency of occurrence of the second type of behavior in the analysis data ZD (population). First, the decision unit 213 counts the number of unit data representing the second type of behavior N (D_T2) for each unit data in the analysis data ZD. The decision unit 213 also counts the total number of unit data in the analysis data ZD N (N_all). The decision unit 213 takes the value obtained by dividing the number of unit data representing the second type of behavior N (D_T2) by the total number of unit data in the analysis data ZD N (N_all) as the importance EventNum(T2) based on the frequency of occurrence of the second type of behavior. That is, the importance based on the frequency of occurrence of the second type of behavior is shown by the following formula (3).
[0065]
[0066] The value of equation (3) above increases as the number of unit data N(D_T2) representing the second type of behavior increases. The determination unit 213 performs the calculation of equation (3) for all types (T1 to Tm) included in the analysis data ZD and calculates the importance of each type of behavior.
[0067] Thus, in Method 2, the determination unit 213 determines the importance of the second type based on the number of unit data corresponding to the second type included in the analysis data ZD. The determination unit 213 assigns a higher importance to the second type as the number of unit data corresponding to the second type increases. This makes it possible to identify major behaviors that occur frequently as highly important behaviors. Alternatively, the determination unit 213 may assign a higher importance to the second type as the number of unit data corresponding to the second type decreases. This makes it possible to identify minor behaviors that occur infrequently as highly important behaviors.
[0068] [Method 3] Method 3 will be explained using the example of calculating the importance of the third type (T3) among several types. Method 3 is a method of calculating importance (score) based on the characteristicness of the third type of behavior in the analysis data ZD (population). In Method 3, the importance increases as the number of users U who perform the third type of behavior is small, and as the number of times the same user U performs the third type of behavior increases.
[0069] First, the decision unit 213 counts the number of unit data points N (D_T3) that represent the third type of behavior within the analysis data ZD. The decision unit 213 also counts the number of users U (U_T3) who performed the third type of behavior among users U[1] to U[n]. The decision unit 213 then calculates the importance level EventFreq(T3) based on the frequency of occurrence of the third type of behavior by dividing the number of unit data points N (D_T3) that represent the third type of behavior by the number of users U (U_T3) who performed the third type of behavior.
[0070]
[0071] Next, the decision unit 213 calculates UserFreq(T3), the percentage of users U who performed the third type of action among all users U included in the analysis data ZD, using the following formula (5). In the following formula (5), N(U_T3) represents the number of users U who performed the second type of action, and N(U_all) represents the total number of users U included in the analysis data ZD.
[0072]
[0073] The decision unit 213 then calculates a score (T3) indicating the importance of the third type of behavior using the following formula (6). The value of formula (6) increases as the number of users U who perform the third type of behavior decreases. Also, the value of formula (6) increases as the number of times the same user U performs the third type of behavior increases. The decision unit 213 performs calculations using formulas (4) to (6) for all types (T1 to Tm) included in the analysis data ZD to calculate the importance of each type of behavior.
[0074]
[0075] Thus, the determination unit 213 determines the importance of the third type based on the number of unit data corresponding to the third type included in the analysis data ZD and the number of users U who performed the third type action among multiple users U. The determination unit 213 assigns a higher importance to the third type if the number of users U who performed the third type action is small, and if the number of users U who performed the third type action is large. This makes it possible to identify actions that are repeatedly performed by a small number of users U as actions with high importance.
[0076] The second reception unit 214 accepts the designation of selected users U[q] to be extracted from the analysis data ZD for personal data PD. Analyst B designates selected users U[q] from users U[1] to U[n] specified in the extraction conditions of the first request. Analyst B may designate multiple selected users U[q].
[0077] The display control unit 215 displays objects on the display device 25 based on at least one unit data representing an action performed by a user U selected from among multiple users U (selected user U[q]). The display control unit 215 changes the display manner of the objects based on the determined importance. More specifically, the display control unit 215 displays first objects corresponding to types with higher importance more prominently than second objects corresponding to types with lower importance.
[0078] In this embodiment, "high importance" means that the importance value calculated using any of the methods 1 to 3 described above is large. The display control unit 215 may, for example, designate types of actions with an importance value of 1 or higher as "high importance types" and other types of actions (with an importance value less than 1) as "low importance types." Alternatively, the display control unit 215 may, for example, designate types of actions with an importance value of 1 or higher as "high importance types," types of actions with an importance value less than 1 but 2 or higher (2nd threshold < 1st threshold) as "moderate importance types," and types of actions with an importance value less than 2nd threshold as "low importance types."
[0079] Figures 7B to 7D are schematic diagrams showing examples of how the timeline TL is displayed. In the timeline TL shown in Figures 7A to 7D, including Figure 7A which was explained earlier, objects based on unit data correspond to icon ICs. More specifically, one unit data corresponds to one icon IC.
[0080] Figures 7A to 7D display icons IC representing behaviors of type α, type β, type γ, and type δ. Of these behavior types, types α and γ are judged to be of high importance, while types β and δ are judged to be of low importance.
[0081] In the timeline TL2 shown in Figure 7B, the icon ICs representing high-importance types α and γ (corresponding to the first object) and the icon ICs representing low-importance types β and δ (corresponding to the second object) are displayed in different colors. Specifically, for example, the icon ICs representing types α and γ are displayed in chromatic colors, while the icon ICs representing types β and δ are displayed in achromatic colors. In other words, the display control unit 215 displays the first object using chromatic colors and the second object using achromatic colors.
[0082] Generally, icon ICs displayed in chromatic colors are more visible than icon ICs displayed in achromatic colors. Therefore, analyst B perceives that icon ICs corresponding to high-importance types are displayed more prominently than icon ICs corresponding to low-importance types. Note that the difference in display color is not limited to chromatic and achromatic colors; it could also be, for example, a difference in brightness or hue.
[0083] In the timeline TL3 shown in Figure 7C, the icons IC representing high-importance types α and γ (corresponding to the first object) and the icons IC representing low-importance types β and δ (corresponding to the second object) have different sizes (display area). Specifically, for example, the icons IC representing types α and γ are displayed larger than the icons IC representing types β and δ.
[0084] Generally, icon ICs with a relatively large display area are more visible than icon ICs with a smaller display area. Therefore, analyst B perceives that icon ICs corresponding to high-importance types are displayed more prominently than icon ICs corresponding to low-importance types. The display configuration shown in Figure 7C can be further developed to hide icon ICs corresponding to low-importance types.
[0085] In the timeline TL4 shown in Figure 7D, only icons IC (corresponding to the first object) representing the high-importance types α and γ are displayed. Where icons IC (corresponding to the second object) representing the low-importance types β and δ were displayed in Figure 7C, etc., points PT (or icons with a different shape from the icons IC representing the high-importance types α and γ) are displayed. Figure 7D is an example of a display mode in which icons IC corresponding to low-importance types are hidden.
[0086] Furthermore, not only the timeline TL, but also the customer journey map MP may be displayed in a way that reflects the importance of the actions. Figure 6B is a schematic diagram showing an example of the display method of the customer journey map MP. The customer journey map MP2 shown in Figure 6B shows the same content as the customer journey map MP1 shown in Figure 6A. In the customer journey map MP, multiple points PT that indicate the actions taken by the selected user U[q] are objects based on unit data.
[0087] For example, among the actions shown in Figures 6A and 6B, "adding to cart in an online shop" and "making an inquiry in the response service" are considered to be high-priority actions. The display control unit 215 displays point PT-1, which corresponds to "adding to cart in an online shop," and point PT-2, which corresponds to "making an inquiry in the response service," in a different display manner from the other points PT. Specifically, as shown in Figure 6B, for example, the display control unit 215 displays points PT-1 and PT-2 as star marks, while the other points PT are circular.
[0088] The display configuration shown in Figure 6B is just one example. For instance, the display colors of points PT-1 and PT-2 may be made more visible than the display colors of the other points PT, or the size of points PT-1 and PT-2 may be made larger than the other points PT.
[0089] In this way, the display control unit 215 changes the display mode of the object corresponding to the action based on the importance of the type of action. As a result, analyst B can immediately grasp the object corresponding to the action of the most important type among multiple objects, and efficiently perform behavioral analysis of selected user U[q].
[0090] Figures 6B and 7B-7D illustrate a display method where importance is categorized into two levels: "high" and "low." However, the importance level may be divided into three or more levels, and the display method of the corresponding objects may be changed accordingly. In this case as well, objects corresponding to relatively high importance should be displayed with more emphasis than objects corresponding to relatively low importance.
[0091] [Operation of Behavioral Analysis Device 2] Figure 8 is a flowchart showing an example of the operation of the processing device 21 of the behavioral analysis device 2. The processing device 21 functions as a first receiving unit 211 and receives the specification of target conditions and extraction conditions from analyst B (step S1). The processing device 21 functions as an acquisition unit 212 and transmits a first request including the target conditions and extraction conditions to the data processing server 3 and acquires analysis data ZD from the data processing server 3 (step S2). The processing device 21 stores the acquired analysis data ZD in the storage device 22.
[0092] The processing unit 21 functions as a decision unit 213 and determines the importance of each of the multiple types of behaviors included in the analysis data ZD (step S3). The processing unit 21 functions as a second reception unit 214 and receives the selection of the selected user U[q], who is the subject of the personal analysis, from the analyst B (step S4). The processing unit 21 extracts unit data representing the behavior of the selected user U[q] from the analysis data ZD and creates personal data PD (step S5).
[0093] The processing unit 21 functions as a display control unit 215 and displays personal data PD on the display device 25 (step S6). More specifically, the processing unit 21 displays a timeline TL, etc., based on the personal data PD on the display device 25. The timeline TL, etc., includes one object corresponding to one action. The processing unit 21 changes the display mode of each object based on the importance of the type of action corresponding to each object.
[0094] [Effects of the Embodiment] As described above, the behavioral analysis device 2 according to the embodiment displays objects that correspond one-to-one with the actions performed by the selected user U[q], and changes the display mode of the objects based on the importance of the type of action corresponding to the object. As a result, the analyst B can visually grasp the importance of the actions corresponding to each object and efficiently perform behavioral analysis of the selected user U[q].
[0095] Furthermore, when determining the importance of different types of behavior, the behavioral analysis device 2 uses analysis data ZD, which includes logs from users U other than the selected user U[q]. Therefore, the characteristics of the population to which the selected user U[q] belongs can be reflected in the importance. For this reason, the importance can be set more appropriately compared to, for example, using only the logs of the selected user U[q] or determining importance using a machine learning model trained on a large amount of training data.
[0096] Furthermore, the behavioral analysis device 2 assigns a higher importance to the first type of behavior in the analysis data ZD if a large number of users U perform the first type of behavior. This makes it possible to identify major behaviors performed by many users U as behaviors with high importance.
[0097] Furthermore, the behavioral analysis device 2 assigns a higher importance to the second type of behavior in the analysis data ZD as the number of corresponding unit data points increases. This makes it possible to identify major behaviors that occur frequently as highly important behaviors.
[0098] Furthermore, the behavioral analysis device 2 assigns a higher importance to the third type of behavior when the number of users U who performed the third type of behavior is small, and when the number of times users U who performed the third type of behavior performed the third type of behavior is high. This makes it possible to identify behaviors that are repeatedly performed by a small number of users U as highly important behaviors.
[0099] Furthermore, the behavioral analysis device 2 displays the first object corresponding to a high-importance type with greater emphasis than the second object corresponding to a low-importance type. Therefore, important types of behavior that require particular attention are highlighted, reducing the cognitive load on analyst B during behavioral analysis and lowering the analysis cost (time required) in individual analysis.
[0100] [Modifications] The following are examples of specific modifications that may be added to each of the embodiments exemplified above. Two or more embodiments may be arbitrarily selected from the following examples and combined as appropriate, as long as they do not contradict each other.
[0101] [Modification 1] In the embodiment described above, the personal data PD used for personal analysis was created from analysis data ZD which includes logs from multiple services. However, the personal data PD may also be created by extracting unit data of the selected user U[q] from the log of one service (e.g., service data SD1 or intermediate data XD1). In this case, the personal data PD will show the history of the selected user U[q]'s actions in one service. The importance of the types of actions is determined using the log of one service (e.g., service data SD1 or intermediate data XD1).
[0102] [Modification 2] In the embodiment described above, analyst B used the behavior analysis device 2 to analyze user U's behavior, but the disclosure is not limited thereto. For example, the functions of the behavior analysis device 2 may be performed by the data processing server 3. In this case, the data processing server 3 is an example of a behavior analysis device.
[0103] [Other] (1) In the embodiments described above, ROM and RAM were given as examples of storage devices, but the storage device may be a flexible disk, magneto-optical disk (e.g., compact disk, digital multipurpose disk, Blu-ray® disk), smart card, flash memory device (e.g., card, stick, key drive), CD-ROM (Compact Disc-ROM), register, removable disk, hard disk, floppy® disk, magnetic strip, database, server, or other suitable storage medium.
[0104] (2) In each of the embodiments described above, the information, signals, etc. may be represented using any of the various different techniques. For example, the data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be mentioned throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.
[0105] (3) In each of the embodiments described above, the input and output information may be stored in a specific location (e.g., memory) or managed using a management table. The input and output information may be overwritten, updated, or appended to. The output information may be deleted. The input information may be transmitted to other devices.
[0106] (4) In each of the embodiments described above, the determination may be made by a value represented by one bit (0 or 1), by a Boolean value (true or false), or by a numerical comparison (for example, a comparison with a predetermined value).
[0107] (5) The processing procedures, sequences, flowcharts, etc., exemplified in each of the embodiments described above may be rearranged in order, as long as there is no inconsistency. For example, in the methods described herein, various step elements are presented using an exemplary order and are not limited to the specific order presented.
[0108] (6) Each function illustrated in the diagrams referenced in each embodiment is realized by any combination of at least one of hardware and software. Furthermore, the method of realizing each function block is not particularly limited. That is, each function block may be realized using one device that is physically or logically coupled, or it may be realized using two or more physically or logically separated devices that are directly or indirectly connected (for example, using wired, wireless, etc.). A function block may be realized by combining the one or more devices with software.
[0109] (7) The programs illustrated in each of the embodiments described above should be broadly interpreted to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, execution threads, procedures, functions, etc., regardless of whether they are called software, firmware, middleware, microcode, hardware description languages or by other names.
[0110] Furthermore, software, instructions, information, etc., may be transmitted and received via a transmission medium. For example, if software is transmitted from a website, server, or other remote source using at least one of wired technology (such as coaxial cable, fiber optic cable, twisted pair, or digital subscriber line) and wireless technology (such as infrared or microwave), then at least one of these wired and wireless technologies is included in the definition of a transmission medium.
[0111] (8) In each of the above-mentioned forms, the terms “system” and “network” shall be used interchangeably.
[0112] (9) The information, parameters, etc. described in this disclosure may be expressed using absolute values, relative values from a given value, or other corresponding information.
[0113] (10) In the embodiments described above, the portable device may be a Mobile Station (MS). A Mobile Station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or several other appropriate terms. In this disclosure, terms such as “mobile station,” “user terminal,” “user equipment (UE),” and “terminal” may be used interchangeably.
[0114] (11) In the embodiments described above, the terms “connected,” “coupled,” or any variation thereof means any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are “connected” or “coupled” with each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, “connection” may be read as “access.” As used in the present disclosure, two elements may be considered to be “connected” or “coupled” with each other using at least one of one or more wires, cables, and printed electrical connections, and, in some non-limiting and non-exclusive examples, electromagnetic energy having wavelengths in the radio frequency domain, microwave domain, and optical (both visible and invisible) domain.
[0115] (12) In the embodiments described above, the phrase "based on" does not mean "based solely on" unless otherwise specified. In other words, the phrase "based on" means both "based solely on" and "based at least on".
[0116] (13) The terms “determinating” and “deciding” as used in this disclosure may encompass a wide variety of actions. “Determinating” and “deciding” may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, search, inquiry (for example, searching in a table, database or other data structure), and confirming. Furthermore, "judgment" and "decision" may include considering something as a "judgment" or "decision" based on actions such as receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, and access (e.g., accessing data in memory). Additionally, "judgment" and "decision" may include considering something as a "judgment" or "decision" based on actions such as resolving, selecting, choosing, establishing, and comparing. In short, "judgment" and "decision" may include considering something as a "judgment" or "decision" based on some action. Furthermore, "judgment (decision)" may be reinterpreted as "assuming," "expecting," or "considering."
[0117] (14) Where the terms “include,” “including,” and variations thereof are used in the embodiments described above, these terms are intended to be inclusive, as is the term “comprising.” Furthermore, the term “or” as used in this disclosure is not intended to be exclusive OR.
[0118] (15) In the present disclosure, if articles are added by translation, for example, a, an, and the in English, the present disclosure may include the fact that the noun following these articles is plural.
[0119] (16) In this disclosure, the term “A and B are different” may mean “A and B are different from each other.” The term may also mean “A and B are each different from C.” Terms such as “separate” and “combine” may be interpreted in the same way as “different.”
[0120] (17) Each aspect / embodiment described herein may be used individually, in combination, or switched between as needed during implementation. Furthermore, notification of certain information (e.g., notification that "it is X") is not limited to explicit notification, but may also be implicit (e.g., by not providing such notification).
[0121] 1...Terminal device, 2...Behavioral analysis device, 3...Data processing server, 21...Processing device, 22...Storage device, 23...Communication device, 24...Input device, 25...Display device, 100...Data processing system, 211...First reception unit, 212...Acquisition unit, 213...Decision unit, 214...Second reception unit, 215...Display control unit, B...Analyst, NW...Network, U...User.
Claims
1. An action analysis device comprising: an acquisition unit that acquires log data recording multiple actions performed by each of multiple users; the log data includes multiple unit data that correspond one-to-one with the multiple actions performed by each of the multiple users; the multiple actions are classified into multiple types; each of the multiple unit data includes identification information of the user who performed the corresponding action and information that identifies the type of the corresponding action; a determination unit that determines the importance of each of the multiple types based on the log data; and a display control unit that displays an object on a display device based on at least one unit data indicating an action performed by a user selected from among the multiple users; the display control unit changes the display mode of the object based on the determined importance.
2. The behavioral analysis device according to claim 1, wherein the plurality of types includes a first type, and the determination unit determines the importance of the first type based on the number of users among the plurality of users who performed the first type of behavior.
3. The behavioral analysis device according to claim 2, wherein the decision unit increases the importance of the first type of behavior as the number of users who performed the first type of behavior increases.
4. The behavioral analysis device according to claim 1, wherein the plurality of types includes a second type, and the determination unit determines the importance of the second type based on the number of unit data corresponding to the second type included in the log data.
5. The behavioral analysis apparatus according to claim 4, wherein the determination unit increases the importance of the second type as the number of unit data corresponding to the second type increases.
6. The behavioral analysis device according to claim 1, wherein the plurality of types includes a third type, and the determination unit determines the importance of the third type based on the number of unit data corresponding to the third type included in the log data and the number of users among the plurality of users who performed the third type of action.
7. The behavioral analysis device according to claim 6, wherein the determination unit increases the importance of the third type of behavior as the number of users who performed the third type of behavior decreases and the number of times the users who performed the third type of behavior performed the third type of behavior increases.
8. The behavioral analysis apparatus according to claim 1, wherein the object is a first object corresponding to the type with high importance, and further comprises a second object corresponding to the type with low importance, and the display control unit displays the first object with more emphasis than the second object.
9. The behavioral analysis apparatus according to claim 8, wherein the display control unit displays the first object using chromatic colors and the second object using achromatic colors.
10. A computer-based behavioral analysis method comprising: acquiring log data recording multiple actions performed by each of multiple users; the log data including multiple unit data corresponding one-to-one with the multiple actions performed by each of the multiple users; the multiple actions being classified into multiple types; each of the multiple unit data including identification information of the user who performed the corresponding action and information identifying the type of the corresponding action; determining the importance of each of the multiple types based on the log data; and displaying an object on a display device based on at least one unit data indicating an action performed by a user selected from among the multiple users; and changing the display mode of the object based on the determined importance.
