JP5481242B2 - User characteristics and usage trend analysis system, and processing method and program thereof - Google Patents

Description

  The present invention relates to a user characteristic and usage trend analysis system that analyzes usage trends of users who use screen information such as web pages, and a processing method and program thereof.

A content provider who distributes content such as a web page to a terminal connected to the Internet achieves the purpose by using the content among the users who use the content (for example, if a product sales website is used, purchase a product) Analyzes are made on the interest in the content for each of the users who have completed the goal and those who have not. For example, in electronic commerce, the customer's interest is analyzed from user attribute information and purchase history, and effective direct marketing is performed. However, only purchase results remain in the purchase history, and the background to the purchase is unknown. In addition, the trend of customers who did not make purchases is unknown.
Here, user operation history, user attributes (user name, age, gender, address, occupation, family structure, hobbies, etc. registered in advance when using the system), content information used by the user (genre, content, etc.) Based on the above, a technique for analyzing user trends using a data mining technique is disclosed.

JP 2002-132800 A

  However, in the above-described patent document technology, the level of detail of user operation history data used for analyzing user trends is limited, and in various features such as ability features, preference features, and lifestyle features It is impossible to analyze in detail whether a user having such characteristics uses screen information, and it is impossible to obtain a detailed user trend by analyzing a user trend based on the feature.

  Therefore, the present invention provides a user feature and usage trend analysis system capable of analyzing a trend of a user group in which those features match based on a feature of a user who uses screen information, and a trend of a user, and a processing method thereof And aims to provide a program.

In order to achieve the above object, the present invention provides an operation history receiving unit that receives operation history information for each object constituting screen information, a type of the object indicated by the operation history information, or an operation time of the object, Alternatively, the user characteristic of the screen information is determined based on at least one of the number of operations of the object or the operation type for the object and a user characteristic determination algorithm for determining a characteristic of the user who uses the screen information. User trend analysis for identifying each user's trend analysis information using at least a user feature determination unit, a type of the object to be operated indicated by the operation history information of each user, and a user trend analysis algorithm And the user trend analysis unit is provided for each user. Using the operation history information, the final operation object is identified among the objects to be operated for the predetermined type of screen information, and the final operation is performed with respect to the type of the screen information indicating the leaving part constituted by the final operation object. It is a user feature and usage trend analysis system characterized in that trend analysis information is specified by calculating the appearance ratio of screen information indicating the departure location with respect to the total number of objects .

Further, the present invention is the above-described user feature and usage trend analysis system, wherein the user trend analysis unit operates in the same session based on identification information of the object to be operated indicated by the operation history information of each user. The identification information related to the predetermined object to be output in the screen information is specified by identifying the objects that have been identified, and all of the identification information indicated by the predetermined number of objects operated last among the objects, When the difference between the operation time of the object operated at the earliest time of the predetermined number of objects operated last and the operation time of the object last operated is equal to or greater than a predetermined time, the cause of withdrawal is determined. The trend analysis information indicating the object is specified by determining that the object is the object .

Further, the present invention is the above-described user feature and usage trend analysis system, wherein the user trend analysis unit is configured to perform predetermined processing in the same session based on identification information of the object to be operated indicated by the operation history information of each user. If the identification information of the object that appears more than once is specified, and there are multiple pieces of identification information for the identified object, and each of these objects constitutes different screen information, the screen analysis information indicating the confusing trend location is displayed as trend analysis information. It is characterized by specifying as .

The user feature and usage trend analysis system may further include a user storage unit that stores the determined user feature for each user, and the user trend analysis unit includes the determined user feature. the reading from the user storage unit, wherein the user characteristic to identify the trend analysis information regarding a plurality of users in common.

The user feature and usage trend analysis system may further include a user storage unit that stores the user attribute information for each user, and the user trend analysis unit stores the user attribute information as the user attribute information. read from the user storage unit, wherein the attribute information for specifying the trend analysis information regarding a plurality of users in common.

The user feature and usage trend analysis system may further include a user storage unit that stores input content information input by the user for each user, and the user trend analysis unit includes the input content information. the reading from the user storage unit, wherein the input content information to identify the trend analysis information regarding a plurality of users in common.

The present invention is also a processing method in a user feature and usage trend analysis system, wherein the operation history receiving unit of the user feature and usage trend analysis system receives operation history information for each object constituting screen information. The user feature determination unit of the user feature and usage trend analysis system includes at least the type of the object indicated by the operation history information, the operation time of the object, the number of operations of the object, or the operation type for the object. Based on any one and a user feature determination algorithm for determining a feature of a user who uses the screen information, the user feature of the screen information is determined, and a user trend analysis of the user feature and usage trend analysis system Section of the operation target indicated by the operation history information of each of the users In identifying the trend analysis information of each of the users using at least the object type and the user trend analysis algorithm, the object to be operated on the screen information of a predetermined type using the operation history information of each of the users The final operation object is identified, and the screen information type indicating the leaving location constituted by the final operation object is used to calculate the trend by calculating the appearance ratio of the screen information indicating the leaving location with respect to the total number of the final operation objects. It is a processing method characterized by specifying analysis information .

In the processing method described above, the user trend analysis unit may identify an object operated in the same session based on identification information of the object to be operated indicated by the operation history information of each user. In the case where all of the identification information indicated by the predetermined number of objects operated last among the objects is identification information related to the predetermined object output by the screen information, the last operated When the difference between the operation time of the object operated at the earliest time of the predetermined number of objects and the operation time of the object operated last is equal to or longer than the predetermined time, the cause of the separation is the target object It is characterized by identifying the trend analysis information indicating the object by determining .

According to the present invention, in the above-described processing method, the user trend analysis unit is configured to detect an object that appears more than a predetermined number of times in the same session based on identification information of the object to be operated indicated by the operation history information of each user. When identification information is specified, and when there are multiple pieces of identification information for the specified object and the objects constitute different screen information, each screen information indicating the lost trend location is specified as trend analysis information. And

The present invention, in the processing method described above, the user trend analysis unit, the characteristics of the user who the determined read from the user storage unit, said user characteristic to identify the trend analysis information regarding a plurality of users in common It is characterized by.

The present invention, in the processing method described above, the user trend analysis unit, the attribute information of the user read from the user storage unit, to identify the trend analysis information regarding a plurality of users, wherein the attribute information is common Features.
The present invention, in the processing method described above, the user trend analysis unit, the input contents information read from the user storage unit, to identify the trend analysis information regarding a plurality of users, wherein the input content information is common Features.

Further, the present invention provides a computer of a user feature and usage trend analysis system, an operation history receiving means for receiving operation history information for each object constituting screen information, the type of the object indicated by the operation history information, or the object Based on at least one of the operation time, the number of operations of the object, or the operation type for the object, and a user feature determination algorithm for determining a feature of the user who uses the screen information. User feature determination means for determining a user feature , identifying the trend analysis information of each user using at least the type of the object to be operated indicated by the operation history information of each user and a user trend analysis algorithm Before the operation of each user The total number of the final operation objects with respect to the type of the screen information that identifies the final operation object among the objects to be operated with respect to the screen information of the predetermined type using the information, and indicates the leaving part constituted by the final operation object It is a program characterized by functioning as user trend analysis means for identifying trend analysis information by calculating the appearance ratio of screen information indicating the departure location with respect to .

Further, in the present invention , the user trend analysis means identifies an object operated in the same session based on identification information of the object to be operated indicated by the operation history information of each user, and among the objects, All of the identification information indicated by the predetermined number of objects operated on is identification information related to the predetermined object output by the screen information, and the predetermined number of objects operated last If the difference between the operation time of the object operated at the earliest time and the operation time of the object operated last is equal to or greater than a predetermined time, the target is determined by determining that the cause of the withdrawal is the object. It is a program characterized by specifying trend analysis information indicating an object .

In the present invention , the user trend analysis means identifies identification information of an object that appears more than a predetermined number of times in the same session based on identification information of the object to be operated indicated by the operation history information of each user. When there are a plurality of pieces of identification information of the identified objects and the screens are different from each other, the program is characterized by identifying each piece of screen information indicating the lost trend location as trend analysis information.

According to the present invention, the “user characteristics” can be automatically estimated based on the user operation of the web page, so that information about the user can be efficiently collected, and their accumulation and statistics It can be easily used as data.
In addition, according to the above-described processing, it is possible to estimate “user characteristics” that are more detailed than human estimation, and by obtaining detailed user trends that cannot be obtained in the past, it is possible to improve accuracy of web page usability evaluation and direct marketing. Effectiveness can be improved.
In addition, according to the above-described processing, a detailed “user feature” can be estimated more than human estimation, and the detailed “user feature” or “user profile information” including the detailed “user feature” can be used for general data mining. Content analysis is performed by a technique, and thereby, a detailed analysis result can be obtained as compared with the conventional analysis technique.
From the above, it is possible to create more targeted content (contents suitable for the user, providing method, screen design, etc.) and more effective direct marketing.

It is a block diagram which shows the structure of the analysis system of a user characteristic and usage trend. It is a figure which shows the functional block of each apparatus which comprises the analysis system of a user characteristic and usage trend. It is a 1st figure which shows the processing flow of the analysis system of a user characteristic and a usage trend. It is a 2nd figure which shows the processing flow of the analysis system of a user characteristic and a usage trend. It is a figure which shows the example of user characteristic determination rule data. It is a figure which shows the determination outline | summary of a user's price sensitivity. It is a figure which shows the determination outline | summary of a user's IT skill. It is a figure which shows the determination outline | summary of a user's contract interest degree. It is a figure which shows the determination outline | summary of a user's product knowledge degree. It is a figure which shows the web page for the determination of a life pattern. It is a figure which shows the determination outline | summary of a user's life pattern. It is a figure which shows the determination outline | summary of a user's communication flexibility. It is a figure which shows the determination outline | summary of a user's audio | voice dependence degree. It is a 1st figure which shows the processing flow of a user trend analysis process. It is a 2nd figure which shows the processing flow of a user trend analysis process. It is a 3rd figure which shows the processing flow of a user trend analysis process. It is a figure which shows the analysis result by a decision tree algorithm. It is a figure which shows the analysis result by a basket analysis algorithm.

Hereinafter, a user characteristic and usage trend analysis system according to an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of a user feature and usage trend analysis system according to the embodiment.
In this figure, reference numeral 1 denotes a web server that distributes contents such as web pages (screen information) on the Internet. Reference numeral 2 denotes an operation history system for acquiring an operation history on a web page of a user who browses / uses the web page. Reference numeral 3 denotes a profiling system that performs processing for detecting input content information input to a web page by the user and extracting user characteristics based on an operation history of the web page. Reference numeral 4 denotes a web application analysis system (hereinafter referred to as a WebAP analysis system) that analyzes a user's trend of using a web page based on user operation history, information input to the web page, and user characteristics. is there.

  Reference numeral 5 is an operation history database device (hereinafter referred to as an operation history DB device) that records operation history information of web pages for each user in a database provided in the own device, and 6 is input content information input to the web page by the user. An input content database device (hereinafter referred to as an input content DB device), which is recorded for each user in a database provided in its own device, 7 is input content information registered by the profiling system 3, user feature information, and input from a web server It is a user profile database device (hereinafter referred to as a user profile DB device; user storage unit) that records user attribute information for each user in a database provided in the own device. Reference numeral 8 denotes a terminal used by the user to display a web page.

FIG. 2 is a diagram showing functional blocks of each device constituting the system for analyzing user characteristics and usage trends.
As shown in this figure, the operation history acquisition system 2 includes an operation history receiving unit 21 that receives operation history information from the web browser of the terminal 8.
The operation history DB device 5 acquires operation history information from the operation history acquisition system 2 and records the operation history information in the operation history DB included in the own device, and outputs the operation history information in response to a request from the profiling system 3. The cooperation part 51 is provided.
The profiling system 3 also includes a user feature determination unit 31 that receives operation history information and input content information and estimates a user feature based on the information and user feature determination rule data.
In addition, the user profile DB device 7 stores the user characteristic information determined by the profiling system 3, the input content information received from the profiling system 3, and the terminal 8 received from the web server 1 into the user profile DB included in the own device. A user profile DB cooperation unit 71 that records attribute information of a user to be used and transmits information on the user characteristics and input content information based on a request from the WebAP analysis system 4 is provided.
The WebAP analysis system 4 also analyzes the user trend based on the user feature information acquired from the user profile DB device 7, the input content information, and the operation history information acquired from the operation history DB device 5. Is provided.
The terminal 8 includes an operation history acquisition unit 81 by executing a web browser program. A program for executing the function of the operation history acquisition unit 81 may be included in the content, and the terminal 8 may be provided with the function of the operation history acquisition unit 81 by executing this program.
The user feature and usage trend analysis system according to this embodiment includes at least an operation history acquisition system 2, a profiling system 3, a WebAP analysis system 4, an operation history DB device 5, an input content DB device 6, and a user profile DB device 7. Are connected via a communication network. In addition to these systems and devices, the definition of the analysis system for user characteristics and usage trends may include a configuration that further includes the web server 1 or a configuration that further includes the terminal 8.

The user feature and usage trend analysis system according to the present embodiment receives the operation history information for each object constituting the web page, and the type of the object indicated by the operation history information, the operation time of the object, or the A process for determining a user feature of a web page is performed based on at least one of the number of operations of the object or an operation type for the object and a user feature determination algorithm for determining a feature of a user who uses the web page. .
The user characteristic and usage trend analysis system is based on at least one or both of the type of the object to be operated indicated by the operation history information of each user and / or the appearance frequency in the operation history information, and the user trend analysis algorithm. Then, the process for identifying the trend analysis information of each user is performed.
Further, the user trend analysis unit reads the determined user characteristics, and the operation history information of a plurality of users having the same characteristics and the type or operation of the operation target object indicated by each of the plurality of operation history information of each user Based on at least one or both of the appearance frequencies in the history information and the user trend analysis algorithm, processing for identifying trend analysis information regarding a plurality of users is performed.
Through such processing, the user feature and usage trend analysis system analyzes the trend of user groups that match those features and the trend of individual users based on the features of users who use screen information such as web pages.
In the present embodiment, an example in which the screen information is a web page will be described. However, the screen information may be other information, for example, moving image data. That is, the screen information may be any screen information as long as it is configured by an object.

Next, the processing flow of the user feature and usage trend analysis system will be described in order.
FIG. 3 is a first diagram showing a processing flow of the analysis system for user characteristics and usage trends.
First, the terminal 8 accesses the web server 1 by the function of the web browser and receives content such as a web page. Then, the web page is displayed on the display unit of the terminal 8. The operation history acquisition unit 81 of the terminal 8 acquires the operation history information when a web browser operation is performed. For example, the operation history information includes the operation content for each object and the operation time. The object is a button, an input column, a scroll bar, or the like displayed on the web page constituting the web page. The operation content includes button press, input of a character string to the input column, operation start time and end time of the object, scrolling of the display position of the web page using a scroll bar, and the like. Then, the operation history acquisition unit 81 transmits operation history information including an operation history and a terminal ID to the operation history acquisition system 2 at a predetermined timing (for example, when the use of the web browser is stopped or at a predetermined time interval). . When receiving the operation history information (step S101), the operation history receiving unit 21 of the operation history acquisition system 2 transmits the operation history information to the operation history DB device 5 (step S102). Then, the operation history DB cooperation unit 51 of the operation history DB device 5 registers the operation history information in the operation history DB provided in the own device (step S103).

  The web server 1 receives from the terminal 8 input content information such as a character string input in the input column of the web page displayed on the terminal 8 (step S104), and the input content information and the input content information are received. The input terminal ID is transmitted to the input content DB device 6 (step S105). Then, the input content DB cooperation unit 61 registers the input content information in association with the terminal ID of the terminal 8 in the input content DB provided in the own device (step S106). Furthermore, the web server 1 receives the attribute information of the user who uses the web page from the terminal 8 (step S107), and transmits the attribute information to the user profile DB device 7 (step S108). The user profile DB cooperation unit 71 of the user profile DB device 7 registers the received attribute information in the user profile DB of the own device (step S109). Then, the profiling system 3 receives the operation history information, input content information, and attribute information from each DB device (step S110), and when receiving a predetermined timing or an instruction to start processing, The user feature determination process is started (step S111).

FIG. 4 is a second diagram showing a processing flow of the analysis system for user characteristics and usage trends.
Next, details of the determination process of the user characteristics of the web page will be described with reference to FIG.
First, when detecting the start of the user feature determination process (step S201), the user feature determination unit 31 requests the operation history DB device 5 to transmit operation history information (step S202). Then, as a response, operation history information is received from the operation history DB cooperation unit 51 (step S203). The user feature determination unit 31 requests the input content DB device 6 to transmit input content information (step S204). And the input content information is received from the input content DB cooperation part 61 as a response (step S205). Upon receiving the operation history information and the input content information, the user feature determination unit 31 sorts based on the terminal IDs stored in the information, and performs the operation for each user who uses the web page distributed by the web server 1. The history information and the input content information are temporarily recorded in a memory or a storage unit (step S206). Further, the user feature determination unit 31 reads the user feature determination rule data from the storage unit in the own device (step S207). Then, using the operation history information, input content information, and user feature determination rule data, the user feature is determined for each user (step S208).

FIG. 5 is a diagram illustrating an example of user feature determination rule data.
In the present embodiment, there are seven user feature determination rule data, (1) price sensibility (price interest level), (2) IT (Information Technology) skill, (3) contract interest level, and (4) product knowledge. This is data for determining the characteristics of degree, (5) life pattern, (6) communication flexibility, and (7) voice dependence.
More specifically, the user feature determination rule data for determining whether the degree of price sensibility of (1) is high or low is that the user who repeatedly browses the web page related to the charge has a high degree of price sensibility. This is data for specifying that the user is a user. For example, in this rule data, the identification information of the transition button for accessing the fee related web page among the buttons displayed on the access source web page, and the return information from the fee related web page to the access source web page. Transition button identification information, transition button identification information for accessing the fee breakdown details page displayed on the fee-related web page, number of points when each transition button is pressed, price sensibility judgment algorithm, etc. Is stored.

  Further, (2) the user feature determination rule data for determining whether the degree of IT skill is high or low indicates that a user who has a fast input time when inputting characters in the input column is a user with a high degree of IT skill. This is data for identification. For example, the rule data includes identification information of a web page including an input column for determining an IT skill, an IT skill determination algorithm such as calculation of an input time, and an input time for determining that the input time is short. Threshold (determined that the IT skill level is high when the input time is less than the threshold) is stored.

  In addition, (3) the user feature determination rule data for determining whether the degree of contract interest is high or low is that the user who has browsed the web page related to the contract for a long time has entered into the contract interest (contract items such as products and services). Data for specifying that the user has a high degree of interest). For example, in this rule data, the identification information of the web page related to the contract and the viewing time when the web page is focused (the web page is selected with the mouse or the like on the terminal 8) are calculated. An algorithm and a threshold of browsing time for determining that the degree of contract interest is high (when the browsing time is equal to or higher than the threshold, it is determined that the degree of contract interest is high) are stored.

  In addition, (4) the user feature determination rule data for determining whether the degree of product knowledge is high or low is that the user who has a short time to select a purchase target product from the product list and completes the purchase procedure This is data for identifying a user having a high degree of degree. For example, in this rule data, the identification information of the product list web page and the identification information of the product purchase button are purchased from the time when the product list web page is focused until the time when the product purchase button is pressed. The algorithm for calculating the time (time when the product purchase button is pressed-time when the web page of the product list is focused) and the like are stored.

  Further, (5) the user characteristic determination rule data for determining the lifestyle pattern is the day of the week and time of the desired date selected by the user (for example, if the product is purchased, the day of the product delivery desired date and the delivery time). Data for specifying a user's life pattern (weekday type / holiday type, home time, etc.). For example, this rule data stores a desired date information acquisition instruction and a life pattern determination algorithm.

  Further, (6) the user characteristic determination rule data for determining whether the degree of communication flexibility is high or low is based on whether or not the person in charge of the web page management company can be contacted and the communication content selected by the user. This is data for specifying whether the degree of the error is high or low. For example, this rule data stores input field identification information determined to have high communication flexibility, force field identification information determined to have low communication flexibility, and communication flexibility determination algorithm. .

  Also, (7) user feature determination rule data for determining whether the degree of voice dependence is high or low is transmitted from the web server 1 when the voice guidance button constituting the web page is pressed, and the terminal 8 is sent to the speaker. This is data for determining whether the frequency of use of the output voice guidance is high or low. For example, in this rule data, the voice dependence degree that determines whether the voice dependence degree is high or low based on the identification information of the voice guidance button and whether the number of times the voice guidance button is pressed is greater or less than a threshold value. Stores judgment algorithm and the like.

FIG. 6 is a diagram showing an outline of determination of price sensitivity of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of the price sensitivity of the user who uses the web page, the user feature determination unit 31 reads the operation history information of one user (more specifically, the operation history information of one access). , Identification information of the transition button of the access source web page on which the transition button for accessing the fee related web page is displayed, and identification information of the transition button for returning to the access source web page displayed on the fee related web page Is determined to appear in the operation history information more than the number of times indicated by a predetermined threshold within a predetermined time. If the two transition buttons are pressed in the operation history information more than the number of times indicated by a predetermined threshold within a predetermined time, between the access source web page and the access destination web page It can be determined that the browsing of the web page is switched and the web page related to the fee is being browsed in detail. Therefore, the user feature determination unit 31 gives points to the user according to the operation state, and records the number of points in association with the terminal ID used by the user in the price sensitivity determination result table, for example. In addition, points are also given when the type information of the transition button for accessing the detailed page of the fee breakdown displayed on the fee-related web page appears in the operation history information. For example, the price sensitivity determination result table shows the user's The number of points is recorded in association with the terminal ID to be used. And this process is performed with respect to all the users who accessed the web page. And the user characteristic determination part 31 counts the total number of users to whom the point within the said point range was provided for every point range (for example, 0 point-100 points, 101 points-201 points ...). For example, the total number of users given 0-100 points is calculated, the stated number of users given 101-200 points is calculated, and the total number of users for each point range is sequentially calculated in this way, Data that can generate a graph as shown in FIG. 6B is generated.

  Then, it is determined that a user who is given a point equal to or greater than a predetermined point threshold is a user with a high degree of price sensitivity, and for example, the information is associated with the ID of the terminal used by the user, and the price sensitivity determination result table Register with. In addition, it is determined that a user who is given a point less than a predetermined point is a user with a low degree of price sensibility, and that information is registered in the price sensibility determination result table in association with the ID of the terminal used by the user, for example. It may be. These processes are performed based on a user determination algorithm recorded in the user feature determination rule data. When the user feature determination unit 31 receives the instruction, the user feature determination unit 31 generates a graph as shown in FIG. 6B based on the information registered in the price sensibility determination result table and outputs the graph to a monitor or the like. . As for the point assignment algorithm, for example, 100 points are added when one web page related to a fee appears in the operation history information, and 100 points when one transition button for transitioning to the web page of the access source appears in the operation history information. In addition, when one transition button for accessing the detailed page of the fee breakdown appears in the operation history information, a process of giving points such as 150 points is performed.

FIG. 7 is a diagram showing an outline of determination of IT skills of a user who uses a web page.
When the user feature determination unit 31 detects the determination of the IT skill of the user who uses the web page, the user feature determination unit 31 reads the operation history information of a single user and, for example, inputs the IT skill determination from the operation history information. In the contractor information input page where the column is displayed, the time A when the first character of the character string input by the user is input to the input column and the time B when the last character string of the character string is input are read. . Then, the input time of one character is calculated by {(time B−time A) ÷ number of characters included in character string}. And the user characteristic determination part 31 counts the number of users for every range of this input time. For example, a character input time of calculating the number of a ₁ millisecond ~a ₂ ms and 1 character input time of calculating the number of b ₁ millisecond ~b ₂ ms, such input time The number of users for each range is sequentially calculated, and data that can generate a graph as shown in FIG. 7B is generated.

Then, a user having a predetermined input time or more is determined as a user having a high degree of IT skill, and the information is registered in the IT skill determination result table in association with, for example, an ID of a terminal used by the user. Further, a user having a predetermined input time may be determined as a user having a low IT skill level, and the information may be registered in the IT skill determination result table in association with, for example, an ID of a terminal used by the user. These processes are performed based on the IT skill determination algorithm stored in the user feature determination rule data. When receiving the instruction, the user feature determination unit 31 generates a graph as shown in FIG. 7B based on the information registered in the IT skill determination result table and outputs it to a monitor or the like.
In addition, as shown in FIG. 7A, in the case where two input columns of sex and name are designated on the web as input columns for determining the degree of IT skill, a user feature determination unit 31 reads each of the first input time c ₁ of letters and input to the input columns and input time c ₂ of the last character from the operation history information. Then, the user feature determination unit 31 calculates the sum of the input times when the character string of the contractor's name is input to the two input columns of the surname and the surname using the formula of Σ (c ₂ −c ₁ ), The degree of IT skill may be determined by comparison with the threshold of the input time.
In addition, when two input columns of gender and name are designated on the web instead of the input time to the input column, the user feature determination unit 31 sets the time d ₁ when the cursor is focused on each input column. read the time d ₂ the focus is deviated from the operation history information. Then, the user feature determination unit 31 calculates the sum of input times when the character string of the contractor's name is input to the two input columns of the surname and the first name by the calculation formula of Σ (d ₂ −d ₁ ), The degree of IT skill may be determined by comparison with the threshold of the input time.

FIG. 8 is a diagram showing an outline of determination of the contract interest level of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of the contract interest level of the user who uses the web page, the user feature determination unit 31 reads the operation history information of a certain user and, for example, determines the contract interest level from the operation history information. The time e ₁ when the web page recorded in association with the web page identification information is focused and the time e ₂ when the focus is removed are read. Then, the user feature determination unit 31 calculates the sum of the browsing time when the contract interest level determination web page is focused by the calculation formula of Σ (e ₂ −e ₁ ), and calculates the browsing time. Compare with threshold. Then, the user feature determination unit 31 determines that the contract interest level is high if the browsing time is longer than the threshold value, and determines that the contract interest level is high if the browsing time is shorter than the threshold value. Is registered in the contract interest level determination result table in association with the ID of the terminal used by the client. These processes are performed based on the contract interest level determination algorithm stored in the user feature determination rule data. When receiving an instruction, the user feature determination unit 31 generates a graph as shown in FIG. 8B based on the information registered in the contract interest level determination result table and outputs the graph to a monitor or the like. .

FIG. 9 is a diagram showing an outline of determination of a product knowledge level of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of the product knowledge level of the user who uses the web page, the user feature determination unit 31 reads the operation history information of a certain user and uses the operation history information to determine the product knowledge level web. A time recorded in association with identification information of a page (for example, a web page on which a product list is displayed and a web page on which a purchase decision button for any product in the product list is displayed) is read. The time f ₁ recorded in association with the identification information of the web page on which the product list is displayed is the display start time of the web page, and the identification information of the web page on which the purchase decision button is displayed. association is with Remembered time is the time f ₂ that purchase decision button of the web page is pressed (clicked). The user feature determination unit 31 calculates a purchase determination time from when the product list is displayed until the product is purchased by using the formula (f ₂ −f ₁ ), and compares the purchase determination time with a threshold value. Then, the user feature determination unit 31 determines that the product knowledge level is high if the purchase determination time is longer than the threshold value, and determines that the product knowledge level is not high if the purchase determination time is shorter than the threshold value. Are associated with the ID of the terminal used by the user and registered in the product knowledge level determination result table. These processes are performed based on the product knowledge level determination algorithm stored in the user feature determination rule data. When receiving the instruction, the user feature determination unit 31 generates a graph as shown in FIG. 9B based on the information registered in the product knowledge level determination result table and outputs the graph to a monitor or the like. .

FIG. 10 is a diagram showing a web page for determining a lifestyle pattern.
FIG. 11 is a diagram showing an outline of determination of a lifestyle pattern of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of the life pattern of the user who uses the web page, the user feature determination unit 31 reads the operation history information and input content information of a single user, and from the operation history information and input content information, The selected construction day of the week, the construction time zone, or the construction recorded in association with the identification information of the life pattern judgment web page (for example, the construction page selection screen web page) as shown in FIG. Read information on whether there is a day consultation. When the day of the week indicates a weekday (Monday-Friday), the user feature determination unit 31 determines the user who has designated the construction day as a weekday type, and when the day of the week indicates a holiday (Saturday, Sunday, holiday) The user who specified the construction date is determined to be a holiday type. Further, the user feature determination unit 31 determines the time zone of the selected construction, and determines the home time based on the time zone. In addition, the user feature determination unit 31 determines whether the life pattern is indeterminate based on the selected information on whether or not to consult the construction day (if there is no construction day consultation, it is determined that the life pattern is fixed, If there is a daily consultation, it is determined that the lifestyle pattern is indeterminate). And the user characteristic determination part 31 matches the information of those determination results with ID of the terminal which a user uses, for example, and registers it into a lifestyle pattern determination result table. These processes are performed based on the life pattern determination algorithm stored in the user feature determination rule data. When receiving the instruction, the user feature determination unit 31 generates a graph as shown in FIG. 9B based on the information registered in the product knowledge level determination result table and outputs the graph to a monitor or the like. . Note that the user feature determination unit 31 may read the access start time and access end time to the life pattern determination web page from the operation history information, and determine the home time from the time zone.

FIG. 12 is a diagram showing an outline of determination of communication flexibility of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of communication flexibility of a user who uses a web page, the user feature determination unit 31 reads operation history information and input content information of a single user. Then, the user feature determination unit 31 corresponds to the identification information of the web page for communication flexibility determination (for example, the web page of the contact procedure screen) as shown in FIG. 12B from the operation history information and the input content information. The type information of the selected contact means that is attached and recorded is read. Then, the user feature determination unit 31 determines that the communication flexibility is “low” if the contact means is “e-mail”, and determines the communication flexibility is “medium” if the contact means is “phone”. In the case where the item of “any” is selected for the communication means, the communication flexibility is determined as “high”. And the user characteristic determination part 31 matches the information of those determination results with ID of the terminal which a user uses, for example, and registers it into a communication flexibility determination result table. These processes are performed based on the communication flexibility determination algorithm stored in the user feature determination rule data. When the user feature determination unit 31 receives an instruction, the user feature determination unit 31 generates a graph displaying the determination result as shown in FIG. 12C based on the information registered in the communication flexibility determination result table. Output to a monitor.

FIG. 13 is a diagram showing an outline of determination of the voice dependence of a user who uses a web page.
When the user feature determination unit 31 detects the start of determination of the voice dependency of the user who uses the web page, the user feature determination unit 31 reads the operation history information of one user, and shows the operation history information as shown in FIG. The number of clicks of the voice function button on the web page recorded in association with the identification information of the web page for determining communication flexibility (for example, the web page on the contact procedure screen) is detected. This voice function button is a function for the terminal 8 to output the guidance voice of the web page when clicked. The voice guidance data is transmitted from the web server 1 to the terminal 8, and the terminal 8 outputs it from the voice output unit. Then, when the number of times is greater than the threshold, the user feature determination unit 31 determines the voice dependency as “high”, and the determination result and the number of detections of the identification information of the voice function button in the operation history information (click Frequency) is registered in the voice dependence determination result table in association with the ID of the terminal used by the user, for example. These processes are performed based on the voice dependence determination algorithm stored in the user feature determination rule data. When the user feature determination unit 31 receives an instruction, the user feature determination unit 31 displays a voice dependency determination result as shown in FIG. 13C based on information registered in the voice dependency determination result table. Is generated and output to a monitor or the like. This graph shows the number of users for each number of clicks of the voice function button (voice function use count).

  As described above, the user feature determination unit 31 determines the features of each user and stores them in each data table. Then, user feature information including each data table indicating the user feature is output to the user profile DB device 7, and the user profile DB cooperation unit 71 stores the user feature information in the user profile DB. Also, the user feature determination unit 31 outputs the input content information received from the input content DB cooperation unit 61 to the user profile DB device 7, and the user profile DB cooperation unit 71 stores the input content information in the user profile DB. . Note that the user feature determination process performed by the user feature determination unit 31 is not limited to this, and various user feature determination processes may be performed by other methods. Then, user feature information indicating the features of the user who uses the web page is sequentially accumulated in the user profile DB device 7, and then the user trend analysis unit 41 stores the user feature information accumulated in the user profile DB device 7. Is used to analyze the trend of users who use web pages.

  In the user trend analysis process of the present embodiment, (A) a process of determining a web page to be operated on which a trend related to a user's withdrawal (browsing cancellation) is seen (hereinafter referred to as a withdrawal trend determination process), (B) a user Processing for determining an operation target web page in which a trend related to repetitive operations due to the user's ambiguity is observed (hereinafter referred to as a ambiguity trend location determination process), (C) an operation target web in which a trend regarding the cause of the user's withdrawal is observed User trend analysis processing in each case of page determination processing (hereinafter referred to as “leave cause determination processing”) will be described.

(A) Withdrawal Trend Determination Processing FIG. 14 is a first diagram illustrating a processing flow of user trend analysis processing.
The user trend analysis unit 41 of the WebAP analysis system 4 makes a transmission request for user feature information, input content information, and user attribute information to the user profile DB device 7, and operation history information to the operation history DB device 5. Is transmitted (step S301). As a result, the user profile DB device 7 transmits the user characteristic information, input content information, and user attribute information to the WebAP analysis system 4, and the operation history DB device transmits the operation history information to the WebAP analysis system 4. Send. When the user trend analysis unit 41 of the WebAP analysis system 4 receives an instruction to start the departure trend determination process (step S302), the user trend analysis algorithm (user trend analysis) is determined from the user trend analysis rule indicating the departure trend determination process. (Algorithm) is read (step S303), and the following processing is started based on the algorithm. First, the user trend analysis unit 41 reads identification information of a user's final operation object (button or column) from the operation history information (step S304), and sets the web page ID constituted by the operation object as an operation object. It acquires from the data table which shows the correspondence of ID of the web page comprised by it (step S305). Then, the type information (for example, the price related web page, the contract related web page, the product display web page, etc.) indicated by the web page with the web page ID is read (step S306). The user trend analysis unit 41 determines that the user has left when a page other than the analysis target web page is accessed, and detects the last operation object in the analysis target web page.

  Then, the user trend analysis unit 41 generates the appearance ratio of the type information of the web page constituting the final operation object of a user, that is, the web page that is the leaving part (appearance with respect to the number of all other final operation objects in the operation history information). Ratio) is calculated (step S307). The user trend analysis unit 41 associates the type information of the web page and the ratio of the leaving part with the ID of the terminal 8 used by the user as the trend analysis information, and stores it in the storage unit provided in the own device. Recorded in the departure trend determination result table (step S308). When receiving the output request, the user trend analysis unit 41 determines the appearance ratio for each type information of each web page indicating the leaving location (appearance ratio with respect to the number of all other final operation objects in the operation history information), You may make it output for every user.

The user trend analysis unit 41 may perform the leaving trend determination process only for a user having a specific user characteristic. In this case, the user trend analysis unit 41 receives user feature identification information, information on the degree of user feature “high”, and an instruction to start the departure trend determination process. Then, the user trend analysis unit 41 reads the determination result table indicated by the user feature identification information from the user feature data, and shares that the degree of the user feature is high among the users stored in the determination result table. The ID of the terminal 8 used by the user shown is read. Then, based on the ID of the terminal 8, the withdrawal trend determination process for each user is performed in the same manner as described above. As a result, when the user feature is IT skill and the user feature level is “high”, it is possible to perform the withdrawal trend determination process for the user with high IT skill, and the web where the user with high IT skill leaves The type of page can be determined.
Note that the user trend analysis unit 41 uses the ID information of the terminal 8 used by the user having the same user characteristics, instead of performing the leaving trend determination process indicated by the user group having the same characteristics, Based on the IDs of the terminals 8 used by the same user, a departure trend determination process indicated by a user group having the same attribute information may be performed. Further, the user trend analysis unit 41 may perform a detachment trend determination process indicated by a user group having the same input content information based on the ID of the terminal 8 used by the same user whose acquired input content information is the same.

(B) Lost trend location determination processing FIG. 15 is a second diagram illustrating a processing flow of user trend analysis processing.
When the user trend analysis unit 41 of the WebAP analysis system 4 accepts an instruction to start the lost trend location determination process (step S401), the user trend analysis algorithm (user trend analysis) from the user trend analysis rule indicating the lost trend location determination process. (Algorithm) is read (step S402), and the following processing is started based on the algorithm. First, the user trend analysis unit 41 reads identification information of an operation object (button, column, or web page itself) related to a certain user's web page operation in the same session between the web server 1 and the terminal 8 from the operation history information ( Step S403). Then, the user trend analysis unit 41 determines whether or not a plurality of pieces of operation object identification information appearing a predetermined number of times can be detected among the operation object identification information (step S404). Then, when a plurality of pieces of identification information of operation objects that appear more than a predetermined number of times can be detected, the user trend analysis unit 41 determines that the user is repeatedly browsing between web pages constituting the operation objects ( Step S405). For example, when the identification information of the product price web page and the product detail description web page is detected from the operation history information in the same session more than a predetermined number of times, the user repeatedly browses the product price and the product detail description. It can be determined that the purchase of the product is lost. Then, the user trend analysis unit 41 uses the identification information of the plurality of web pages detected from the operation history information in the same session more than a predetermined number of times as the confused trend location information (trend analysis information), and the ID of the terminal 8 used by the user Are recorded in the stray trend location determination result table in the storage unit of the device (step S406).

  In addition, the user trend analysis unit 41 accumulates the lost trend location information regarding a certain user in the lost trend location determination result table, and then detects the type of the web page from the identification information of the web page indicated by the lost trend location. And the classification of the said web page with which a user may be at a loss in browsing of a web page is specified by the classification. For example, a certain user repeatedly browses a predetermined number of times between a web page displaying the price of the product X and a web page displaying the price of the product Y, and displaying the price of the product X; When the user repeatedly browses a web page displaying the price of the product Z more than a predetermined number of times, the user makes a confusing trend (repeated browsing) on the price-related web page regardless of the type of the product. It can be determined that there is a tendency.

In addition, the user trend analysis unit 41 may perform a lost trend location determination process only for a user having a specific user characteristic. For example, the user trend analysis unit 41 receives user feature identification information, information on the degree of user feature “low”, and an instruction to start a confusing trend location determination process. Then, the user trend analysis unit 41 reads the determination result table indicated by the user feature identification information from the user feature data, and among the users stored in the determination result table, the user trend analysis unit 41 uses it. The ID of the terminal 8 is read. Then, based on the ID of the terminal 8, the lost trend location determination process for each user is performed in the same manner as described above. Thereby, when the user feature is IT skill and the degree of user feature is “low”, it is possible to perform the lost trend location determination process for the user with low IT skill, and the user with low IT skill shows the lost trend. The type of the web page at the location can be determined.
Note that the user trend analysis unit 41 obtains separately acquired attribute information instead of performing the confusing trend location determination process indicated by the user group having the same feature based on the ID of the terminal 8 used by the user having the same user feature. Based on the ID of the terminal 8 used by the same user, the confusing trend location determination process indicated by the user group having the same attribute information may be performed. Further, the user trend analysis unit 41 may perform a confusing trend location determination process indicated by a user group having the same input content information based on the ID of the terminal 8 used by the same user with the acquired input content information. .

(C) Withdrawal Cause Determination Processing FIG. 16 is a third diagram illustrating a processing flow of user trend analysis processing.
When the user trend analysis unit 41 of the WebAP analysis system 4 receives an instruction to start the withdrawal cause determination process (step S501), the user trend analysis processing algorithm (user trend analysis algorithm) is extracted from the user trend analysis rule indicating the leave cause determination process. ) (Step S502), and the following processing is started based on the algorithm. First, the user trend analysis unit 41 extracts the last n operation objects in the same session from a certain user's operation history information (step S503). Of the last n operation objects, the same type as the last operation object is continuously operated a predetermined number of times to the end (step S504), and the same type of operation objects Each of them does not appear more than once (there is no repeated browsing) (step S505), and the browsing time indicated by the time from the first operation time of the consecutive operation object of the same type to the operation time of the last operation object. It is determined whether the reference threshold value is exceeded (step S506). And when all are Yes, the user trend analysis part 41 determines with the user having left | separated by the cause specified by the classification of the said web page (step S507). In other words, for example, the user “operation start (time t0)... → function browse button (t1) of product X on the product list page → function detail browse button (t2) → function detail window scroll (t3) → function detail window When the operation is “scroll (t4) → browser / close button (final operation) = leaving (t5)”, the type of all operation objects in the last operation from t2 to t5 is “function of product X”. When the browsing time related to the function of the product X obtained by calculating “t5-t2” is equal to or greater than a predetermined threshold, the operation is ended despite the time spent browsing the function of the product X. I understand that. Therefore, in such a case, the user trend analysis unit 41 determines that the cause of the user leaving is the function of the product X. Then, the user trend analysis unit 41 associates the type information of the operation object determined to be the cause of withdrawal as the trend analysis information with the ID of the terminal 8, and stores it in the withdrawal cause determination result table in the storage unit of the own device. Recording is performed (step S508).

Further, the user trend analysis unit 41 may perform the leaving cause determination process only for a user having a specific user characteristic. For example, the user trend analysis unit 41 receives user feature identification information, information on the user feature degree “high”, and an instruction to start the cause determination process. Then, the user trend analysis unit 41 reads the determination result table indicated by the user feature identification information from the user feature data, and among the users stored in the determination result table, the user trend analysis unit 41 uses it. The ID of the terminal 8 is read. Then, based on the ID of the terminal 8, the user's withdrawal cause determination process indicating the feature is performed in the same manner as described above. As a result, when the user feature is IT skill and the user feature level is “high”, it is possible to perform a leave cause determination process for a user with high IT skill, and the cause of the leave on the web page of the user with high IT skill. Can be determined.
Note that the user trend analysis unit 41 uses the ID information of the terminal 8 used by the user having the same user characteristics, instead of performing the leaving cause determination process indicated by the user group having the same characteristics, Based on the IDs of the terminals 8 used by the same user, it is possible to perform a cause-of-leaving determination process indicated by a user group having the same attribute information. Further, the user trend analysis unit 41 may perform a leaving cause determination process indicated by a user group having the same input content information based on the ID of the terminal 8 used by the same user whose acquired input content information is the same.

<Use of decision tree algorithm>
The user trend analysis unit 41 uses the decision tree algorithm to generate a decision tree graph as shown in FIG. 17 (a diagram showing the analysis result by the decision tree algorithm), and the user trend and preferences regarding the use of the web page. And attributes may be determined. For example, the user trend analysis unit 41 uses a decision tree algorithm based on user characteristic data, operation history information, determination processing results by the user trend analysis unit 41, user attribute information, and input content information. A user who purchased a product using a web page, a user who did not purchase a product, what kind of characteristics it is, what kind of attributes it is a user, what kind of trend it is It is possible to determine what kind of web page operation / selection the user has performed.
That is, the user trend analysis unit 41 uses the user characteristics, the operation history information, the user attribute information, the input content information, and the decision tree algorithm to affect the operation history information, the user characteristics. Then, a decision tree graph of any one of the user attribute information and input content information and any of operation history information, user characteristics, user attribute information, and input content information that does not affect the achievement of the object is generated.
<Use of basket analysis algorithm>
Further, the user trend analysis unit 41 uses a basket analysis algorithm to determine correlations between user trends, preferences, and attributes regarding the use of web pages, as shown in FIG. 18 (a diagram showing an analysis result by the basket analysis algorithm). You may do it. For example, the user trend analysis unit 41 uses a basket analysis algorithm based on user characteristic data, operation history information, determination processing result by the user trend analysis unit 41, user attribute information, and input content information. What kind of features correlates with users who have purchased a product using a web page and users who have not purchased a product correlate with users who show what attributes, and what trends It is possible to determine whether the user is correlated with the user who indicates, what kind of web page operation / selection the user is correlated with, and the like.
That is, the user trend analysis unit 41 uses the user characteristics, the operation history information, the user attribute information, the input content information, and the basket analysis algorithm, and the user characteristics, the operation history information, the user attribute information, A basket analysis result graph showing a correlation with input content information is generated.

As mentioned above, although this embodiment was described, according to the above-mentioned processing, based on user operation of a web page, it can estimate "user's characteristic" automatically, and collects information about a user efficiently. In addition, it can be easily stored and used as statistical data.
In addition, according to the above-described processing, it is possible to estimate “user characteristics” that are more detailed than human estimation, and by obtaining detailed user trends that cannot be obtained in the past, it is possible to improve accuracy of web page usability evaluation and direct marketing. Effectiveness can be improved.
In addition, according to the above-described processing, a detailed “user feature” can be estimated more than human estimation, and the detailed “user feature” or “user profile information” including the detailed “user feature” can be used for general data mining. Content analysis is performed by a technique, and thereby, a detailed analysis result can be obtained as compared with the conventional analysis technique.
From the above, it is possible to create more targeted content (contents suitable for the user, providing method, screen design, etc.) and more effective direct marketing.

  Each of the above systems and devices has a computer system inside. Each process described above is stored in a computer-readable recording medium in the form of a program, and the above process is performed by the computer reading and executing the program. Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

DESCRIPTION OF SYMBOLS 1 ... Web server 2 ... Operation history acquisition system 3 ... Profiling system 4 ... Web application analysis system 5 ... Operation history DB apparatus 6 ... Input content DB apparatus 7 ... User profile DB device 8... Terminal

Claims (15)

An operation history receiving unit for receiving operation history information for each object constituting the screen information;
At least one of the type of the object indicated by the operation history information, the operation time of the object, the number of operations of the object, or the operation type for the object, and the characteristics of the user who uses the screen information are determined. A user feature determination unit that determines a user feature of the screen information based on a user feature determination algorithm;
A user trend analysis unit that identifies trend analysis information of each of the users using at least a type of the object to be operated indicated by the operation history information of each of the users and a user trend analysis algorithm;
The user trend analysis unit
Using the operation history information of each user, identify the final operation object among the objects to be operated for the predetermined type of screen information, and regarding the type of the screen information indicating the leaving part constituted by the final operation object, A system for analyzing user characteristics and usage trends, wherein trend analysis information is specified by calculating an appearance ratio of screen information indicating the departure location with respect to the total number of final operation objects . The user trend analysis unit
Based on the identification information of the object to be operated indicated by the operation history information of each user, the object operated in the same session is specified, and the identification of the predetermined number of objects operated last among these objects The operation of the object operated at the earliest time among the predetermined number of objects operated last, when all of the information is identification information related to the predetermined object output by the screen information When the difference between the time and the operation time of the last operated object is equal to or longer than a predetermined time, the trend analysis information indicating the target is specified by determining that the cause of withdrawal is the target The user feature and usage trend analysis system according to claim 1. The user trend analysis unit
Based on the identification information of the object to be operated indicated by the operation history information of each user, identification information of an object that appears more than a predetermined number of times in the same session is identified, and there are a plurality of identification information of the identified object. 3. The user characteristics and usage trends according to claim 1 , wherein, when the object constitutes different screen information, each screen information indicating the lost trend location is specified as trend analysis information . 4. Analysis system. A user storage unit that stores the determined user characteristics for each user, and
The user trend analysis unit reads the characteristics of the user who the determination from the user storage unit, according to claim claim 1, wherein the identifying the trend analysis information regarding a plurality of users, wherein the user characteristic is common 4. The user feature and usage trend analysis system according to any one of items 3 to 4 . A user storage unit that stores the attribute information of the user for each user,
The said user trend analysis part reads the said user's attribute information from the said user memory | storage part, The said trend analysis information regarding the some user with the said common attribute information is specified. The Claims 1-3 characterized by the above-mentioned. The user characteristic and usage trend analysis system according to any one of the above. A user storage unit that stores, for each user, input content information input by the user,
The said user trend analysis part reads the said input content information from the said user memory | storage part, The said trend analysis information regarding the some user with which the said input content information is common is specified. The Claims 1-3 characterized by the above-mentioned. The user characteristic and usage trend analysis system according to any one of the above. A processing method in an analysis system for user characteristics and usage trends,
The operation history receiving unit of the user feature and usage trend analysis system receives operation history information for each object constituting the screen information,
The user feature determination unit of the user feature and usage trend analysis system is at least one of the type of the object indicated by the operation history information, the operation time of the object, the number of operations of the object, or the operation type of the object And determining a user feature of the screen information based on one and a user feature determination algorithm for determining a feature of the user using the screen information,
A user trend analysis unit of the user characteristic and usage trend analysis system uses at least the type of the object to be operated indicated by the operation history information of each user and a user trend analysis algorithm, and In identifying trend analysis information, the user specifies the final operation object among the objects to be operated with respect to the screen information of a predetermined type by using the operation history information of each user, and a leaving part constituted by the final operation object The trend analysis information is specified by calculating the appearance ratio of the screen information indicating the leaving part with respect to the total number of the final operation objects . The user trend analysis unit
Based on the identification information of the object to be operated indicated by the operation history information of each user, the object operated in the same session is specified, and the identification of the predetermined number of objects operated last among these objects The operation of the object operated at the earliest time among the predetermined number of objects operated last, when all of the information is identification information related to the predetermined object output by the screen information When the difference between the time and the operation time of the last operated object is equal to or longer than a predetermined time, the trend analysis information indicating the target is specified by determining that the cause of withdrawal is the target The processing method according to claim 7. The user trend analysis unit
Based on the identification information of the object to be operated indicated by the operation history information of each user, identification information of an object that appears more than a predetermined number of times in the same session is identified, and there are a plurality of identification information of the identified object. 9. The processing method according to claim 7 or 8, wherein when the object constitutes different screen information, each screen information indicating the lost trend part is specified as trend analysis information . The user trend analysis unit, the characteristics of the determined user read from the user storage unit, the processing according to claim 9, wherein the identifying the trend analysis information regarding a plurality of users, wherein the user characteristic is common Method. The user trend analysis unit, the attribute information of the user read from the user storage unit, the processing method according to claim 9, wherein the identifying the trend analysis information regarding a plurality of users, wherein the attribute information is common . The user trend analysis unit, the processing method according to claim 9, the input contents information read from the user storage unit, wherein the input content information to identify the trend analysis information regarding a plurality of users in common . Computer of user characteristics and usage trend analysis system,
Operation history receiving means for receiving operation history information for each object constituting screen information;
At least one of the type of the object indicated by the operation history information, the operation time of the object, the number of operations of the object, or the operation type for the object, and the characteristics of the user who uses the screen information are determined. User feature determination means for determining a user feature of the screen information based on a user feature determination algorithm ;
In identifying the trend analysis information of each user, at least using the type of the object to be operated indicated by the operation history information of each user and the user trend analysis algorithm, the operation history information of each user is determined. The final operation object among the objects to be operated with respect to the screen information of a predetermined type is specified, and the type of the screen information indicating the leaving part configured by the final operation object is the total of the final operation objects. A program that functions as a user trend analysis unit that identifies trend analysis information by calculating an appearance ratio of screen information indicating a leaving location . The user trend analysis means includes:
Based on the identification information of the object to be operated indicated by the operation history information of each user, the object operated in the same session is specified, and the identification of the predetermined number of objects operated last among these objects The operation of the object operated at the earliest time among the predetermined number of objects operated last, when all of the information is identification information related to the predetermined object output by the screen information When the difference between the time and the operation time of the last operated object is equal to or longer than a predetermined time, the trend analysis information indicating the target is specified by determining that the cause of the departure is the target
The program according to claim 13. The user trend analysis means includes:
Based on the identification information of the object to be operated indicated by the operation history information of each user, identification information of an object that appears more than a predetermined number of times in the same session is identified, and there are a plurality of identification information of the identified object. When the screen information is different for each object, the screen information indicating those lost trends is identified as trend analysis information.
The program according to claim 13 or claim 14, characterized in that:

Images