JP5838888B2 - User information management apparatus and user information management method - Google Patents

Description

  The present invention relates to a user information management device and a user information management method.

  As a device that manages user information, smartphones that store outgoing and incoming calls, email transmission and reception, and camera shooting as life logs, pad terminals that store document creation logs and web access logs, and personal computers There are computers (hereinafter referred to as client terminals).

  In addition, web search and access logs are stored in a server, and information related to the information is automatically displayed on a client terminal.

Patent No. 3629438 JP 2011-90687 A

  In any of these user information management apparatuses, the user voluntarily inputs or acquires information and manages the information.

  These user information management devices manage information input or acquired by the user on the client terminal, and cannot easily collect various information obtained by the user in the real world.

  Further, these user information management devices handle all information input or acquired by the user equally, and are not managed based on the user's interests, the amount of contact with the information, or the amount of contribution.

  Therefore, an object of the present invention is to provide a user information management apparatus that manages information collected in the real world based on user interests and the like.

A first aspect of the user collection information management apparatus is a user collection information management apparatus having a client terminal and a user information management server capable of communicating with the client terminal via a network,
The user information management server
A user gaze interval determination unit that inputs and accumulates unit interval data having real world collection information collected in the real world in a unit interval of time, and determines whether or not the unit interval data is information that the user gazes;
Each word collection importance calculator for calculating the collection importance of words included in the user gaze interval data;
In addition to the user word group to be displayed according to the collection importance of the words included in the user gaze interval data, a display processing unit that displays words of the user word group on the display unit of the client terminal ,
The client terminal is
In response to a user operation that associates a word of the user word group displayed on the display unit with a certain interest, the interest management unit associates the user operated word with the interest.

  According to the first aspect, user convenience is improved.

It is a block diagram of the user collection information management system in this Embodiment. It is a figure which shows the example of a display of the panel of client terminal Tc1. It is a figure which shows the example of the object at the time of linking the information produced on client terminal Tc1, and putting it together as an interest. It is a figure which shows the example of a display of the panel of client terminal Tc1. It is a graph which shows the example of a change of the information and interest in a fixed period. It is a figure which shows the flow of collection of real world information. It is a figure which shows the rough flow from collection | recovery of real world information until it displays on the display panel of a client terminal. It is a detailed block diagram of a user information management server. It is a figure which shows the flow of a process of the real world information by a server. It is a figure which shows the relationship between each word importance degree table 47C, the display priority table 47D, the user word group 45C, and the display panel 20. FIG. It is a block diagram of the user information management apparatus system in 1st Embodiment. It is a flowchart figure which shows an example of a process of a user attention area analogy part. It is a block diagram of the user information management system in 2nd Embodiment. It is a flowchart figure of a process of the area collection importance determination part. It is a flowchart figure of the calculation process of Stable evaluation value. It is a figure which shows the process after adding to each word importance table. It is a block diagram of the user information management system in 3rd Embodiment. FIG. 10 is a flowchart of processing of a section collection importance level determination unit 10C in the present embodiment. It is a flowchart figure of the score USED calculation process of stationary information acquisition. It is a figure which shows the process after adding to each word importance table. It is a figure which shows the example of a panel when a user shares an interest with other users.

[Definition of terms]
In the following embodiments, each term is defined as follows.

  The user information management device refers to all devices that manage user information. Specifically, a mobile phone, a smartphone, a pad terminal, a personal computer, a server, a fax machine, a fixed telephone with a telephone directory function, etc. used by the user can be cited. However, in the following embodiments, a description will be given by taking as an example a system in which a client terminal used by a user and an information management server are realized by dividing the entire processing. A user terminal dedicated to the user displays user information on the panel for the user, collects user information data in the user information management server, and the user information management server processes, manages, and stores the history of the user information.

  Both the server side and the client side may be a case where a plurality of users (may be 0 users) are assigned to one device, or a plurality of systems are constructed on one device. good. Further, for example, one system function may be distributed to a plurality of devices, for example, a part of functions and a part of data may be separated and held on another device. Conversely, functions realized as a plurality of devices may be realized by a single device.

  The panel is a screen displayed on the client terminal, and the content displayed on the panel is access-controlled by the user information management server. Therefore, the panel display is performed on the assumption that user A is confirmed as a user. In a situation where the user A is not confirmed, it is assumed that the panel is presented in a state where it can be presented to all users.

  Collecting means that the user information management system collects information (such as TV audio) from the real world. The system collects real world information without user operation. Since this collected real-world information includes various miscellaneous information, check whether the importance of collection is high based on the information that the user has watched or the past history. Add to the user words displayed on the panel.

  Collected real world information is stored in the real world information database. For example, when analogizing whether or not the collected real world information is gaze information, the real world information database is referred to when calculating the similarity with the storage unit interval data in the real world information database.

  “Use” means use by an application in a client terminal. The words searched on the web or the like by the search application are accumulated in the usage history database. Each word application usage history database is referred to when calculating the collection importance of collected real world information.

  The interest-related information formation phase is the time period when the user is trying to acquire new information as a concern (new interest-related information formation phase) or other time period (continuous interest-related information formation phase) Means. In the case of a new interest-related information formation phase, words that have no past usage history in the collected information increase the importance of collection. In the continuous interest-related information formation phase, the importance of collecting words that have a history of use is increased based on whether or not the words used in the past in the collected information have been used stably in the past. , Increase the importance of collecting words with no usage history.

[Outline of User Information Management System in Embodiment (Introduction)]
FIG. 1 is a configuration diagram of a user information management system in the present embodiment. The client terminal Tc1 and the user information management server Ts can communicate with each other via a network. Further, the client terminal Tc1 has a display panel, on which user information is displayed, associated with interests, and made available by an application program (hereinafter referred to as an application).

  In addition to the client terminal Tc1, the user information management server Ts is connected to a device Tc2 that collects information in the real world so that communication is possible. Real-world information such as audio data obtained from a TV or the like can be obtained without user operation. Entered. The user information management server Ts calculates the collection information processing unit 10 that processes the keywords (words) included in the input real world collection information, and the importance of the keywords (words) included in the real world collection information. Each word processing unit 12, a panel display processing unit 14 for selecting a user word to be displayed based on the importance of the word, an object processing unit 16 for processing an object displayed on the panel, and the like. Details of these processing units will be described later.

  The user language selected by the user information management server Ts is displayed on the display panel of the client terminal Tc1. That is, the user displays the user words obtained in the real world such as a television and recognized as important on the display panel of the client terminal Tc1 without any special operation by the user. The user can use the displayed user word to link (associate) with interests or use it in an application.

  The user gaze information collection device Tc3 collects user gaze information indicating the degree of gaze of the user when the real world information collection device Tc2 collects real world information. Then, the user information management server Ts acquires the user gaze information and stores it inside. Based on the user gaze information, the user information management server Ts determines whether or not the input real-world information of the unit interval time is the information that the user gazes. It calculates the contribution score to the interval collection importance of the degree and the word contained in it, and the word collection importance. The user gaze information collection device Tc3 is, for example, a video camera, and the user gaze information is, for example, moving image data or still image data.

  In the example of FIG. 1, a web browser is installed in the client terminal Tc1, and a web server program is installed in the user information management server Ts. However, the client terminal Tc1 may have both a web browser function and a web server function, and the client terminal may manage user information performed by the server Ts.

[Panel display example of client terminal (1)]
FIG. 2 is a diagram showing a display example of the panel of the client terminal Tc1. The display example in FIG. 2 is an example of a private panel that is access-controlled so that only a certain user A can see it. On the panel 20, the information object 24 such as “Drinking Party” and “Tori Fuji Yokohama Store” linked to the object 22 of interest and the information alone such as “Yokohama” “Fujiya” “Minato Mirai” float on the panel. The object 26 to be displayed is displayed.

  The largest interest object 22 (1) on the right side in FIG. 2 is linked (associated) with five information objects 24, and the contents of each information are directly character strings as shown in FIG. Some objects store data, while others only store user information and reference information to photo files.

  In addition, available application icons 28-34 are displayed on the panel of the client terminal Tc1. It is possible to perform desired processing for these applications using objects on the panel.

  FIG. 3 is a diagram showing an example of an object when information created on the client terminal Tc1 is linked and collected as an interest. In FIG. 3, both the interest and other information are managed as the same object. That is, the object Id = 00001 is an object of interest (22 in FIG. 2), and the object Id = 00002 is a character string object (24 in FIG. 2). In this way, all objects displayed on the panel shown in FIG. 2 are managed by the object management unit in the user information management server Ts.

  The object displayed on the panel is used in cooperation with the application on the client terminal. As an example, by dragging and dropping an object of interest onto the schedule icon button 30, a schedule is registered using an information object managed with Id = 00002 and Id = 00004 in FIG. By dragging and dropping the data on the panel, it can be handled as information or interest on the panel. Similarly, by dragging and dropping information and interest on the panel to the icon button 32 for searching the Web, the Web search is performed, or conversely, the data searched on the Web is handled as information and interest on the panel. In addition, an information object can be dragged and dropped onto the contact icon 28 and registered in the telephone directory.

  In this way, on the panel, in order for the user to view the information explicitly added by the user or the information added by using the application later or to use it for another application, the user can A group of information is created as the object 22 of interest.

  An information object 26 floating on the left side of the panel is, for example, a user language included in real world information. That is, the user can find on the client terminal panel the user words determined to be important among the keywords collected in the real world such as a television, without performing any special operation. The user can link to the object of interest or use it in the application by dragging and dropping the information object floating on these panels.

  Regarding the deletion of this information, when creating information, the display period (expiration date) is determined, the application-specific display period is determined, the default fixed period is used as the expiration date, or a certain timing (the device is restarted). It may be done by deciding to depend on (startup). An information object floating on the panel is an object to be erased, but an information object linked as a matter of interest by the user's operation can be erased only by an operation by the user in principle.

[Summary of collection importance of real-world collection information]
Collected real-world information includes various miscellaneous information. Therefore, among the keywords (words) included in the collection information, these are the words included in the data of the section that the user has watched. In addition, the phase in which information related to the user's interests is formed, applications in the past, etc. Only words that are determined to be important (words having a high degree of importance) are displayed as information objects 26 on the panel of the client terminal based on the fact that they are used in or linked as interests.

  The phase of information formation regarding interests is as follows. Basically, a certain period of time (for example, one day or one week) after a new interest 22 is added on the panel by a user operation, a new interest-related information formation phase for forming information on the new interest. To do. After a new period of interest-related information formation phase has passed, a continuous interest-related information formation phase for forming information on continuous interests is set.

  In the new interest-related information formation phase, the interest added on the panel is treated as a new concern (there may be several new interests at the same time), and the other concerns are treated as continuous concerns.

  In the new interest-related information formation phase, users tend to acquire new information within a certain short period of time. Therefore, it is desirable to add keywords (words) of real world collection information to the user word group displayed on the panel, considering that words that have not been used in the past are more important.

  On the other hand, in the continuous interest-related information formation phase, it is considered that the keywords (words) of real-world collected information include words that are continuously acquired, such as hobbies, and words that are newly acquired. It is done. Therefore, the tendency of information acquisition is calculated from the keywords (words) of the collected real world information and the keywords (words) that have been used in the past are used according to the acquisition tendency. It is desirable to increase the importance of keywords (words) that are not higher or lower.

  Therefore, for example, the user information management server calculates the degree of information acquisition tendency of primary information acquisition tendency and steady information acquisition tendency. If the same word used in the past (in an application, etc.) is included in the real-world collection information, there is a strong tendency to acquire constant information. On the other hand, if the real world collection information does not include words similar to those used in the past, the primary information acquisition tendency is strong.

  The method of calculating the importance of collecting this word will be described in detail later, but in general, in the new interest-related information formation phase, it is assumed that the user has a strong tendency to acquire primary information, and there is a past usage record. It is preferable to consider the words that are not present (words without usage history) as important and add them to the user word group displayed on the panel. On the other hand, in the continuous interest-related information formation phase, the degree of information acquisition tendency is calculated by the above method, and depending on which acquisition tendency is strong, words that have been used (words with usage history) It is preferable to give a contribution score of the importance of collection to new words that have never been used (words without usage history).

[Panel display example of client terminal (2)]
FIG. 4 is a diagram showing a display example of the panel of the client terminal Tc1. The display example of FIG. 4 shows the transition of interest on the panel from the display example of FIG. In FIG. 4, the interest object 23 (3) and the information object 26 (5) that were not on the panel of FIG. 2 appear, while the interest object displayed on the panel is the same as in FIG. 2. 22 (1), 22 (2) and information object 26 (4).

  User A does not add an information object to the interest Id = 00001 (22 (1) in FIG. 2) shown in FIG. 3 between FIG. 2 and FIG. In addition, the information object 26 (4) including the address character string “Yokohama City Kohoku Ward Okurayama…” in the lower left of FIG. 4 is also displayed at the fixed position. User A keeps an interest in the information and interests that are continuously present in the panel, but does not acquire new information about them. Instead, we are trying to get information about three concerns linked to the interest link 22 (3) added at the time between FIG. 2 and FIG.

[An example of information acquisition tendency]
Here, as an example, an example will be described in which the degree of information acquisition tendency is calculated as an average change in interest over a certain period.

  FIG. 5 is a chart showing an example of changes in information and interest during a certain period. Fig. 5 shows the results of checking the changes up to 2:00 AM each day since the previous day's counting and counting the increments. For example, in the calculation on July 1, 2011, since the previous day's calculation, two new interests (primary information acquisition) were linked to a total of 10 pieces of information, and stationary interests (steady information acquisition) ) Is not linked to any information. The three-day incremental average from July 3rd to 5th, 2011 is a total of 4 links to information on primary and stationary concerns. Of these, there are 4 links to primary concerns and 0 links to stationary concerns. In other words, for these three days, the primary information acquisition tendency was 4/4 = 100%, and the steady information acquisition tendency was 0/4 = 0%.

  Such an information acquisition tendency is used when calculating the word collection importance as an index as to whether or not to add to the user words displayed on the panel, as described above.

[Flow of processing real-world collected information]
Next, the outline of the processing flow for information collected in the real world will be described. The information collected in the real world is an enormous amount with various miscellaneous information. Therefore, in the present embodiment, it is checked whether or not the unit interval data is a user's gaze unit, and further, the importance of each keyword (word) in the user gaze unit interval data is checked and displayed on the panel. Add to the group.

  FIG. 6 is a diagram showing a flow of collecting real world information. In this example, the user is sitting in front of the television and watching the television. The real-world information collection device Tc2 such as a smartphone or a voice recorder collects real-world information Rin such as voice data heard from a television. A user gaze information collection device Tc3 such as a video camera collects user gaze information Fin such as moving image data of a user's facial expression. The collected information Fin and Rin are input to the user information management server Ts as input data Vin and processed.

  Specifically, in the example of FIG. 6, the smartphone Tc2 in the user's pocket is a real-world information collection device, and the collected environmental sounds such as TV sounds and user tweets are all real-world information Rin. Transmit to the information management server Ts. The web camera Tc3 provided on the television, as a user gaze information collecting device, displays user gaze information Fin indicating a user's state (smile, etc.) as to whether or not the user is gazes at real world information. Send to Ts. The collected information Rin and Fin are processed by the information management apparatus Ts and processed as information object candidates to be presented on the panel of the user A client terminal Tc1.

  FIG. 7 is a diagram showing a rough flow from collection of real world information to display on the display panel of the client terminal.

  The unit interval data Rin, which is unit time information collected from the real world information collection device Tc2 and the user gaze information Fin collected from the user gaze information collection device Tc3, of the real world information contacted by the user A, As input data Vin that is information of the same section by time matching, it is input to the user information management server Ts and processed. The input data Vin includes keywords (words) included in the real world information Rin.

  In the user information management server Ts, when the user gaze information determination unit S1 adds the user gaze information Fin to the input data Vin after the preprocessing, the time when the user gaze is performed (for example, the user is laughing) It is only necessary to determine the preceding and following sections as user gaze sections. That is, a time zone in which the user is gazing, such as when the user's smile is reflected in the image data of the camera Tc3, is determined as the user gazing interval.

  On the other hand, in situations where the user is out of the shooting range of the user gaze information collection device Tc3, the user gaze information Fin may not be added to the real world information collected from the input data Vin. In this embodiment, even if the user gaze information Fin is not added, the user gaze interval analogy unit S2 stores the accumulated data indicating whether the collected real world information may be of interest to the user. Estimate based on DB1. This analogy method will be described in detail later.

  As a result of this user gaze interval analogy, if it is determined that the user may be interested in the collected real world information Rin, the input data Vin is regarded as the user gaze interval.

  Next, the section collection importance calculation unit S3 calculates the section collection importance of the input data Vin determined as the user gaze section. This section collection importance is the collection importance of the real world collection information in the unit section. If this section collection importance is high, the extraction keyword (word) of the section will also have a high collection importance. It is added to the user word group of the panel display target.

  For example, (1) For each keyword (word) in the input user gaze interval, the word collection importance is calculated based on the number of times collected in the past. (2) Referring to the use history of words used in the past in the past or the like, the section collection importance is determined. The cumulative number and usage history are stored in the history database DB2.

  In the case of the word collection importance determination method of (1) above, when the input data is processed as collection information, in addition to the collected real world information and user gaze information in the accumulated data DB1, The cumulative number of collections is recorded. Then, the importance of the word is determined by the cumulative number of times each word is collected, and the word having the high importance is added to the user word group to be displayed on the panel.

  The cumulative collection count for the panel display is reset when the panel display is performed. When the cumulative number of times of collection is handled as the importance level of each word, each word importance level determination unit (described later) records the cumulative number of times of sampling in each word importance level data table (described later). Each word importance degree data table exists for each application, and depending on the application, the score calculation in each word importance degree table may be designed so as to give more importance to the degree of frequent occurrence and temporal novelty.

  The interval collection importance determination method of (2) will be described in detail later. When the interval collection importance is calculated, the collection importance contribution score of each word contributing to the interval collection importance is evaluated. Accumulated in each word as a value. This collection importance contribution score is each word collection importance.

  Next, the user word group organizing unit S5 adds to the user word group to be displayed on the panel based on the importance of each word (such as the cumulative number of collections, the cumulative value of the collection importance contribution score, etc.).

  Then, the panel display updating unit S6 displays the user word group on the panel and updates the display on the panel. The user language displayed on the panel is floated and displayed as the information object 26 described with reference to FIG. When the user language is used for an application or the like, the usage history is accumulated in the history database DB2.

[Detailed configuration of user information management server and flow to real-world information display]
Before describing the first, second, and third embodiments in detail, a more detailed configuration of the user information management server and how the collected real world information is displayed on the panel will be described. After understanding these descriptions, the first, second, and third embodiments will be described.

  FIG. 8 is a detailed configuration diagram of the user information management server. The server Ts includes a communication unit 40, a user interface unit 41, a preprocessing unit 42 for input data such as extracting keywords from audio data of real world information, and an external cooperation unit 43 for extracting data used in the application. .

  Further, the server Ts has a program as a program, a collected information processing unit 10 that processes information collected from the real world, a word processing unit 12 that processes words included in the collected information, and a user word group on the panel of the client terminal. Panel display processing unit 14 that performs panel display processing such as displaying words as objects, and objects on the panel (words, interests, links between interests and other objects, etc. included in the user word group) And an object processing unit 16 that performs processing of the above.

  The collection information processing unit 10 includes a user gaze section determination unit 10A that determines whether or not unit section data collected from the real world is a user gaze section, similarity to information collected in the past when there is no user gaze information Fin, and the like User attention segment analogy unit 10B (explained in the first embodiment) for analogizing the user gaze interval, and segment collection importance determination unit 10C (explained in the second and third embodiments) for determining the collection importance of the segment ), An appearance opportunity prediction unit 10D (described in the third embodiment).

  Each word processing unit 12 includes a word importance determining unit 12A that determines the importance of each word from the section collection importance, and a display priority determining unit 12B that determines a display priority from the importance of each word.

  The panel display processing unit 14 updates the display priority table according to the importance of each word, and updates the user word group table by adding higher-order words in the display priority table to the user word group. A user word group organizing unit 14B, a panel display updating unit 14C for displaying the words of the user word group in a panel, and a panel usage history managing unit 14D for managing the usage history of the words in the panel to the application or interest.

  The object processing unit 16 includes an object management unit 16A and an object management history management unit 16B.

  The database of the server Ts includes a storage database 44 that stores data collected from the real world, a panel-related database 45, a history database 46 that stores history such as usage on the panel, and an evaluation value database 47 that stores evaluation values. Have

  The accumulation database 44 includes real world information 44A for each unit section collected and preprocessed, gaze real world information 44B determined as a user gaze from among them, user gaze information (Fin) such as a camera, and cumulative collection of each word It has count data 44D. The real world information and the gaze real world information are aggregated for each unit section, and further aggregated for each collection scene such as moving on a train or shopping, or for each user. The gaze real world information 44B is information of a section that the user gazes out of the real world information 44A, and does not necessarily need to be accumulated as accumulated data.

  The panel-related database 45 includes object information 45A, link information 45B to interests, and user word group data 45C.

  The history database 46 includes object management history data 46A and usage history data 46B for each word application. Each word application usage history data 46B is tabulated for each unit section, and further tabulated for each scene.

  FIG. 9 is a diagram showing a flow of processing of real world information by the server. Voice data such as television is collected in the real world and input to the user information management server as real world information Rin, and keywords in the real world information (voice data) Rin are extracted by the input data preprocessing unit 42 and input data. Vin and stored in the real world information 44A in the storage database 44 for each unit section. The data for each unit section of the real world information 44A includes information such as extraction keywords, collection time zones (holidays, weekdays, etc.), collection scenes, unit section data Rin, and users.

  In addition to real world information, user gaze information Fin in a unit section is also collected by a camera or the like and input to the user information management server as input data Vin. The user gaze information Fin for each unit section is accumulated in the user gaze information 44C in the accumulation database 44.

  The user gaze section determination unit 10A determines that the user gaze section is a user gaze section based on the user gaze information Fin, for example, the smile of the user (S1). If the user gaze information Fin is not input, the user gaze interval analogy unit 10B resembles the various information of the collected real world information with the various information in the accumulated real world information data 44A. It is inferred that it is the user gaze interval based on whether or not it is (S1).

  For the keywords (words) of the real world information that the user has watched or guessed, the section collection importance determining unit 10C determines the section collection importance, and the contribution score of each word included in the section collection importance Is determined (S2). The calculation method of the determination of the interval collection importance and the contribution score of each word will be described in detail in the second and third embodiments. For example, the interval collection importance is determined based on the word included in the interval data. It is calculated by the total number of collections in a certain period, or based on the word usage history.

  The information acquisition tendency degree 47D and the word-by-word collection importance contribution score 47B in FIG. 9 will be described in detail later.

  When the importance of each word is determined by the above calculation, it is added to each word importance table 47C. Then, the display priority determination unit 12B updates the display priority table (S5-A) based on the cumulative importance of each word importance table 47C.

  FIG. 10 is a diagram showing the relationship among each word importance table 47C, display priority table 47D, user word group 45C, and display panel 20. As shown in FIG. Each word importance table 47C is classified for each object displayed on the panel, and is classified into words collected from the real world (Real), words used in web search (Web), words used in the schedule application (Sched), etc. being classified. And, for the words collected from the real world, the collection importance is accumulated as the importance. In the example of FIG. 10, it is the cumulative number of times of collection. The importance of words used in the application is calculated based on, for example, usage frequency or novelty.

  The panel display management unit 14A performs display priority table update processing S5-A, and updates the display priority table 47D in descending order of importance in each word importance table 47C for each object. In the display priority table 47D, all objects are mixed and arranged in the display priority order. The timing of the update is arbitrary, for example, when there is a margin in computer processing.

  Returning to FIG. 9, if there is a vacancy in the user word group 45C, the user word group organizing part 14B additionally registers the upper word of the display priority table 47D in the user word group and updates the user word group ( S5-B). The user word group 45C is displayed on the panel. When the user explicitly deletes from the panel or when a certain period of time elapses, it is deleted from the user word group and is not displayed on the panel. However, it is desirable that words that are linked as a concern are not deleted after a certain period of time. However, even if it is deleted from the user word group and is no longer displayed on the panel, it is not deleted from each word importance table 47C. Therefore, the importance may be increased again and added to the user word group and displayed.

  The panel display updating unit 14C displays the words in the user word group 45C on the panel (S6). Thereby, the words in the user word group are displayed in the panel 20 as words 1-4, and the user can use them arbitrarily. Since the displayed words include keywords collected from the real world by the system, the convenience of the user can be improved.

  When the user adds the word displayed on the panel 20 to the link of interest (S12), the panel usage history management unit 14D adds the linked history to the link information 45D.

  When the user uses the displayed user word (S7-A), the panel usage history management unit 14D adds the usage history to each word application usage history data 46B. In addition, when a word input by the user is used in the application (S7-B), the panel usage history management unit 14D adds the usage history of the application usage word 49 to each word application usage history data 46B. Based on the application usage frequency and novelty (S13), it is added to or added to each word importance table 47C.

  In this way, the usage history used in the application is accumulated in each word application usage history data 46B, and the section collection importance level determination unit 10C and each word importance level determination unit 12A refer to the usage history data to be described later. To calculate the importance of each.

  With the above processing flow, the user's attention and importance are evaluated for information automatically collected from the real world by the system, and displayed on the panel according to the importance of each word. Since the information collected from the real world by the system is cobblestone chaos, by appropriately performing these evaluations, the word that the user wants or wants to use as an interest is displayed as an object on the panel.

  The above is the rough configuration and processing flow of the present embodiment. Hereinafter, specific embodiments will be described in detail.

[First embodiment, estimation of user gaze interval]
FIG. 11 is a configuration diagram of the user information management apparatus system according to the first embodiment. Unlike the configuration diagram of FIG. 1, a user gaze section determining unit 10A that determines input data Vin of unit section data from which the collection information processing unit 10 has collected real world information based on user gaze information Fin, and user gaze information Fin are And a user gaze interval analogy unit 10B that performs analogy based on the similarity with the real world information Vi accumulated when it cannot be obtained. Furthermore, the configuration of FIG. 11 differs from FIG. 1 in that each word processing unit 12 determines each word importance determining unit 12A that determines the collection importance of each word and the contribution score of each word to the collection importance. And a display priority determining unit 12B that determines whether or not to display in addition to the user word group. The client terminal Tc4 of another user different from the user A is also connected to the user information management server Ts via the network, and the real world collection information of the other user is accumulated in the server Ts. The rest is the same as FIG. Each of the panel display processing unit 14 and the object processing unit 16 has the configuration shown in FIG.

  In FIG. 11, as in FIG. 1, the client terminal Tc1 may have both a web browser function and a web server function, and the client terminal may manage user information performed by the server Ts. In this case, the client terminal inputs real world information in a stand-alone manner, manages user information, and displays user words.

  In the first embodiment, it is configured to infer whether or not the input data collected by the user gaze interval analogy unit 10B is a user gaze interval. Other than that, it is equivalent to the schematic description of FIG. That is, the user information management server Ts determines whether or not the real-world information Vin collected and input is information on a unit section watched by the user based on the user watching information Fin. , The contribution score to the interval collection importance of the interval collection importance and the words included in the interval collection importance is calculated. On the other hand, if the user is not in the vicinity of the sensor that collects the user gaze information (for example, in front of the camera), the user gaze information cannot be obtained. Even in that case, in the first embodiment, it is estimated (or analogized) based on the accumulated data whether or not the collected real world information may be of interest to the user. When it is estimated that the user gaze interval, the input information is regarded as a user gaze interval, and the user information management server Ts performs a process of calculating a segment collection importance and a contribution score of each word to the interval collection importance.

FIG. 12 is a flowchart illustrating an example of processing of the user gaze interval analogy unit. First, the input data V _in to the user gaze interval analogy unit, for example, a content as follows.
Vin = {Extraction Keyword K _in : Minato Mirai, Fujiya, Yokohama, Taste Log, ...
Collection time: Holiday collector: hoge
Unit interval data Rin: rfile_hoge_20110702120000.wav
User gaze information Fin: None}
There was no user gaze information Fin that was temporally associated with the unit time audio file Rin cut out as a unit section. In addition to the above, the input data Vin may also include information on the scenes collected, such as “moving by train” and “shopping”, in addition to the analysis of GPS information at the time of voice file Rin and Rin collection. good.

  On the other hand, the input real world collection information data to which the user gaze information collected before is added is accumulated as the accumulation database 44 in the user information management terminal Ts. This is the real world information 44A shown in FIG. As shown in FIG. 12, the accumulated data is organized and accumulated for each unit section in a relatively collected form. On the other hand, it is stored in a state that can be organized for each scene and each user.

  The real-world information 44A of the storage database shown in FIG. 12 has data V1, V2,... Organized for each unit section, and each data Vi includes an extraction keyword, a sampling time zone, a sampling scene S1, and unit time data. R1 etc. The same applies to data organized for each scene. In addition, the real world information 44A of this accumulation database is real world information collected by the user who owns the client terminal Tc1 and other plural users, and the real world information collected by other users on the panel. It may be displayed as an object in an understandable way. For example, real-world information collected by other people with close interests.

  The user gaze interval analogy unit 10B compares the input data Vin and the real world information Vi, which is stored data organized for each unit interval, from a plurality of comparison viewpoints (S21). There are three comparative viewpoints illustrated in Fig. 12. The first is the similarity s_xi of the collection time zone of the real world information Vi and the input data Vin, and the second is the extracted keyword for each Vi and Vin collector. The similarity s_yi, and the third is the voice similarity s_zi between the unit interval data Ri and Rin of Vi and Vin. Each of these similarities is normalized to 1 and becomes 1 at the maximum.

  The similarity s_xi between the sampling times of Vi and Vin, which is the similarity on the X axis, is calculated as follows. In this calculation, the sampling time periods where there is a difference in information acquisition, such as the coincidence of holidays or weekdays, the coincidence of time periods such as day and night, the coincidence of time periods at work and time spent at home, etc. Set. In the example of FIG. 12, since V1 and Vin are collected on holidays, the collection time zones coincide with each other, and the similarity s_x1 on the X axis is 1.

  The similarity s_y1 of the extracted keywords for each collector of V1 and Vin, which is the similarity on the Y axis, is calculated as follows. In other words, simply compare the extracted hoge of Vin and the extracted keyword for each user accumulated by the V1 collector (in this case, if the same person is naturally the same, but the collector who is close to the language used) Then, it becomes a value closer to 1). In this example, the extracted keyword is not so much matched and is similar to general match / mismatch, so the similarity is set to 1/2.

  In the calculation of the voice similarity s_z1 between the unit interval data R1 and Rin of V1 and Vin, which are similarities on the Z axis, there is not much similarity in the example of FIG.

  Next, the weight of each comparison viewpoint for the three similarities is determined based on the similarity type (S22, S23). In the example of FIG. 12, the similarity type is a context-dependent type on the X axis, a mass data dependent type on the Y axis, and a non-dependent type on the Z axis.

  That is, the context-dependent type is a type in which the degree of similarity changes depending on the context. For example, if there is a tendency for the degree of similarity related to the sampling time to be acquired only by people on holidays, it is not necessary to consider the similarity of the sampling time on weekdays. Set to 0. Therefore, this weight depends on the user.

  The mass data dependency type is a similarity type that reduces the reliability of the similarity calculation result unless the similarity determination is performed based on a large amount of data. Explaining the extracted keywords for each sampler as an example, if there are only a few extracted keywords for each sampler, the tendency of the extracted keywords for sampler A and collector B is similar or not similar It is difficult to judge. Therefore, if the number of collected keywords is small, the weight is reduced, and if it is large, the weight is increased.

  The non-dependence type is a type that can explain that the result of simply comparing data A and data B is the similarity between data A and data B. In this example, the similarity of audio files is this type. The weights Wx, Wy, and Wz are set in consideration of the type of each similarity. In the example, the X-axis weight Wx is set to 1/3 without adjusting the weight because the user does not depend on the context. Since it is a weight for one of the three similarities, it is simply one third. The Y-axis weight Wy is set to 1/6 because the number of extracted keywords is small and a large amount of data is insufficient. The weight Wz of the Z axis is calculated as 1- (Wx + Wy) and is 1/2.

As a result, the similarity S1 between Vin and V1 is (s_x1 · Wx _, s_y1 · Wy _, s_z1 · Wz) = (0.33, 0.13, 0.05) (S24). Similarly, S2, S3, S4,... Sn are mapped to the similarity space (S20, S21).

  After mapping all similarities Si to be compared to the similarity space, it is determined whether there is a similarity Si exceeding a certain value (threshold) (circle TH in the vector space diagram). In the example of FIG. 12, since S4 exceeds the threshold value among the four similarities, it is inferred as a user gaze interval.

  In the above example, it is determined whether a certain value has been exceeded after mapping all the similarities Si to the similarity space, but at the time of mapping to each similarity, whether a certain value (threshold value) has been exceeded. You may make a judgment. In addition, all the stored Vis may be targeted for verification, but the stored data to be used may be filtered, such as narrowing down to the most recent data.

  In addition, if it cannot be regarded as the user gaze interval even when compared with all verification targets, neighborhood similarity determination S26 may be performed as a remedy. As shown in FIG. 12, in the neighborhood similarity determination S26, another constant value TH2 is set inside a predetermined constant value TH1, and the similarity Si corresponding to a plurality of continuous unit interval data Vi is determined. If the value exceeds this constant value inside, it is determined that it is regarded as a gaze interval.

  The user information management server Ts calculates the importance for each word with respect to the keyword (word) in the input data determined or inferred as the user attention section, and the user is based on the importance for each word. A word group (user word group) to be displayed on the panel is determined. In the first embodiment, the calculation of the degree of importance of each word is determined by the cumulative number of times each word has been explained with reference to FIG.

  Subsequent updating of the display priority table based on the importance of each word, updating of the user word group, display on the panel, link to the interest, usage history in the application, etc. are explained in FIGS. It is the same.

[Second embodiment, interval collection importance and each word collection importance contribution score]
In the second embodiment, for the input unit interval data, the interval collection importance is calculated based on the usage history of the application or the like, and a keyword (word) having a high contribution score to the interval collection importance is given priority. Include in the input when updating the user word group. This is different from the case where the cumulative number of collections in the first embodiment is used as the collection importance of each word. The rest is the same as the first embodiment.

  FIG. 13 is a configuration diagram of a user information management system according to the second embodiment. Unlike the first embodiment of FIG. 11, the user information management server Ts includes a section collection importance degree determination unit 10 </ b> C in the collection information processing unit 10. The user information management server Ts calculates the interval collection importance for the input data that is considered to be the user attention interval, and preferentially selects words included in the unit interval data with high interval collection importance. Add to the group.

  FIG. 14 is a flowchart of processing of the section collection importance level determination unit. In this determination process, the interval collection importance is calculated in consideration of the degree of information acquisition tendency at the time of collection of real world information.

  As preprocessing, refer to each word application usage history data 46B for each word of the extracted keyword Kin of the input data, and depending on the usage history, there is a usage history word Kin_use with usage history and usage history without usage history Sorted into none word Kin_notuse (S30). In the example of FIG. 14, the usage history word Kin_use is “Minato Mirai, Fujiya, Yokohama” among the “Minato Mirai, Fujiya, Yokohama, Tabelog, and other five” extraction keywords Kin of the input data. There are six words Kin_notuse without history, “eating log, five others”. As will be described later, these words with usage history and words without usage history are information acquisition trends (the tendency to acquire new primary information, or constant information such as hobbies that are acquired continuously) The importance of collection varies depending on whether or not

  The interval collection importance for the words in the user's gaze interval is calculated as follows. To summarize, as the number of words with usage history in the interval increases, the steady information acquisition score USED of the collection severity scores increases. The higher the number of words without usage history, the higher the UNUSED score for primary information acquisition among the collection importance scores. These scores are corrected by the weight w1 of the steady information acquisition tendency and the weight w2 of the primary information acquisition tendency.

  The section collection importance determination unit 10C first determines the information formation phase regarding the interest (S31). The information formation phase regarding interests includes a new interest-related information formation phase and a continuous interest-related information formation phase. For example, a predetermined period (for example, 1 day to 1 week) after adding a new interest on the panel of the client terminal is a new interest-related information forming phase with a high tendency to acquire new words. On the other hand, the other period is the continuous interest-related information formation phase.

  If it is determined that the new interest-related information formation phase is established (S31), the weight w1 of the steady information acquisition tendency is set to a low value, for example, 0 so that new information can be acquired as much as possible (S32). As a result, the weight w2 of the primary information acquisition tendency is set to a value higher than w1, for example, w2 = 1−w1 = 1, and the weight for the primary information acquisition score UNUSED becomes the maximum value 1, and the interval collection importance is increased. The contribution score of becomes higher. A method for calculating the contribution score will be described later.

  If it is determined that it is the continuous interest-related information formation phase (S31), the information acquisition tendency weights w2 and w1 are calculated for the current information acquisition tendency (S33). In other words, the continuous interest-related information formation phase is simply a period other than the period of the new interest-related information formation phase, and thus there is a strong tendency for constant information acquisition in the continuous interest-related information formation phase. In some cases, there is a strong tendency to acquire new primary information. Therefore, the word included in the user gaze section is examined closely to calculate which tendency is strong.

  The weight w1 of the steady information acquisition tendency is determined to some extent when the word usage history is the same when the current time point is compared with the past time point. If the words used in the collected real-world information that have a usage history (words with usage history) have not changed much in the past and present, there is a steady information acquisition tendency. If there is a stronger (steady weight of information acquisition tendency w1), and a new word is used more strongly and the words used in the past and the present are significantly different, the primary information acquisition tendency is stronger ( The weight w2 of the primary information acquisition tendency is heavy).

Since these are integrated with each other, in the example of FIG. 14, the weight w2 of the primary information acquisition tendency is set to w2 = 1−w1 (weight w1 of the steady information acquisition tendency). Further, in the example of FIG. 14, a stable evaluation value (the same degree of use history) is performed for each word of the extracted keyword, and a change in the use history is calculated on average when each word is viewed as the center. In step S33 of FIG. 14, the formula for calculating the weight w1 is
w1 = Stable / u = (Addition of stable evaluation of words with usage history) / u
It is said. u is the number (3) of words with usage history. That is, the weight w1 of the steady information acquisition tendency as the collection interval is obtained by the average value of the stable evaluation values of the words having the usage history.

  Here, the description of FIG. 14 is interrupted, and the stable evaluation value of each usage history word is described below.

[Stable1 evaluation]
FIG. 15 is a flowchart of a stable evaluation value calculation process. The calculation of the stable evaluation value is also performed by the section collection importance determining unit 10C.

  In the example of FIG. 14, among the extracted keywords (words) in the user attention interval for which the collection importance is to be calculated, the words Kin_use that have been used are the three words “Minato Mirai”, “Fujiya”, and “Yokohama”. For each of these words, the degree of use history is evaluated to the same extent (S33A). In this example, the state of the database is expressed using “same section collected word-used set”. In other words, the “same section collected word-used set” for the word A refers to a word that has been collected in the past as the same unit section as the word A and has been used.

  For each word in the word set Kin_use with usage history among the words in the collection section to be evaluated, the collected words from the same section up to the previous time [time t]-used set and the current time [time t + 1] The same section sampled word-used set is obtained (S33B, S33C). For example, for “Minato Mirai”, referring to the real-world information in the accumulation database 44, the same section collected words collected in the same section as “Minato Mirai” are extracted, and each word application usage history in the history database 46 is extracted. Referring to 46B, a word that has been used up to the time t in the past is searched for among the words collected in the same section, and the same section collected word-usable set [time t] is obtained (S33B). Similarly, a word that has been used up to the current time t + 1 among the words collected in the same section is searched to obtain the same section collected word-used set [time t + 1] (S33C).

  Next, the above-mentioned two same-segment collected words-used sets are compared, and the change is detected to calculate the same degree of use history (S33D). Specifically, in the case of “Minato Mirai”, the set up to the present time was 1000 words, and the set up to the past time and 999 words were common. Therefore, to some extent, the stable evaluation value with the same usage history for “Minato Mirai” is 999/1000. Similarly, “Fujiya” is 4/4 and “Yokohama” is 1700/2000.

  The basis of the above calculation is the word A if it is collected in the same situation as the word A and the available word does not change much between the past and the current point in time. In the vicinity of, it is based on the assumption that the usage will not change significantly, that is, the words are used stably.

  In addition to the calculation method of FIG. 15, to the same extent as the usage history, for the usage history existing words in the collection interval, for example, the usage frequency or the number of usages up to a past time point and the current time point, for example, The frequency of use or the number of times of use in a predetermined period may be compared and obtained based on the same degree.

Returning to FIG. 14, in step S33, the weight w1 of the steady information acquisition tendency is obtained by the following equation.
w1 = Stable / u = (addition of stable evaluation of words with usage history) / u = (999/1000 + 4/4 + 1700/2000) /3=0.95
That is, the weight w1 of the steady information acquisition tendency as the section is obtained by averaging the Stable evaluation values of the words with usage history in the section.

  On the other hand, the weight w2 of the primary information acquisition tendency is w2 = 1−w1 = 1−0.95 = 0.05. Since these weights w1 and w2 are the weights in the section, the higher the stable value of the word with usage history, the higher the weight of the steady information acquisition score USED for the word with usage history in the section. The lower the stable value of a certain word, the higher the weight of the score UNUSED for acquiring the primary information for the word without usage history in that section.

  In the example of FIG. 14, as a result of the stable evaluation, it has been found that there is not much change even when the usage histories when the words with usage history included in the input data are collected in the past are compared. As a result, the weight w1 of the steady information acquisition tendency in that category is as high as 0.95, and the weight w2 of the primary information acquisition tendency is 0.05.

Next, interval collection importance p is calculated (S34). The formula for calculating the interval collection importance p is as follows.
Collection importance p = USED ・ w1 + UNUSED ・ w2
USED is a constant information acquisition score and UNUSED is a primary information acquisition score. In the example of FIG. 14, the words with no usage history and the ratios of the words with usage history to the total number of words are used. . That is, words with no usage history are 6 out of 9 words collected, words with usage history are 3 out of 9 words collected, USED = 3/9, UNUSED = 6/9.

  That is, if the collected unit interval data contains many words with usage history, the steady-state information acquisition score increases, and if many words without usage history are included, the primary information acquisition score increases. . However, these scores are weighted by the weights w1 and w2 corresponding to the aforementioned phases.

As a result of substituting into the above formula, the interval collection importance p is
p = 3 / 9x0.95 + 6 / 9x0.05 = 0.35
It becomes.

  In this example, since the weight w1 of the steady information acquisition tendency is larger than the ratio of words with usage history to the total number of extracted keywords (steady information acquisition score USED), this increases the importance of interval collection. ing. However, the ratio of words without usage history (primary information acquisition score UNUSED) was high, but the weight w2 of the primary information acquisition tendency was small, so the contribution to the interval collection importance was small.

  Next, the contribution score to the interval collection importance of each word is obtained by the formula in FIG. In the example of FIG. 14, this contribution score is divided equally, USED · w1 is divided into three equal parts for three words with usage history, and UNUSED · w2 is six for six words without usage history. The contribution scores (evaluation values) are 0.11 and 0.01. The contribution score of each word collection importance is recorded in the word collection importance contribution score 47B in the evaluation value database 47 and added to the importance of each word importance table 47C.

  In the example of FIG. 14, the words “Minato Mirai”, “Fujiya”, and “Yokohama” with usage history are words that have been used in the past, but the stable evaluation value is used to determine whether or not the word is used stably. If it is stable (if the Stable rating or w1 is high), it is estimated that the user will want to display it on the panel. On the other hand, the word “taste log” with no usage history is a word that has never been used in the past, but if the degree of instability (corresponding to W2) is high, the primary information acquisition tendency is strong. It is presumed that the user would have displayed words that were not used in the past on the panel.

  In this way, the number of words with usage history in the section (words) has a large number of words with usage history, the steady information acquisition score USED will increase, but depending on the stability (Stable value) of the words with usage history, Evaluation of words with history changes. Conversely, if the number of words without usage history is large and the primary information acquisition score UNUSED is large, and the stability of words with usage history is low, the primary information acquisition weight w2 for that section increases and the usage is increased. The evaluation of words without history (contribution score of each word collection importance) becomes high.

  However, in the above example, the evaluations for all the words in the section are averaged and treated equally in the words with usage history and the words without usage history.

  FIG. 16 is a diagram showing processing after being added to each word importance table. This process is the same as in the first embodiment (FIG. 10). The difference from FIG. 10 is that the importance of the real world information is the cumulative value of the contribution scores of the word collection importance for each word. Other than that, it is the same as FIG.

  FIG. 14 shows a process in which words included in the real world collection information are registered on the importance table. The words used in each application in this manner are given importance according to the operation of each application. In the importance of the real world information shown in the second embodiment, an example is shown in which contribution scores to the interval collection importance of each word are accumulated. In the example shown in FIG. 14, the steady information acquisition tendency is strong, and as a result, the evaluation value for a word having a usage record is high. Therefore, in such a state, the word having a usage record raises the evaluation value to the same rate. To go.

  In the present embodiment, the case where the interval collection importance is calculated in consideration of the degree of information acquisition tendency (w1, w2) at the time of real world information collection is illustrated. However, it is also possible to record a plurality of importance levels, such as the importance level when the degree of information acquisition tendency (w1, w2) is not taken into account, and use them according to user preferences.

  In addition, display prioritization may be performed with reference to information obtained from an external service. For example, given a word group and a word as input, if there is an external service indicating the degree of association between the word group and the word, the word group and the user word group included in the interest on the user's panel from the external service. The degree of relevance with the candidate word may be obtained, and the display may be in the vicinity of the interest with strong relevance. In this case, according to the degree of each information acquisition tendency, if the primary information acquisition tendency is strong, information that can be added in the vicinity of the newly added interest may be displayed.

  Further, in the present embodiment, the interval collection importance p is once obtained. When the section collection importance p is lower than the specified value, it may be excluded from the user gaze section data.

[Third Embodiment]
FIG. 17 is a configuration diagram of a user information management system according to the third embodiment. The difference from the second embodiment of FIG. 13 is that the collection information processing unit 10 has an appearance opportunity prediction unit 10D.

  In the second embodiment, the contribution scores to the section collection importance level p of words with usage history and words without usage history are averaged. That is, the contribution scores of all words with usage history are the same, and the contribution scores of all words without usage history are the same. However, words with usage history are collected again in the future even if the frequency of appearance opportunities is high and the importance is not particularly high by analyzing the rare degree of appearance opportunities as collected words from the usage history. It can be distinguished whether the appearance opportunity is rare, and if it is not particularly important this time, it will not be collected again in the future and will not be displayed.

  Thus, in the third embodiment, in the calculation of the interval collection importance shown in the second embodiment, when there is a steady information acquisition tendency, it is difficult to accumulate words in the system. The case where importance is calculated in consideration of a certain degree will be described.

  FIG. 18 is a flowchart of processing of the section collection importance level determination unit 10C in the present embodiment. The difference from the second embodiment of FIG. 14 is that, in the collection importance score calculation step S35, the steady information acquisition score USED is calculated based on the rare degree of appearance opportunity (OImportance). In the contribution score to the interval collection importance degree p, there is a difference between the three usage history words.

  In the second embodiment of FIG. 14, the collection importance calculation process is performed in the collection information processing, and the collection importance contribution score of each word obtained in the calculation process is accumulated. The words with high importance are preferentially displayed on the panel. As in the example shown in FIG. 14, when the steady information acquisition tendency is strong, the calculation of each word collection importance contribution score as in the second embodiment will result in the use of words that have been used. As the evaluation value becomes higher, if the steady information acquisition tendency continues, words with a track record of use will increase in importance over time.

  On the other hand, in the present embodiment in FIG. 18, the degree to which the input data to be examined is difficult to be accumulated in the system by including the word combination appearance opportunity prediction process in the collected information processing. Taking into account, the contribution score to the importance of collection for each word is calculated. As a result, a difference occurs in the contribution scores of words with usage history.

  In FIG. 18, preprocessing (S30) is performed at the start of input data collection importance calculation, phase determination is performed (S31), and the weight of information acquisition tendency is determined (S33). Similar to the example. Then, in the collection importance score calculation step S35 for each tendency, the collection importance scores UNUSED and USED for the primary / steady information acquisition tendency are obtained. As for the score UNUSED for words without usage history, the ratio of the number of words without usage history Kin_notuse to the number of extracted keywords Kin of input data is used as in the example of FIG.

  On the other hand, the score USED for words with usage history is not simply set to the score USED, but the higher the degree of difficulty in accumulating in the system, the higher the score. As a result, words that rarely appear are given a contribution score to a high degree of importance in one collection opportunity, and are easily accumulated in the system.

  FIG. 19 is a flowchart of the score USED calculation process for routine information acquisition. In this example, the rare degree of appearance opportunity (OImportance: Opportunity Importance) is obtained for each word with a usage history of the extracted keyword (word). Rare degree of appearance opportunity The basic idea of OImportance is (1) For a word A, the words collected in the same unit section in the past, the more words that have been used, the more I think that there are many appearance opportunities. The fact that there are many appearance opportunities in the past is predicted that there will be many appearance opportunities in the future, and it is likely to be included in the input data at other timings in the future. , It is not very important at this collection opportunity.

  Also, (2) For input data that is subject to importance calculation, there are so many words that have been used in the same unit section and have been used in the past. It can be assumed that there is a high possibility that a word that may be used in the future is included in the same interval collected word-a word with a use-set (in common with a word with a use history of input data).

  Using these two concepts, OImportance is calculated for a rare degree of opportunity.

  As shown in FIG. 19, the calculation of OImportance is rarely performed for each word (Minato Mirai, Fujiya, Yokohama) included in the word set Kin_use with extracted keyword usage history (S35A).

  First, by referring to the accumulation database and the history database, the same-interval-collected word-used set for each word of the usage history word set Kin_use is obtained (S35B). In the example of FIG. 19, 1000 words for “Minato Mirai”, only 4 words for “Fujiya”, and 2000 words for “Yokohama” It is a word with a usage history.

  Then, for each word, the rare opportunity OImportance is obtained as follows (S35C). In other words, among the words obtained in step S35B, the number of words that are also in the words with the usage history of the input data to be evaluated this time is counted, and the ratio of the number to the section collected word-usage set is rare. About OImportance.

  For example, “Fujiya” has 4 words in the same interval collected word-used set, 2 words are also included in the extracted keyword usage history set Kin_use, and OImportance for “Fujiya” is 2/4 And evaluate. Similarly, “Minato Mirai” was 2/1000 and “Yokohama” was 2/2000.

  In step S35D, the average OImportance of these three words is calculated as 0.17. This is used as the collection important score USED of the steady information acquisition tendency.

  However, the rare degree of appearance of each word is “Fujiya”, and common words such as “Minato Mirai” and “Yokohama” are low. These are reflected in the contribution score to the importance of collection. In other words, “Fujiya” has only four words in the same-word collection-use-with-set, and has not appeared at other occasions, so it rarely appears for a long time. I think there is no opportunity to do it. In other words, OImportance's denominator is small enough to rarely appear. Furthermore, the words with usage history collected at the same time as “Fujiya” are the same as “Fujiya”, which is the same-word collected word-used-with-set word. OImportance, which is a rare opportunity to appear, is highly appreciated in the case of “Fujiya” because the numerator is larger than the denominator.

  The calculation of the rare degree of appearance opportunity may be obtained by simplifying as follows. That is, for each word with a usage history that is used in each word application usage history database among the words included in the user gaze interval data, the same unit section as the word with the usage history and the word with the usage history in the past If the number of past collections of words collected in (2) is large, the rare chance of appearance is low, and if the past collections are small, the rare degree of appearance is high. Even in such a calculation, a word with a usage history among words in the user's gaze interval data is less likely to reappear with words collected in the same unit interval as the word with the usage history in the past (the chance of appearance is rare). It is possible to calculate whether the degree is high) or low (the rare degree of appearance is low).

Returning to FIG. 18, in the collection importance calculation step S34, first, the section collection importance p, which is the collection importance of the entire extracted keywords, is obtained as follows.
p = USED ・ w1 + UNUSED ・ w2 = 0.17 x 0.95 + 6/9 x 0.005 = 0.19
Then, the contribution score of each word to the collection importance is calculated as a score contribution rate to the section collection importance p. As shown in FIG. 14, the score of the word with the use history is calculated by the following formula.
Contribution score to each word collection importance = (word OImportance / | Kin_use |) x w1
The score of words without usage history is the same as in the second embodiment. The reason for dividing by | Kin_use | (number of words with usage history) and | Kin_notuse | (number of words without usage history) is to normalize the evaluation value that can be obtained by one collection. This embodiment is also the same.

  As a result, the contribution score to the collection importance of each word is 0.00 for “Minato Mirai” and “Yokohama”, 0.16 for “Fujiya”, and 0.01 for “words without usage history”. In other words, even for words with the same usage record, “Fujiya”, which was judged to have a rare appearance opportunity, had a higher word evaluation value than “Minato Mirai” and “Yokohama”, which had many appearance opportunities. ing. This is different from the second embodiment.

  FIG. 20 is a diagram illustrating processing after addition to each word importance table. In this process, as in the second embodiment (FIG. 16), the contribution score to each word collection importance is added to the importance of each word in the importance table. However, the difference from FIG. 16 is that the contribution score to the importance of collection for each word added to the importance of the real world information is 0.00 for “Yokohama” and “Minato Mirai”, as shown in FIG. Is 0.16, and their cumulative values are different.

  For this reason, “Fujiya”, which has rare appearance opportunities, is more important than “Yokohama” and “Minato Mirai”, which have many appearance opportunities, and is more easily displayed on the display panel.

  Furthermore, in the embodiment in which the contribution score to the collection importance level considering the appearance opportunity is emphasized, the time series change is stored in the importance level calculation process, and the change in the importance level is not a gradual increase, but specific. The display order may be given priority with respect to the splashed words (that is, words such as “Fujiya” that have rare appearance opportunities and high contribution scores). As a result, it is possible to suppress the possibility of being buried because the appearance opportunity is rare.

  Also, the cumulative importance level is stored separately for the importance level of each word in the steady information acquisition tendency and the primary information acquisition tendency, and when the display priority number is assigned to the object, the current information acquisition is performed. Display priority may be determined at a rate that matches the trend.

[Community with multiple users]
FIG. 21 is a diagram illustrating an example of a panel when a user shares an interest with another user. The panel displays a private panel Close that displays only its own interests as shown in FIG. 1 and the like and is not disclosed to other users, and a public panel Open that shares the interests with other users.

  This panel shows two panels, a closed panel Close and a travel system open panel Open. The closed panel Close is a screen that only user A can see. Therefore, information objects automatically extracted from the user word group are floating. Therefore, with this closed panel Close, you can accumulate your interests from application usage history and collected real world information.

On the other hand, the public panel Open can be viewed by other users of the information management device T _s , for example. User A discloses travel-related concerns and names the public panel as a travel-related public panel.

  FIG. 20 shows an example of the panel display viewed by the user A. For example, a panel for viewing the community D to which a plurality of users A, B, and C belong may be prepared. As in the case of FIG. 20, the community-oriented panel includes two or more panels, and includes a private panel managed only by a plurality of users registered in the community and a panel disclosed outside the community.

  User A publishes his / her interests to an unspecified number of users by publishing his / her travel-related concerns on the travel system public panel on the panel of FIG. However, for example, by dragging and dropping the interests of the travel community onto their private panel, the travel related interests can be managed only by the users A, B, and C belonging to the travel community.

  On the panel, for example, additional information (contact history) is collected, such as how long the user has kept the object on the panel as an interest, and whether the user has viewed the object close to the display. By using this contact history, it may be possible to provide support such as collecting and presenting information on the interests of interest.

  On the panel, additional information (contribution history) is collected, such as which interests of a community a user is providing information about. By using this contribution amount, active information collection by the server such as requesting provision of information at a timing that seems to have time according to the convenience of the user may be performed.

  In addition, on the panel, how many users refer to the information provided by a user from the history that other users have acquired the information provided by a user in their community. And the degree of influence (influence) of the user on the community may be calculated. By using this influence, it is possible to adopt a service form in which information on users with strong influence is preferentially presented within the community.

  As described above, according to the first, second, and third embodiments, it is determined whether or not the real-world information automatically collected is the information that the user has watched, and the user has been watching or For the keywords (words) included in the real-world information estimated to be watched, the importance of collection is calculated, and the words with the highest importance are displayed on the panel. Therefore, the word that the user wants to use can be displayed on the panel, and the user can arbitrarily link it to the interest or use it in the application, so that convenience can be improved.

  The above embodiment is summarized as follows.

(Appendix 1)
A user collection information management apparatus having a client terminal and a user information management server capable of communicating with the client terminal,
The user information management server
A user gaze interval determination unit that inputs and accumulates unit interval data having real world collection information collected in the real world in a unit interval of time, and determines whether or not the unit interval data is information that the user gazes;
Each word collection importance calculator for calculating the collection importance of words included in the user gaze interval data;
In addition to the user word group to be displayed according to the collection importance of the words included in the user gaze interval data, a display processing unit that displays words of the user word group on the display unit of the client terminal ,
The client terminal is
A user information management apparatus comprising: an interest management unit that associates the user-operated word with the interest in response to a user operation that associates a word of the user word group displayed on the display unit with the interest.

(Appendix 2)
In Appendix 1,
The user information management server further includes:
A real-world collection information database for accumulating words included in the unit interval data together with at least the collection time;
Each word application usage history database for storing the usage history of the words used in the application program in the client terminal,
Each of the word collection importance calculators
Of the words included in the user gaze section data, the usage history up to the past time and the usage history up to the current time for each usage history word that is used in each word application usage history database Obtain the same usage history indicating the degree (stable),
In the new interest-related information formation phase after the user operation for associating with the interest at the client terminal is performed, the weight w1 of the steady information acquisition tendency is set to the first value, and the primary information acquisition The weight w2 is a second value higher than the first value,
In the continuous interest-related information formation phase after the new interest-related information formation phase, the weight w1 is a value based on the same degree of use history, and the weight w2 is 1-w1,
A regular information acquisition score (USED) based on the number of words with usage history that the user gaze interval data has, and a primary information acquisition score (UNUSED) based on the number of words without usage history that the user gaze interval data has , Obtaining the interval collection importance weighted by the weight w1 of the steady information acquisition tendency and the weight w2 of the primary information acquisition tendency, respectively, and for the words with usage history and the words without usage history included in the user gaze interval data The contribution score to interval collection importance is calculated as the collection importance of each word.
User information management device.

(Appendix 3)
In Appendix 2,
The same degree of use history is the same unit section as the word with use history in the past for each use history word used in each word application use history database among the words included in the user gaze section data. It is obtained by the degree of common between the words that have been used in the past and that have a history of use up to the past time, and the words that have a history of use up to the present time.
User information management device.

(Appendix 4)
In Appendix 2,
The stationary information acquisition score is a ratio of the number of words with usage history to the total number of words of the user gaze interval data,
The primary information acquisition score is a ratio of the number of words without usage history to the total number of words included in the user gaze interval data.
User information management device.

(Appendix 5)
In Appendix 1,
The user information management server further includes:
A real-world collection information database for accumulating words included in the unit interval data together with at least the collection time;
Each word application usage history database for storing the usage history of the words used in the application program in the client terminal,
Each of the word collection importance calculators
Of the words included in the user gaze section data, the usage history up to the past time and the usage history up to the current time for each usage history word that is used in each word application usage history database Obtain the same usage history indicating the degree (stable),
In a new interest-related information formation phase after a user operation for associating with the interest at the client terminal is performed, the weight w1 of the steady information acquisition tendency is the first value, and the weight of primary information acquisition Let w1 be a second value higher than the first value,
In the continuous interest-related information forming phase after the new interest-related information forming phase, the weight w1 is a value based on the same degree of use history, and the weight w2 is 1-w1,
Among words included in the user gazing interval data, for each word with usage history that is used in each word application usage history database, in the same unit interval as the word with usage history and the word with usage history in the past Obtained as a rare degree of opportunity for the collected words based on the number of past collections,
The words with usage history included in the user gaze interval data are weighted with the weight w1 to a rare degree of appearance opportunities, and the words without usage history with respect to the total number of words of the user gaze interval data are used for the words without usage history. The primary information acquisition score (UNUSED), which is a percentage of the number, is weighted by the weight w2, and the contribution score to the interval collection importance is calculated as the collection importance of each word.
User information management device.

(Appendix 6)
In Appendix 5,
The same degree of use history is the same unit section as the word with use history in the past for each use history word used in each word application use history database among the words included in the user gaze section data. It is obtained by the degree of common between the words that have been used in the past and that have a history of use up to the past time, and the words that have a history of use up to the present time.
User information management device.

(Appendix 7)
In Appendix 5,
The primary information acquisition score is a ratio of the number of words without usage history to the total number of words included in the user gaze interval data.
User information management device.

(Appendix 8)
In Appendix 5,
The rare degree of the appearance opportunity is the same as the word with usage history in the past for each word with usage history that has been used in each word application usage history database among the words included in the user gaze interval data. The words with the usage history included in the collected word / usage-use set included in the same-section collected word / usage-use set with respect to the number of words of the same-section collection word / usage-use set that are words collected in the unit interval and used in the past. The ratio of numbers,
User information management device.

(Appendix 9)
In Appendix 1, 2, or 5,
The display processing unit adds to the user word group according to the cumulative importance of collecting each word,
User information management device.

(Appendix 10)
In Appendix 1, 2, or 5,
The user information management server inputs user gaze information having a gaze state of the user when the user collects the real world collection information,
The user gaze interval determination unit determines whether or not the unit interval data is information that the user gazes based on the user gaze information.
User information management device.

(Appendix 11)
In Appendix 1, 2, or 5,
A real world collection information database for accumulating the unit interval data together with at least the collection time;
The user gaze interval determination unit calculates a similarity between the input unit interval data and the accumulated unit interval data accumulated in the real world collection information database, and the user gaze interval when the similarity exceeds a threshold value ,
User information management device.

(Appendix 12)
A user information management method in a user collection information management apparatus having a client terminal and a user information management server capable of communicating with the client terminal via a network,
The user information management server
A user gazing interval determination step for inputting and accumulating unit interval data having real world collection information collected in the real world in a unit interval of time, and determining whether the unit interval data is information that the user has observed;
Each word collection importance calculating step for calculating the word collection importance included in the user gaze interval data;
In addition to the user word group to be displayed according to the collection importance of the word included in the user gaze interval data, a display processing step of displaying the word of the user word group on the display unit of the client terminal,
The client terminal is
A user information management method for performing a concern management step of associating the user-operated word with the interest in response to a user operation for associating the word of the user word group displayed on the display unit with a certain interest.

(Appendix 13)
In Appendix 12,
The user information management server further includes:
A real-world collection information database for accumulating words included in the unit interval data together with at least the collection time;
With reference to each word application usage history database that accumulates the usage history of words used in application programs at the client terminal,
In each word collection importance calculation step,
A regular information acquisition score (USED) based on the number of words with usage history that the user gaze interval data has, and a primary information acquisition score (UNUSED) based on the number of words without usage history that the user gaze interval data has , Obtaining the interval collection importance weighted by the weight w1 of the stationary information acquisition tendency corresponding to the acquisition tendency of the stationary information and the weight w2 of the primary information acquisition tendency corresponding to the acquisition tendency of the new information, respectively, The contribution score to the interval collection importance for words with usage history and words without usage history included in the gaze interval data is calculated as the collection importance of each word.
User information management method.

Tc1: Client terminal
Ts: User information management server
Rin: Real World Information
Fin: User gaze information

Claims (10)

A user collection information management device having a user information management device capable of communicating with a terminal,
The user information management device includes:
A user gaze interval determination unit that inputs and accumulates unit interval data having real world collection information collected in the real world in a unit interval of time, and determines whether or not the unit interval data is information that the user gazes;
Each word collection importance calculator for calculating the collection importance of words included in the user gaze interval data;
In addition to the user word group to be displayed according to the collection importance of the words included in the user gaze interval data, a display processing unit that displays the words of the user word group on the display unit of the terminal,
A user information management device in which the user-operated word is associated with the interest in response to a user operation that associates a word of the user word group displayed on the display unit with a certain interest. In claim 1,
The user information management device further includes:
A real-world collection information database for accumulating words included in the unit interval data together with at least the collection time;
Access to each word application usage history database that accumulates the usage history of words used in application programs in the terminal,
Each of the word collection importance calculators
Of the words included in the user gaze section data, the usage history up to the past time and the usage history up to the current time for each usage history word that is used in each word application usage history database Obtain the same degree of usage history indicating the degree,
In a new interest-related information formation phase after a user operation for associating with the interest at the terminal is performed, the weight w1 of the steady information acquisition tendency is set to the first value, and the weight for primary information acquisition Let w2 be a second value higher than the first value,
In the continuous interest-related information formation phase after the new interest-related information formation phase, the weight w1 is a value based on the same degree of use history, and the weight w2 is 1-w1,
The steady information acquisition score based on the number of words with usage history that the user gaze section data has and the primary information acquisition score based on the number of words without usage history that the user gaze section data has The section collection importance weighted by the weight w1 of the tendency and the weight w2 of the primary information acquisition tendency is obtained, and the section collection importance for the words with usage history and the words without usage history included in the user gaze section data is obtained. Calculate the contribution score as the importance of collecting each word,
User information management device. In claim 2,
The same degree of use history is the same unit section as the word with use history in the past for each use history word used in each word application use history database among the words included in the user gaze section data. It is obtained by the degree of common between the words that have been used in the past and that have a history of use up to the past time, and the words that have a history of use up to the present time.
User information management device. In claim 2,
The stationary information acquisition score is a ratio of the number of words with usage history to the total number of words of the user gaze interval data,
The primary information acquisition score is a ratio of the number of words without usage history to the total number of words included in the user gaze interval data.
User information management device. In claim 1,
The user information management device further includes:
A real-world collection information database for accumulating words included in the unit interval data together with at least the collection time;
Access to each word application usage history database that accumulates the usage history of words used in application programs in the terminal,
Each of the word collection importance calculators
Of the words included in the user gaze section data, the usage history up to the past time and the usage history up to the current time for each usage history word that is used in each word application usage history database Obtain the same degree of usage history indicating the degree,
In a new interest-related information formation phase after a user operation for associating with the interest at the terminal is performed, the weight w1 of the steady information acquisition tendency is the first value, and the weight w2 of the primary information acquisition Is a second value higher than the first value,
In the continuous interest-related information forming phase after the new interest-related information forming phase, the weight w1 is a value based on the same degree of use history, and the weight w2 is 1-w1,
Among words included in the user gazing interval data, for each word with usage history that is used in each word application usage history database, in the same unit interval as the word with usage history and the word with usage history in the past Obtained as a rare degree of opportunity for the collected words based on the number of past collections,
The words with usage history included in the user gaze interval data are weighted with the weight w1 to a rare degree of appearance opportunities, and the words without usage history with respect to the total number of words of the user gaze interval data are used for the words without usage history. A primary information acquisition score that is a percentage of the number is weighted with the weight w2, and a contribution score to the interval collection importance is calculated as the collection importance of each word.
User information management device. In claim 5,
The rare degree of the appearance opportunity is the same as the word with usage history in the past for each word with usage history that has been used in each word application usage history database among the words included in the user gaze interval data. The words with the usage history included in the collected word / usage-use set included in the same-section collected word / usage-use set with respect to the number of words of the same-section collection word / usage-use set that are words collected in the unit interval and used in the past. The ratio of numbers,
User information management device. In claim 1, 2 or 5,
The display processing unit is a user information management device that adds to the user word group according to the cumulative collection importance of each word. In claim 1, 2 or 5,
The user information management device inputs user gaze information having a gaze state of the user when the user collects the real world collection information,
The user gaze interval determination unit determines whether or not the unit interval data is information that the user gazes based on the user gaze information.
User information management device. In claim 1, 2 or 5,
The user information management device is made accessible to a real world collection information database that accumulates the unit interval data together with at least the collection time,
The user gaze interval determination unit calculates a similarity between the input unit interval data and the accumulated unit interval data accumulated in the real world collection information database, and the user gaze interval when the similarity exceeds a threshold value ,
User information management device. A user information management method in a user collection information management apparatus having a user information management apparatus capable of communicating with a terminal,
The user information management device includes:
A user gazing interval determination step for inputting and accumulating unit interval data having real world collection information collected in the real world in a unit interval of time, and determining whether the unit interval data is information that the user has observed;
Each word collection importance calculating step for calculating the word collection importance included in the user gaze interval data;
In addition to the user word group to be displayed according to the collection importance of the word included in the user gaze interval data, a display processing step of displaying the word of the user word group on the display unit of the terminal,
A user information management method in which the user-operated word is associated with the interest in response to a user operation that associates a word of the user word group displayed on the display unit with a certain interest.