JP4445849B2 - File search method, file search device, and file search program - Google Patents

Description

  The present invention relates to a technique for searching for a file, and more particularly to a technique for searching for a similar file among files browsed by a user in the past.

  As a service for searching for similar files, for example, there is a search site (Google (registered trademark)) for information search provided by Google (registered trademark). This search site provides link destination information of a Web page related to the Web page for each URL of the Web page as a search result. That is, in this search site, the similarity is calculated from the contents of the Web page, the URL of the site that publishes the Web page, etc., and the related Web page is searched.

  Further, as a technique for managing information browsed by the user in the past, for example, there are “favorite” and “history”. “Favorites” and “History” are functions provided by Internet Explorer (registered trademark) which is a Web browser of Microsoft Corporation (registered trademark). The user can easily access Web pages browsed in the past by using the functions of “favorites” and “history”.

Also, in Patent Document 1, an evaluation value of each Web page is calculated based on the user's operation history, and the URL of the Web page highly evaluated by other users having similar evaluation values is provided to the user. A browsing support device is described.
JP-A-10-307845

  In general, when searching for a file similar to a predetermined file (including a Web page), similar files are displayed in the order of similar scores calculated using keywords, file creation date, access date and time, and the like. However, if the number of searched similar files is enormous, files useful for the user are not always displayed at the top. For this reason, it may be difficult for the user to access a useful file from similar files that have been searched.

  On the other hand, a file browsed in the past by a user is often more useful than a file that has never been browsed. Moreover, even if it is the file browsed in the past, it is thought that a useful degree changes according to the condition where the user browsed the said file. For example, it is considered that the latter is more useful for the user when the user simply displays the file and when the user understands the contents of the displayed file.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to search for similar files that are more useful to the user.

In order to solve the above-described problem, for example, in the first invention, a file search method performed by the file search device, the file search device includes a history storage unit in which information of browsed files browsed in the past is stored, And a processing unit. Then, the processing unit refers to the display file acquisition step for acquiring the display file displayed on the display device and the history storage unit, and calculates a similarity score indicating the similarity to the display file for each browsed file. Similar to a similar score calculation step, an experience score calculation step that refers to the history storage unit, calculates an experience score indicating the strength of the user's experience with respect to the browsed file when viewed in the past, and similar The score distribution ratio that determines the ratio between the score and the experience score is multiplied by the similarity score and the experience score calculated for each browsed file to calculate an integrated score, each integrated score, and the integrated score An output step of outputting information of the browsed file stored in the history storage unit corresponding to the score to the display device; Gastric row, in the history storage unit, for each viewing already file, and the time an application becomes active state is displayed on the display device the viewing already file is stored as viewing start time is displayed on the display device The time when the file was switched from the browsed file to the new file is stored as the browse end time, and the experience score calculation step is browsed according to the browse time obtained from the browse start time and the browse end time of the history storage unit. Calculate the experience score for the file.

The second invention is a file search device in a file search system having a file search device and a display device, the history storage means storing information of browsed files browsed in the past, and display on the display means Display file acquisition means for acquiring the displayed file, history storage means, similarity score calculation means for calculating a similarity score indicating the similarity to the display file for each browsed file, and history storage means For each browsed file , the ratio of the experience score calculation means for calculating the experience score indicating the strength of the user's experience with respect to the browsed file when browsing in the past, and the ratio between the similarity score and the experience score was determined. the score distribution ratio, multiplied by the similarity score and experience score was calculated for each viewing already file, integrated score calculation to calculate a combined score Means, and the total score respectively, possess and output means for outputting the information browsing already files stored in the history storage means corresponding to the total score, the display means, and the history storage means, viewing already For each file, the time when the application that was in the active state displayed the browsed file on the display device is stored as the browsing start time, and the file displayed on the display device has been switched from the browsed file to a new file The time is stored as the browsing end time, and the experience score calculating unit calculates the experience score of the browsed file according to the browsing time obtained from the browsing start time and the browsing end time of the history storage unit.

The third invention is a file search program for causing a computer to function as each means constituting the file search device according to the second invention.

  According to the present invention, it is possible to search for similar files that are more useful to the user. This also allows the user to easily access more useful similar files.

  Embodiments of the present invention will be described below.

  FIG. 1 is a configuration diagram of a file search system to which an embodiment of the present invention is applied.

  The illustrated similar file search system is a system for searching for a file similar to a predetermined file. As illustrated, the similar file search system includes a web browser 11, a similar file display unit 12, a browsing file acquisition unit 21, a search management unit 22, a similar score calculation unit 23, and an experience score calculation unit 24. The operation history acquisition unit 25 and the storage unit 26 are provided.

  The web browser 11 is an application for browsing web pages. That is, the Web browser 11 downloads various information (HTML files, images, etc.) from a Web server (not shown) via a network such as the Internet, and displays it on an output device (not shown). The similar file display unit 12 displays a file similar to a predetermined file on the output device.

  The browsing file acquisition unit 21 acquires a browsing file (display file) displayed on the output device. The search management unit 22 searches for a file similar to the browsing file acquired by the browsing file acquisition unit 21. The similarity score calculation unit 23 calculates a similarity score indicating the degree of similarity with the file being browsed for each browsed file stored in the storage unit 26. The experience score calculation unit 24 calculates, for each browsed file stored in the storage unit 26, an experience score indicating the user's operation status when browsing the browsed file in the past. The operation history acquisition unit 25 acquires a browsed file based on a user operation instruction and stores it in the storage unit 26.

  The storage unit 26 stores a browsed file 27, a file attribute information DB 28, and an operation history DB 29. The browsed file 27 is the information body of the browsed file, and is, for example, an HTML file, a text data file, an image data file, or the like that displays a Web page. The file attribute information DB 28 is a database in which attributes of each browsed file 27 are stored. The operation history DB 29 is a database in which operation contents of each browsed file 27 are stored. The file attribute information DB 28 and the operation history DB 29 will be described later.

  The similar file search system described above includes, for example, a CPU, a memory, an external storage device such as an HDD, an input device such as a keyboard and a mouse, and an output device such as a display and a printer, and communication control for network connection. And a general-purpose computer system provided with the apparatus. In this computer system, each function is realized by the CPU executing the similar file search system program loaded on the memory. Note that a memory or an external storage device is used for the storage unit 26.

  In addition, the similar file search system in the present embodiment uses one PC (Personal Computer). However, the present invention is not limited to this. For example, the functions of the Web browser 11 and the similar file display unit 12 may be realized on the PC, and the other functions 21 to 29 may be realized on the server. In this case, the PC and the server each include a CPU, a memory, an external storage device, and the like. The functions of the PC and the server are realized by the CPU of the PC executing the program for the PC and the CPU of the server executing the program for the server, respectively.

  Next, the operation history DB 29 and the file attribute information DB 28 in the storage unit 26 will be described.

  FIG. 2 is a diagram illustrating an example of the operation history DB 29. The operation history DB 29 stores an index (operation content) necessary for calculating an experience score. The operation history DB 29 shown in the figure includes a browsing start time 201, a browsing end time 202, a URL 203, a hash value 204, and a file storage location 205 for each browsed file.

  The browsing start time 201 is set to the time when the browsed file is displayed on the output device. The browsing end time 202 is set to the time when another file other than the browsed file is displayed on the output device (that is, the time when browsing is ended). In the URL 203, an address when the browsed file exists on the Internet is set. For the hash value, a fixed-length pseudorandom number generated from the information body of the browsed file is set using, for example, an MD5 hash function. In the file storage location 205, the storage location of the browsed file 27 stored in the storage unit 26 is set with a full path.

  In the present embodiment, the browsing time calculated from the browsing start time 201 and the browsing end time 202 is used as an index for calculating the experience score. In other words, understanding the information described in a file that has been viewed (performed) over a longer period of time is compared to understanding the information described in a file that has been browsed for a relatively short time in the past. It can be said that it is easy. Therefore, in this embodiment, browsing time is used as an index of the experience score.

  However, the present invention is not limited to this, and as an index for calculating the experience score, there is a print instruction for a browsed file, a mouse click (such as an instruction to move to a link destination), text input, etc. An operation may be used.

  FIG. 3 is a diagram showing an example of the file attribute information DB 28. The file attribute information DB 28 stores an index (file attribute) necessary for calculating a similarity score. The file attribute information DB 28 shown in the figure has a file storage location 301, an author 302, a title 303, an update date 304, a hash value 305, and a URL 306 for each browsed file browsed in the past.

  In the present embodiment, the author 302, the title 303, the update date 304, and the browsed file 27 main body stored in the storage unit 26 are used as indices for calculating the similarity score.

  Next, an operation history acquisition process for a browsed file of this system will be described. In the present embodiment, it is assumed that the operation history acquisition process is repeatedly performed while the OS (Operating System) of the system is activated.

  FIG. 4 is a process flow diagram of the operation history acquisition process. The operation history acquisition unit 25 acquires (hooks) a message that is globally communicated on the OS, and identifies an application that is in an active state (S41). The message notifies the target application of various events received by the OS from an input device such as a mouse or a keyboard. In the illustrated example, the identified active application is assumed to be the Web browser 11 and will be described below.

  Then, the operation history acquisition unit 25 communicates with the Web browser 11 by inter-program communication, and the file displayed on the output device by the Web browser 11 (in this case, an HTML file for displaying the Web page), The URL of the file is acquired (S42). Then, the operation history acquisition unit 25 calculates the hash value of the acquired file body (S43).

  Then, the operation history acquisition unit 25 stores the acquired file in the storage unit 26 as the browsed file 27 (S44). Then, the operation history acquisition unit 25 stores the browsing start time, the URL, the calculated hash value, and the file storage location in the operation history DB 29 (see FIG. 2) of the storage unit 26 (S45). Note that, as the browsing start time, the time included in the message acquired in S41 is estimated as the time when browsing (display) of the file is started, and is set as the browsing start time. In addition, a full path indicating the storage location of the browsed file stored in the storage unit 26 is set as the file storage location.

  Then, the operation history acquisition unit 25 analyzes the browsed file 27 stored in the storage unit 26, and specifies the title, the author, and the update date of the file (S46). In the illustrated example, the browsed file 27 is an HTML file. Therefore, the operation history acquisition unit 25 acquires the title, author, and update date by analyzing the tag of the THML file.

  Then, the operation history acquisition unit 25 uses the file attribute information DB 28 (see FIG. 3) as the title, author, and update date acquired in S46, and the storage location, hash value, and URL of the file stored in the operation history DB 29 in S45. (S47).

  Then, when the file displayed on the output device is switched to a new file, the operation history acquisition unit 25 estimates the time when the new file is displayed and the time when browsing of the file is finished, and the operation history DB 29 Is stored at the browsing end time (S48). In other words, for example, when the operation history acquisition unit 25 acquires a message notifying the display of a new file from the Web browser 11 or when acquiring a message that makes the Web browser 11 inactive, the operation history acquisition unit 25 includes the message. Is stored in the browsing end time of the operation history DB 29. Then, the operation history acquisition unit 25 returns to S41 (or S42), and performs the same processing by setting the file displayed on the output device as a browsed file.

  Next, similar file search processing will be described.

  FIG. 5 is a process flow diagram of the similar file search process. First, the search management unit 22 refers to an initial setting file (not shown) stored in the storage unit 26, and acquires the application type and score calculation method preset in the initial setting file (S51).

  The application type is set with the type of application for searching for similar files. In the illustrated example, the target application type is the Web browser 11, and a similar file similar to the file currently displayed by the Web browser 11 is searched and displayed. Note that the application type may be all applications, and similar files may be searched for and displayed by files displayed by all applications.

  The score calculation method includes, for example, priority file attributes, maximum browsing time, score distribution, and the like. As the file attribute to be prioritized, an attribute of a file to be prioritized (weighted) is set in the processing of the similarity score calculation unit 23 described later. The maximum viewing time is defined as an upper viewing time in the processing of the experience score calculation unit 24 described later. The score allocation defines a ratio (priority level) between the similarity score and the experience score.

  The search management unit 22 may acquire not only the application type and score calculation method from the initial setting file, but also the application type and score calculation method input by the user using the input device. The input screen when the user inputs is described later.

  Then, the search management unit 22 transmits the identification information of the target application (Web browser 11) to the browsing file acquisition unit 21. Then, the browsing file acquisition unit 21 inquires of the Web browser 11 and acquires a currently displayed (browsing) file (HTML file) and URL. Then, the browsing file acquisition unit 21 transmits the acquired file and URL to the search management unit 22 (S52).

  The browsed file acquisition unit 21 identifies the latest browsed file (that is, the currently displayed file) from the operation history DB 29 and identifies the browsed file identified from the URL of the file and the storage location of the file. 27 may be acquired.

  Then, the search management unit 22 transmits to the similarity score calculation unit 23 the file and URL currently being displayed and the file attributes that are prioritized by the score calculation method acquired in S1. Then, the similarity score calculation unit 23 analyzes the currently displayed file received from the search management unit 22, and specifies the title, the author, and the update date (S53). Specifically, the similarity score calculation unit 23 acquires a title, an author, and an update date by analyzing a tag of the THML file.

  Then, the similarity score calculation unit 23 calculates a similarity score (similarity) with the currently displayed file for each browsed file stored in the file attribute information DB 28 of the storage unit 26 (S54). As for the calculation of the similarity score, for example, there are a method of analyzing the contents of a file and assigning a score to each keyword by text matching, a method of using a Web hyperlink structure, and the like.

  In the present embodiment, the similarity score of the information main body of the browsed file, the similarity score of the author, the similarity score of the title, the similarity score of the update date, and the similarity score of the URL are respectively calculated using the text matching technique. Shall. The similarity score calculation unit 23 accesses the information main body of each browsed file 27 stored in the storage unit 26 using the file storage location of the file attribute information DB 28.

  Then, the similarity score calculation unit 23 weights the predetermined similarity score based on the score calculation method (priority file attribute) transmitted from the search management unit 22 and calculates an integrated similarity score (S55). The priority file attribute of this embodiment is, for example, a title. In this case, the similarity score calculation unit 23 calculates an integrated similarity score by the following formula for each browsed file.

Integrated similarity score = similarity score of information main body of browsed file + author's similarity score + (similarity score of title × 2) + similarity score of update date + similarity score of URL A similarity score list in which the calculated integrated similarity score and the file storage location are paired is transmitted to the search management unit 22.

  Then, the search management unit 22 transmits the maximum browsing time of the score calculation method acquired in S51 to the experience score calculation unit 24. Then, the experience score calculation unit 24 refers to the operation history DB 29 and calculates a browsing time for each browsed file. That is, the experience score calculation unit 24 calculates the browsing time by subtracting the browsing start time from the browsing end time. And the experience score calculation part 24 correct | amends each browsing time calculated for every browsed file with a score calculation method (maximum browsing time), and calculates an experience score (S56).

  Note that the maximum browsing time in this embodiment is, for example, 5 minutes. In this case, the experience score calculation unit 24 rounds the browsing time exceeding 5 minutes to make it uniformly 5 minutes. Then, the experience score calculation unit calculates the experience score of each browsed file as 1 point per second. Then, the experience score calculation unit 24 transmits an experience score list in which the experience score calculated for each browsed file and the storage location of the file are paired to the search management unit 22.

  Then, the search management unit 22 calculates an integrated score using the similarity score list transmitted by the similarity score calculation unit 23 and the experience score list transmitted by the experience score calculation unit 24 (S57). That is, the search management unit 22 matches the similarity score list and the experience score list using the file storage location as a matching key. Then, the search management unit 22 calculates the integrated score by multiplying the score distribution ratio acquired in S51 by the integrated similarity score and the experience score. In the score distribution of the present embodiment, the integrated similarity score is “3” and the experience score is “7”. In this case, the search management unit 22 calculates an integrated score of each browsed file by the following formula.

Integrated score = Integrated similarity score × 0.3 + Experience score × 0.7
Then, the search management unit 22 transmits an integrated score list in which the integrated score calculated for each browsed file and the storage location of the file corresponding to the integrated score are paired to the similar file display unit 12. The similar file display unit 12 displays information on the browsed files in descending order of the integrated score on the output device (S58).

  FIG. 6 is a diagram illustrating an example of an output screen on which similar browsed files are displayed. In the illustrated display screen, a similar file toolbar 61 displayed by the similar file display unit 12 and a currently displayed file (Web page) 62 displayed by the Web browser 11 that is a running application are displayed. Yes. On the similar file toolbar 61, a user input area 63 and a similar file display area 64 are displayed.

  The user input area 63 is an area for inputting an application to be searched for similar files and a score calculation method. In the illustrated example, the target application is the Web browser set in the above-described initial setting file (FIG. 5: S51). The user can change the type of the target application at any time by inputting into a desired check box.

  In the illustrated example, a bar indicating score distribution is displayed as the score calculation method. The user can change the integrated score calculation method (the ratio between the integrated similarity score and the experience score) calculated in S57 of FIG. 5 by moving the bar indicating the score distribution using the input device. . Note that the input area 63 may be configured to allow other score calculation methods (priority file attributes, maximum browsing time, etc.) to be input.

  In the similar file display area 64, a predetermined number of pieces of information regarding each viewed file are displayed in descending order of the integrated score. As information about the browsed file, for example, the title of the browsed file, the summary of the information body of the browsed file, the URL, and the like can be considered. The similar file display unit 12 rearranges the integrated list transmitted from the search management unit 22 using the integrated score as a sort key, and specifies a higher-level browsed file. Then, the similar file display unit 12 accesses the file attribute information DB 28, the operation history DB 29, and the information main body of the browsed file 27 using the file storage location as a search key for each of the identified upper browsed files. The predetermined information is displayed in the similar file display area 64.

  The embodiment of the present invention has been described above.

  In the present embodiment, a file similar to the currently displayed file is searched for and displayed from previously browsed files based on the similarity score of the file and the experience score when the file has been browsed in the past. To do. As a result, similar files that are more useful to the user can be extracted and displayed at the top. In addition, by displaying a useful file at the top, the user can easily access the useful file.

  In this embodiment, the file browsing time is used as an index for calculating the experience score. Thereby, the file browsed for a long time is displayed at a relatively high rank, and the user can easily access a file that has been read carefully (understanding the contents) over a long time in the past.

In the present embodiment, designation of an application to be searched for similar files and a score calculation method are accepted on the display screen shown in FIG. That is, it is possible to accept a score calculation method or the like designated by the user and flexibly change the integrated score calculation method or the like.

  In addition, this invention is not limited to said embodiment, Many deformation | transformation are possible within the range of the summary. For example, in this embodiment, the operation history acquisition unit 25 acquires a user's operation history from various messages. However, the operation history may be acquired by diverting the history acquisition function of the Web browser 11. Also, the operation history may be acquired from a Web server connected to the similar file search system via a network.

  Moreover, in said embodiment, the search object range of the similar file was made into the browsed file browsed in the past. However, the similarity score calculation unit 23 may calculate a similarity score from an arbitrary range of files including files (information) that have not been browsed in the past. That is, the similarity score calculation unit 23 communicates with an external search service to search for a file similar to the currently displayed file in an arbitrary range including files that the user has not browsed in the past. Also good.

  The similar file search system may not include the similarity score calculation unit 23. That is, the search management unit 22 of the similar file search system may receive the similarity score calculation result by transmitting the currently displayed file and URL to another external system.

1 is a configuration diagram of a similar file search system to which an embodiment of the present invention is applied. It is a figure which shows an example of the data structure of file attribute information DB. It is a figure which shows an example of the data structure of operation history DB. It is a processing flow figure of operation history acquisition processing. It is a processing flowchart of a similar file search process. It is a figure which shows an example of a similar file display screen.

Explanation of symbols

  11: Web browser, 12: Similar file display unit, 21: Browsing file acquisition unit, 22: Search management unit, 23: Similar score calculation unit, 24: Experience score calculation unit, 25: Operation history acquisition unit, 26: Storage Part, 27: browsed file, 28: file attribute information DB, 29: operation history DB

Claims (3)

A file search method performed by a file search device,
The file search apparatus includes a history storage unit in which information of browsed files browsed in the past is stored, and a processing unit.
The processor is
A display file acquisition step of acquiring a display file displayed on the display device;
A similarity score calculating step of calculating a similarity score indicating a similarity with the display file for each browsed file with reference to the history storage unit,
An experience score calculating step of referring to the history storage unit and calculating an experience score indicating the strength of the user's experience with respect to the browsed file when viewed in the past for each of the browsed files;
An integrated score calculation step of calculating an integrated score by multiplying a score distribution ratio that defines a ratio between a similar score and an experience score by the similarity score and the experience score calculated for each browsed file,
The integration and score each have rows and outputting step, the outputting the information of the history viewing already file stored in the storage unit corresponding to the total score, the said display device,
In the history storage unit, for each browsed file,
The time when the application that has entered the active state is displayed on the display device is stored as the browsing start time,
The time when the file displayed on the display device is switched from the viewed file to a new file is stored as the browsing end time,
The said experience score calculation step calculates the experience score of the said browsed file according to the browsing time calculated | required from the said browsing start time and the said browsing end time of the said log | history memory | storage part . A file search device in a file search system having a file search device and a display device,
A history storage means for storing information of browsed files browsed in the past;
Display file acquisition means for acquiring a display file displayed on the display means;
Similar score calculation means for referring to the history storage means and calculating a similarity score indicating the similarity to the display file for each browsed file;
Referring to the history storage means, experience score calculation means for calculating an experience score indicating the strength of the user's experience with respect to the browsed file when viewed in the past for each of the browsed files;
An integrated score calculation means for calculating an integrated score by multiplying a score distribution ratio that defines a ratio between the similar score and the experience score by the similarity score and the experience score calculated for each browsed file;
Wherein possess integration and score each and output means for outputting information of the history viewing already files stored in the storage means corresponding to the total score, the said display means, and
In the history storage means, for each browsed file,
The time when the application that has entered the active state is displayed on the display device is stored as the browsing start time,
The time when the file displayed on the display device is switched from the viewed file to a new file is stored as the browsing end time,
The file search device, wherein the experience score calculation means calculates an experience score of the browsed file according to a browse time obtained from the browse start time and the browse end time of the history storage means .   A file search program for causing a computer to function as each means constituting the file search device according to claim 2.

Images