JP3530993B2 - Information processing apparatus and method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information processing device and
Information processing method .

[0002]

2. Description of the Related Art With the rapid spread of computers in recent years, the amount of electronic information handled by people has increased rapidly. In order to manage these electronic information efficiently, attribute information is added to each electronic information. For example, in the field of databases, a search key such as a keyword is added as attribute information to individual data (electronic information) to be managed, and data is searched using this search key. Furthermore, a technique (hereinafter referred to as “fuzzy search”) for searching data by adding similarity to other data to each data as the attribute information and setting a value of this similarity has been proposed. There is.

[0003]

In the above data retrieval, there is a demand to obtain a desired number of data as retrieval results. In such a case, in the above fuzzy search, it is not known how much similarity value should be set to obtain the desired number of data. Therefore, the operator actually searches and the search result (data count It is necessary to repeat the operation of resetting the value of the degree of similarity based on the above) and conducting further searches until the desired number of data is obtained. In order to solve this problem, once a fuzzy search is executed, a document search device that displays a list of the number of data items for each similarity (the number of data items obtained when searching with the value of the similarity) is displayed on the operation screen. It is proposed by JP-A-7-129622. In the device described in the publication, since the value of the similarity used for the next search can be selected by referring to the number of data items for each similarity displayed on the screen, a desired search can be performed with a smaller number of searches. You can get a number of data. However, in the device of the above publication, in order to display the number of data for each similarity, the operator must first set an appropriate similarity and perform a search once, and then refer to the number of displayed data. It was necessary to search again.

The object of the present invention is, in fuzzy search,
An object of the present invention is to provide an information processing device and an information processing method capable of obtaining a desired number of data as a search result by a simple operation. More specifically, it is to provide an information processing apparatus and an information processing method capable of obtaining a desired number of data as a search result only by setting the number of data to be obtained as a search result. .

[0005]

According to the present invention, in an information processing apparatus for storing and managing an object together with its attribute information, a selection receiving means for receiving selection of an object by a user and a predetermined search condition. A search means for searching and extracting other objects from the selected object by referring to the attribute information.
The number a user of the object to be extracted result of the search
A unit for accepting a designation to perform a search, the search condition is appropriately set, and then a process of performing the search by resetting the search condition according to the number of objects extracted by the search is repeatedly performed. Thus, the control means for controlling the search means so as to extract the specified number of objects. In the above information processing device, the attribute information may be a similarity of each object to another object, and the search condition may be a value of the similarity. In this information processing device,
The control means is, for example, extracted by the search.
The number of objects is greater than the number of objects specified above.
If it is larger than the above, increase the similarity value before
The search conditions are reset and the objects extracted by the above search are
The number of objects is greater than the number of specified objects
If it is smaller, the test is performed to reduce the similarity value.
The process of resetting the search conditions and performing the search is repeatedly executed.
Preferably, the control means increases or decreases the value of the similarity.
Make the width smaller each time you search. Also according to the invention
The information processing method is to use an object with its attribute information.
Information processing executed by the information processing device that stores and manages
A method of receiving an object selection by a user
User number of objects extracted as a result of search
Is accepted. Here, based on the predetermined search conditions
Then, by referring to the attribute information, the selected
Other objects from the stored objects,
When extracting, set the search conditions appropriately and perform the search.
Then, of the objects extracted by the search
Depending on the number, the process of resetting the search conditions and performing the search
By repeating the execution, the specified number of objects
Extract the project. The attribute information is, for example,
Is the similarity of the object to other objects,
The search condition is the value of this similarity. This information processing
In the method, in resetting the search condition and searching,
For example, the number of objects extracted by the search
Is greater than the number of specified objects,
Reset the search criteria to increase the similarity value
The number of objects extracted by the search
If it is smaller than the number of specified objects,
Reset the search criteria to reduce the similarity value
The process of searching is repeated. Preferably, the
The range of increase / decrease in the similarity value is reduced each time a search is performed.
Here, the “object” means electronic information that is a subject to be stored and managed by the information processing apparatus, for example, text or image data, or a folder in which these are collected. Further, the “attribute information” is secondary information added to the “object” as the main body, for example, similarity of each object to other objects, search key (keyword, color, shape, etc.), data The creator, creation date, size, comment, icon, etc.

[0006]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of an information processing apparatus of the present invention will be described below with reference to the accompanying drawings. Figure 1
Is an information processing apparatus of the present embodiment (hereinafter referred to as a system)
It is a figure which shows the outline of a structure. This system constitutes a database management system for image data that employs a graphical user interface (hereinafter referred to as GUI). In this system, the similarity to other image data is added as attribute information to the image data as an object, and the image data can be searched based on this similarity. Hereinafter, such a search will be referred to as "fuzzy search". This system is
A central processing unit (hereinafter referred to as a CPU) is provided, and is mainly configured by a control device 1 (computer) that controls the entire system. The display 2 displays an image or characters, and also displays various screens for operation. The keyboard 3 and the mouse 4 perform various inputs, instruction operations, and the like. The floppy disk drive 5b is
Data is written to and read from the floppy disk 5a, which is a medium for storing and storing data. The hard disk device 6 stores and manages image data and its attribute information. The printer 7 forms an image on a sheet based on the image data. The scanner 8 reads image data from a sheet-shaped document. The CD-ROM drive 9b reads image data from the CD-ROM 9a in which a large amount of image data is stored in a predetermined format. The image data read by the scanner 8 and the CD-ROM drive 9b is input to the hard disk device 6. The speaker 10 outputs a voice, and the microphone 11 inputs a voice. Each of these devices is connected to the control device 1 via a connection cable. The arrows shown in the figure show the flow of data in this system. Note that these devices may be built in the control device 1 and integrally formed.

FIG. 2 is a block diagram centering on the controller 1. The control device 1 is, for example, an Intel product number i
A ROM 20 configured mainly by a CPU 201 such as 80486DX and storing a search program according to the present invention via a data bus 220 connected to the CPU 201.
3, a RAM 204 for storing various data, a display control circuit 20 for displaying images or characters on the display 2
5, a keyboard control circuit 206 that controls transfer of input from the keyboard 3, a mouse control circuit 207 that controls transfer of input from the mouse 4, a floppy disk drive 5b
Floppy disk drive control circuit 20 for controlling
8, a hard disk control circuit 209 for controlling the hard disk device 6, a printer control circuit 210 for controlling output to the printer 7, a scanner control circuit 211 for controlling the scanner 8, a CD-ROM for controlling the CD-ROM drive 9b.
The ROM drive control circuit 212, the speaker control unit 213 that controls the audio output to the speaker 10, and the microphone control circuit 214 that controls the microphone 11 are connected to each other. Further, the CPU 201 is connected with a clock 202 for generating a reference clock necessary for operating this system, and further, a data bus 22
An expansion slot 215 for connecting various expansion boards is connected via 0. In addition, S in the expansion slot 215
A CSI board may be connected, and the floppy disk drive 5b, the hard disk device 6, the scanner 8, the CD-ROM drive 9b, etc. may be connected via this SCSI board. In this system, the floppy disk 5a and the hard disk device 6 are used as a medium for storing data, but another storage medium such as a magneto-optical disk may be used. Further, although the scanner 8 and the CD-ROM drive 9b are used as the image data input device, other input devices such as a still video camera and a digital camera may be used. Although the printer 7 is used as the output device, other output device such as a digital copying machine may be used. Further, in this system, the search program according to the present invention is stored in the ROM 203. However, copy all or part of this program to floppy disk 5a.
Alternatively, the program and data may be stored in a data storage medium such as a hard disk device 6 or the like, the program and the data may be read from the data storage medium into the RAM 204 and executed as necessary. In the present system, the image data input to the hard disk device 6 from the external input device is stored and managed, but the image data stored in the CD-ROM 9a may be directly managed. Conceivable. In this case, the index data indicating the image data stored in the CD-ROM 9a is stored in the hard disk device 6 with the similarity as attribute information added (the set of this index data and attribute information is called a record). . At the time of searching for image data, a record is first extracted as a search result based on the value of the set similarity, and then the image data main body is read from the CD-ROM 9a based on the index data of the extracted record. do it. As the management target (object) of this system, in addition to still image data and moving image data, character data, voice data, multimedia data which is a combination of these data, and the like can be considered. Furthermore, a folder that collects these data may be an object, or each computer itself may be treated as an object when a network is built.

FIG. 3 is a diagram showing a setting screen displayed on the display 2 as the program is executed. Windows 20, 21, 22 are displayed on the display 2. In the window 20, a list of image data stored and stored in the hard disk device 6 is displayed.
When performing a fuzzy search, one of these image data is selected as a key image by operating the keyboard 3 or the mouse 4. The selected image data is displayed with its peripheral portion highlighted. In the window 21, select the process to be executed,
Display the menu screen to instruct. On the menu screen,
A well-known fuzzy search for performing a search by setting a similarity value in the menu 23 of "data registration" that executes a process of registering new image data input from the scanner 8 or the CD-ROM drive 9b in the image database A “similarity setting” menu 24, a “search number setting” menu 25 for performing a search by setting the number of image data obtained as a result of search, and other menus (not shown) are displayed. The window 22 displays the image data searched by the fuzzy search process, the number of image data, and the similarity at the time of the search.

FIG. 4 is an example of a table showing the degree of similarity between the key image and other image data when one of the image data registered in the image database of the hard disk device 6 is used as the key image. The degree of similarity is represented by a value of 0 to 1. As for the value of the similarity, 1 represents the same, the similarity decreases as the value decreases, and the similarity of 0 represents no similarity. The similarity may be expressed as a percentage, such as 0% to 100%. After setting the key image from the image data displayed in the list in the window 20, select the menu 24 of “similarity setting” and set the similarity to 1
When a fuzzy search is performed by setting to, a matching search is performed and only the key image is searched. Also, in the same way,
When the fuzzy search is performed with the similarity set to 0.70, five image data of image data A, image data B, image data C, and image data D are searched in addition to the key image. In this case, as shown in FIG.
Displays the image data obtained as a result of the search, the number of search results, and the degree of similarity at the time of the search. It should be noted that in the present system, in addition to setting the degree of similarity as described above and performing the search, it is possible to set and search the number of image data obtained as a result of the search. Such a search can be performed by selecting the key image and then selecting the “search number setting” menu 25 and setting a desired search number. In this case, the same number of image data as the set search number are searched. For example, if the number of searches is set to 1, a match search is performed, and only key images are searched. Further, in the definition shown in FIG. 4, when the number of searches is set to 5 and a search is performed, five image data A, image data B, image data C, and image data D are displayed in addition to the key image. To be searched. At this time, as shown in FIG. 6, the window 22 displays the image data obtained as a result of the search, the number of search results, and the similarity at the time of the search. Regarding the search process by setting the number of searches,
The details will be described later.

FIG. 7 is a flowchart of a main routine of control processing executed by the CPU 201 based on a program stored in the ROM 203. When the system is powered on and the program is started, first, initialization processing such as initialization of variables necessary for each processing and display of an initialization screen (see FIG. 3) on the display 2 is performed (step S1). Next, it is determined whether the menu in the window 21 is selected (step S2). here,
When the "Data Registration" menu 23 is selected,
While the image data is taken in from the image input device such as the scanner 8 or the CD-ROM drive 9b, the data registration process for adding the similarity to other image data as attribute information is performed (step S3), and other processes (step S6). )
Go to. When the “similarity setting” menu 24 or the “search number setting” menu 25 related to data search is selected, the image data (key image) selected from the image data displayed in the window 20 as a reference is used as a reference. A search process for searching image data is executed (step S
4). Then, after displaying the search result in the window 22 on the display 2 (step S5), the process proceeds to other processing (step S6). If no menu is selected, the process immediately proceeds to other processing (step S8). When the other processes in step S8 are completed, the process returns to step S2, and the same processes are repeated thereafter. The data registration process (step S
3) and the search process (step S4) will be described later.

FIG. 8 is a flow chart of the data registration process (FIG. 7, step S3). Here, the scanner 8
Image data newly input from an image input device such as the CD-ROM 9a by the CD-ROM drive 9b.
The image data read from the image data to which the degree of similarity with other image data is added as attribute information is stored in the hard disk device 6 as a record of the image database.
First, image data to be newly registered in the hard disk device 6 is selected from the scanner 8 or the CD-ROM drive 9b (step S31). Next, the selected image data,
A process of setting the value of the similarity with other image data already registered in the hard disk device 6 is performed (step S
32). After that, the selected image data is stored in the hard disk device 6 together with the data of the set similarity (step S33). In the similarity setting process of step S32, it is a difficult task to successively set the similarity between the image data newly registered by the operator and each existing image data. Therefore, normally, when the operator sets the degree of similarity with one or several existing image data, the degree of similarity with the remaining existing image data is automatically set according to a predetermined rule. Has become. In addition, the characteristics of the image data to be registered may be automatically extracted, and the degree of similarity to the existing image data may be automatically set according to a predetermined rule.

FIG. 9 shows the retrieval process (FIG. 7, step S4).
It is a flowchart of. First, from the image data displayed in a list in the window 20, the image data that is the source of the search is selected and designated as a key image (step S4).
1). Next, it is determined which of the "similarity setting" menu 24 and the "search number setting" menu 25 has been selected (step S42). Menu 25 for "Set number of searches"
If is selected (YES in step S42)
S) Then, the operator operates the keyboard 3 or the mouse 4 to input the number of image data to be obtained by the search (step S43). The search is performed while adjusting the similarity so that the input number of image data are searched (step S44). This process will be described later. On the other hand, when the "similarity setting" menu 24 is selected (NO in step S42), the operator operates the keyboard 3 or the mouse 4 to set the similarity (step S45). The search is executed with the set similarity (step S46). The search here is a well-known search for image data in which the similarity to the key image is set to a value equal to or higher than the set similarity, and detailed description thereof will be omitted.

FIG. 10 is a flowchart of the search process (FIG. 9, step S44) performed while adjusting the degree of similarity. The number of data obtained as a result of the search cannot be known unless the search is actually performed. Therefore, the similarity is increased or decreased according to the number of previous search results, and the search is executed again using the increased or decreased similarity. By repeating the above, a desired number of data is obtained. First, as an initial setting, the variable i is set to 1, and the intermediate value between the upper limit and the lower limit of the similarity is set to Δ (step S441). The similarity V ₁ is set to Δ (step S442). Set similarity V
_A fuzzy search is executed with _i (the initial value is i = 1) (step S443). As a result, if the number of retrieved data H _i is the same as the number of previously retrieved data (step S44)
If YES, the process ends. On the other hand, if the number of retrieved data H _i is different from the previous retrieval result (step S
444, NO), this number H _i and the number of target data,
That is, the number of data input in step S43 (hereinafter,
This number is represented by T) and is compared (step S44).
5). Here, when H _i is equal to T (step S44)
If YES, the process ends. If H _i and T do not match (NO in step S445), the subsequent similarity V _{i + 1}
Is determined as follows. If the number H _{i of} search results is larger than the target number T (YES in step S446), the similarity V _{i + 1} used in the next search is set to V _i + Δ × 1/2 ⁱ (step S447). If the number of search results H _i is less than the target number T (NO in step S446), the similarity V _{i + 1} used in the next search is set to V _i −Δ × 1/2 ⁱ (step S448). ). After adding 1 to the value of the variable i (step S449), the process returns to step S443.
That is, in the search processing performed while adjusting the above-described similarity, first, using a predetermined similarity (here, an intermediate value between the upper limit value and the lower limit value of the similarity), a normal fuzzy search, that is,
The search is executed by the same method as the search performed by setting the similarity in steps S45 and S46 (step S4).
41-S443). Next, it is determined whether the number of data obtained as a result of this search is larger or smaller than the desired number of data (step S446), and the degree of similarity is increased or decreased based on the result of this comparison (steps S447 and S448). ), And further search is performed (steps S449 and S443). It should be noted that the range of increase or decrease in the degree of similarity is halved each time a search is performed. In the above process, the number of data obtained as a result of the search matches the desired number of data (YES in step S445).
Repeat until S). However, if the number of data obtained as a result of the search is the same as the result of the previous search,
When it is determined that the adjustment range of the similarity has reached the convergence, the process ends (YES in step S444). The processing may be terminated when the number of data obtained as a result of the search falls within a predetermined error range with respect to the desired number of data. After the search is completed, the above-mentioned search result is displayed (see FIG. 6) (step S5). FIG. 11 is a table showing how the similarity converges to a value at which a desired number of data are obtained in the search process (step S44) performed while adjusting the similarity. Hereinafter, the process until the target search number T is obtained will be described using specific numerical values with reference to the flowchart of FIG. 10 described above. The horizontal axis of the graph represents the degree of similarity and the vertical axis represents the number of search results. The intermediate value Δ between the upper limit value and the lower limit value of the similarity is 0.5.
Is. First, the first fuzzy search is performed with the similarity V ₁ = 0.5 (step S443). The number of data H ₁ obtained by this search was not sufficient for the target number of searches T (NO in step S446). Therefore, V ₁ = 0.5
From which Δ / 2 ¹ = 0.25 is subtracted V ₂ = 0.25
The second ambiguous search is performed with (steps S448, S
449, S443). The number H ₂ of data obtained by the _second search has exceeded the target number T of searches this time (YES in step S446). Therefore, V ₂
= 0.25 plus Δ / 2 ² = 0.125 Similarity V ₃ =
Perform a third fuzzy search at 0.375 (step S
447, S449, S443). As a result, the number H ₃ of retrieved data matches the target number T of retrievals (YES in step S445), and the retrieval process is ended.

[0014]

As described above, according to the information processing apparatus and the information processing method of the present invention, the fuzzy search is repeatedly executed until the designated number of data (objects) is obtained while adjusting the value of the degree of similarity. To do. Therefore, the operator can obtain a desired number of data as a search result by a simple operation of designating the number of data.

[Brief description of drawings]

FIG. 1 is a diagram showing an outline of a configuration of an information processing apparatus of this embodiment.

FIG. 2 is a block diagram centering on a control device.

FIG. 3 is a diagram showing a setting screen displayed on the display as the program is executed.

FIG. 4 is a diagram showing an example of a degree of similarity added between other image data when one of the image data stored in the image database of the hard disk device is used as a key image.

FIG. 5 is a diagram showing an example of a result of fuzzy search with the similarity set to 0.7.

FIG. 6 is a diagram showing an example of a result of performing a search with the number of image data obtained as a result of the search set to 5.

FIG. 7 is a flowchart of a main routine of control processing executed by a CPU.

FIG. 8 is a flowchart of a data registration process.

FIG. 9 is a flowchart of a search process.

FIG. 10 is a flowchart of a search process performed while adjusting the degree of similarity.

FIG. 11 is a diagram showing how the similarity actually converges through a search process performed while adjusting the similarity.

[Explanation of symbols]

1 ... Control device 2 ... Display 3 ... Keyboard 4 ... mouse 5a ... Floppy disk 5b ... Floppy disk drive 6 ... Hard disk device 7 ... Printer 8 ... Scanner 9a ... CD-ROM 9b ... CD-ROM drive 10 ... speaker 11 ... Microphone 201 ... CPU 202 ... Data bus 203 ... ROM 204 ... RAM

─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-8-287103 (JP, A) JP-A-8-314950 (JP, A) JP-A-8-171572 (JP, A) JP-A-11- 53366 (JP, A) JP-A-4-333168 (JP, A) JP-A-7-334512 (JP, A) Daigo Motomasa, Nobuyuki Nakajima, Yuko Hanawa, Support for product purchase incorporating automatic change of search conditions System construction, Journal of Japan Fuzzy Society, April 15, 1995, Volume 7, Issue 2, p. 380-389 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06F 17/30

Claims (8)

(57) [Claims] 1. An information processing apparatus for storing and managing an object together with its attribute information, wherein the selection information receiving means receives a selection of an object by a user, and refers to the attribute information based on a predetermined search condition. Search means for searching and extracting other objects from the selected object, and the number of objects extracted as a result of the search to the user
And means for accepting the designation by performing a search by appropriately setting the search condition, and then repeatedly performing a process of resetting the search condition and performing a search according to the number of objects extracted by the search. An information processing apparatus having: a control unit that controls the search unit so as to extract the specified number of objects. 2. The attribute information is the degree of similarity of each object to other objects, and the search condition is
The information processing apparatus according to claim 1, wherein the information processing apparatus is a value of the similarity. 3. The control means is extracted by the search.
The number of objects
If it is greater than the number, try to increase the similarity value
The search conditions are reset to and extracted by the search
The number of objects is greater than the number of objects specified above.
If it is smaller than the above, decrease the similarity value before
Repeat the process of resetting the search conditions and performing the search.
The information processing device according to claim 2. 4. The control means increases or decreases the value of the similarity.
4. The information according to claim 3, wherein the width of is reduced each time a search is performed.
Information processing device. 5. An object is stored together with its attribute information.
Information processing method executed by the information processing device
Method, which accepts the selection of objects by the user and determines the number of objects extracted as a result of the search by the user.
Accept the designation and refer to the attribute information based on predetermined search conditions
To select other objects from the selected object.
When searching for and extracting objects
Set and search, then extract
Re-set the search condition according to the number of objects
By repeatedly executing the search process,
An information processing method for extracting a fixed number of objects. 6. The attribute information is for other objects of each object.
The degree of similarity to an object, and the search condition is
The information processing method according to claim 5, wherein the similarity value is a value. 7. In resetting and controlling the search condition,
The number of objects extracted by the search is the specified
If the number is greater than the number of objects
Reset the search criteria to increase the value of
The number of objects extracted by the search is specified above
If the number is less than the number of
The search conditions are set again to reduce the
The information processing method according to claim 6, wherein the processing is repeatedly executed.
Law. 8. The range of increase or decrease of the value of the similarity is searched.
The information processing method according to claim 7, wherein the information processing method is reduced for each.