JP2003256446A - Data management system, data management method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To simply and efficiently give information for retrieval to multimedia contents by anyone and to prevent the meta data quantity from being increased to no purpose. <P>SOLUTION: When a plurality of data are managed by individually giving meta data used for retrieving for the data thereto, an input sheet carrying predetermined meta data candidates is prepared, a user is caused to select the meta data from the meta data candidates, and the selected meta data is saved in connection with the data. When the existing meta data (b) and (c) given in the past are in the data to which the meta data is given, the data (b) and (c) are synthesized with the new meta data (a) to be given, and the synthesized meta data (d) is given to prevent repetition of the meta data. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data management technique for data retrieval.

[0002]

2. Description of the Related Art Due to the improvement in capacity and cost reduction of small computer systems, home electric appliances are generally equipped with a computer for controlling and processing the information. In addition to video and music supplied as digital data, high-definition contents such as video and audio have been supplied, and multimedia contents in which these are mixed have been disseminated.

Further, due to the spread of the Internet, which is a global computer network, and the cost reduction of mobile communication devices in recent years, it has become an era where these contents can be easily obtained and enjoyed.

[0004] Household video equipment also changes from recording analog broadcasts and enjoying video and music supplied by media to equipment that records moving pictures and audio as digital data that does not deteriorate with high quality. Along with
Video cameras that can be purchased at ordinary homes have emerged due to small and inexpensive video recorders, and the time has come to enjoy watching and watching video at home.

Under such a background, the amount of contents to be video-recorded and edited, which has been conventionally performed by broadcasting companies, has been increasing.

[0006] Here, broadcasting companies and the like have advanced equipment for shooting and editing, and engineers with advanced technology for editing and configuring to make the contents easy to understand. On the other hand, editing the contents shot with a home-use video camera is not often performed due to the price of the device and the need for editing technology.

However, not only moving images taken at home, but also various media and communication means such as the Internet increase the amount of information and contents delivered to the home, which results in multi-media such as moving images and voice. The information has been changed to media information and contents, and the information search function for such information and contents has become very important.

[0008] In addition, due to the expansion of the infrastructure for distributing and transmitting information, including the Internet, the time has come for individuals to transmit information, but the transmission of information centered on video and audio is beginning. . It is expected that the need for such functions as multimedia information and content information retrieval on the Internet will increase greatly in the future. Further, it is desirable that the search function be easy to handle, and at the same time that various retrieved results should be easy to understand.

Furthermore, since it has become easy to deliver the latest information and contents in real time by always using the Internet or by using mobile communication, even broadcasting companies can efficiently obtain the latest images. Then
It is very important to edit, and under such circumstances, a function of quickly searching for existing information or content and reflecting it in new information or content is required.

[0010] However, data such as a moving image, which has a large amount of information and changes with time, is very bulky to store, and it is very difficult to easily search for the contents. was difficult. For the same digital data, if it is document data, etc., it is not difficult for a computer to find the part that contains a character or word by pointing to it, but in the case of video or audio data, it is a perfect match. Is rarely presented and searched, and it is common to search for parts that are similar in some sense.

Further, it is not sufficient in the current technology for a computer to understand the contents expressed by video and audio, and the contents of such data can be searched for in the same manner. Still not enough.

Therefore, with respect to such moving images, still images, sounds, etc., the contents and characteristics are described as linguistic names such as keywords and explanations, and the characteristics of images and sounds in a format easy to process by a computer are described. A method has been started in which such non-linguistic information is added to the contents and information as information for search, that is, so-called metadata, to facilitate the search and use.

[0013]

However, it is not always easy to accurately and efficiently assign this metadata to the managed data. In particular, it is more difficult for an ordinary person to add metadata to an image captured by a home video camera or the like due to lack of specialized knowledge and technology.

On the other hand, such metadata may be added to the same data a plurality of times, or depending on the data, the metadata may already be added. Therefore, the metadata having the same content may be added in duplicate, and the amount of metadata may unnecessarily increase.

Therefore, a main object of the present invention is to allow information for retrieval to be added to managed data easily and efficiently by any person, and to prevent the amount of metadata from being unnecessarily increased. is there.

[0016]

According to the present invention, there is provided a data management system which manages data by individually adding metadata for data retrieval to data to be managed,
Means for accepting the selection of the data to which the metadata is given, means for providing one or more candidates consisting of a predetermined single or a plurality of collections of metadata, and a means for providing the metadata candidates From among, means for accepting a selection of metadata to be added to the selected data, metadata already added to the selected data, and a synthesizing means for synthesizing the selected metadata, There is provided a data management system comprising: a means for associating and storing the selected data and the metadata synthesized by the synthesizing means.

Further, according to the present invention, there is provided a data management method for individually managing data by adding metadata for data search to the data to be managed, and selecting the data to which the metadata is added. And a step of providing one or a plurality of candidates consisting of a predetermined single or a collection of a plurality of metadata, and a step of providing the selected data from the provided metadata candidates. A step of accepting a selection of metadata, a combining step of combining the metadata already added to the selected data and the selected metadata, the selected data, and the combining step And a step of associating and storing the metadata synthesized in
There is provided a data management method comprising:

Further, according to the present invention, in order to individually add the metadata for data retrieval to the data to be managed and manage the data, the computer is required to select the data to which the metadata is added. A step of accepting, a step of providing one or a plurality of candidates composed of a predetermined single or a plurality of sets of metadata, and a step of assigning to the data selected from the provided candidates of the metadata A step of receiving a selection of metadata, a combining step of combining the metadata already added to the selected data and the selected metadata, the selected data, and the combining step, Storing the synthesized metadata in association with each other,
A program is provided to execute.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a configuration example of a computer device in which a data management system according to an embodiment of the present invention is constructed.

In FIG. 1, a CPU is a microprocessor, which performs arithmetic operations for various processes, logical judgments, etc., and is connected to these buses via an address bus AB, a control bus CB, and a data bus DB. Control components. The details of the work are ROM and RA described later.
Directed by the program on M.

A plurality of computer programs can be operated in parallel by the function of the CPU itself and the mechanism of the computer programs.

The address bus AB transfers an address signal designating a component to be controlled by the CPU. The control bus CB transfers and applies a control signal of each constituent element to be controlled by the microprocessor CPU. The data bus DB transfers data between the respective constituent devices.

The ROM is a read-only fixed memory. The ROM includes a microprocessor CPU described later.
A computer program area and a data area that store the control procedure of are stored.

The RAM is a writable random access memory, and is a microprocessor C described later.
It is also used as a computer program area and a data area in which control procedures by the PU are stored, and also as a temporary storage area for various computer programs and various data from each component other than the CPU.

Storage media such as ROM and RAM store computer programs and data for realizing the data management of this embodiment. The CPU reads and executes the program code stored in these storage media. By doing so, the function is realized. The type of storage medium does not matter.

A storage medium storing the program and data according to the present invention is supplied to a system or apparatus,
The program may be copied onto the RAM from a rewritable storage medium such as a RAM. The storage medium may be a CD-ROM, a floppy (registered trademark) disk, or a CD-ROM. A ROM, a hard disk, a memory card, a magneto-optical disk, etc. can be used, and such a system can also be adopted.

Further, as shown in FIG. 2, it is naturally possible to read the data management program of this embodiment from the storage medium storing the program. In addition, in FIG.
21 is a computer device having the configuration shown in FIG.
Reference numeral 22 is a drive device for reading a storage medium, and 23 is a storage medium storing a program.

Returning to FIG. 1, the DISK is a hard disk and functions as an external memory for storing various data and the like. DISK has a built-in storage medium that can read and write a large amount of information at a relatively high speed, and various data and the like can be stored and retrieved as necessary. In addition, the stored computer programs and data are instructed by the keyboard and various computer programs,
It is called completely or partially in RAM when needed.

A MemCard (memory card) is a removable storage medium. Information is stored in this storage medium, and by connecting the storage medium to another device, the stored information can be referenced / transferred. Will be possible.

The KB is a keyboard provided with various function keys such as an alphabet key, a hiragana key, a katakana key, a character / symbol input key such as a punctuation mark, and a cursor movement key for instructing cursor movement. In addition,
A pointing device such as a mouse can also be included.

CR is a cursor register. The CPU can read and write the contents of the cursor register. A CRT controller CRTC, which will be described later, displays a cursor at a position on the display device CRT for the address stored here.

DBUF is a buffer memory for display and stores a pattern of data to be displayed. The CRTC plays a role of displaying the contents stored in the cursor register CR and the buffer DBUF on the display CRT.

The CRT is a display device using a cathode ray tube or the like, and the display pattern of the dot configuration and the display of the cursor on the display device CRT are controlled by the CRT controller. Further, CG is a character generator, which stores patterns of characters and symbols to be displayed on the display device CRT.

The NCU is a communication device for communicating with other computer devices and the like, and by using this, the program and data of this embodiment can be shared with other devices. In FIG. 1, the network (L
A personal computer (PC) via the AN), receiving / storing / displaying devices such as TV broadcasts and videos taken by yourself (TV)
/ VR), a home-use game computer (GC), etc., and information can be freely exchanged with them. Any device may be connected to the system of this embodiment via a network. Further, the type of network does not matter, and the network may not be a closed network as shown in the figure but may be one connected to an external network.

The DTU is a receiving device that realizes the receiving function of broadcast communication using artificial satellites, etc., and receives radio waves broadcast via artificial satellites with a parabolic antenna (ANT), It has the function of extracting the data being broadcast. There are various forms of broadcast communication, such as those broadcast by terrestrial radio waves, those broadcast by coaxial cables, optical cables, etc., those delivered by the LAN or large-scale networks, etc. Although various forms are possible, any form of broadcast communication can be adopted.

<Concept of data management in this embodiment>
Next, the concept of data management in the data management system of this embodiment will be described.

Generally, in data management, the contents of data to be managed may be various, but the contents of data to be handled are predictable to some extent or limited to a certain range. For example, when managing a file of moving image data, the contents include the images of daily real-world events such as ceremonial occasions, children's growth records, images of travel destinations, and images of dramas and news. And can be appropriately classified.

Then, depending on the type of content of each data,
The specific contents can be predicted to some extent empirically. For example, if it's a video of a wedding reception,
It will include videos of individual scenes such as a bride and groom's entrance, guest speeches, toasts, entertainment, etc.In addition, in the case of criminal drama videos, scenes of incident occurrence, investigation, resolution, etc. Is included. Furthermore, the persons, objects, situations, etc. appearing in each scene are also expected to some extent.

As a result, it can be said that the contents of the metadata to be added to these data naturally have common factors depending on the type of the contents of the data.

Therefore, in the present embodiment, a candidate to be given as metadata is created in advance according to the type of the content of the data, and this is presented to the user at the time of data registration so that the user can select the metadata. To be easily and efficiently realized. In addition, we will not realize efficient data retrieval by handling the given metadata by a specific method.

Further, in this embodiment, in order to create a candidate of metadata, etc., it is first classified according to the type of the data to be managed. In this embodiment,
Classify daily events as a unit. For example, wedding receptions and overseas trips can be cited as events.

As a rule of thumb, each event can be expected to be composed of a plurality of events. For example, a wedding reception event consists of small individual events such as bride and groom admission, guest speeches, and toasts. Therefore, in the present embodiment, each event is defined from a plurality of events, and metadata candidates are created for each event.

FIG. 3 is an example of information defining a relationship between events, and is a diagram showing an inclusion relationship of individual events XA to I included in a certain event X. As is apparent from the figure, the event X first includes the events X-A to C, and the event X-A further includes the events X-D and E.

For example, if the event X is a wedding reception, the events XA to C are set to a banquet, a meal, a climax, and the events XD and E are subordinate events of the event XA. Can be used as a bride and groom admission, guest speeches, etc.

Candidates for metadata are objects such as persons, objects, etc. that appear empirically in each event, or situations, and the more subdivided the event, the easier and more accurate the setting will be. .

FIG. 4 is a diagram showing an example of a type of each event and a table of metadata candidates set for each event. In FIG. 4, there is a wedding reception as an event, and as the events below it, there is a bride and groom entrance, speech, and entertainment. Further, as an object or the like appearing at a wedding reception, the bride and groom are empirically ruled out, and thus they are candidates for metadata.
Also, as a situation when the bride and groom enter, a spotlight can be considered as an empirical rule, so it can be seen that this is listed as a candidate for metadata. This information can be stored in DISK or the like in FIG. This embodiment is an example suitable for classifying the contents of the home video contents and the like and adding metadata by using the event as a unit. But,
The present invention can cope with contents other than video by facilitating the use by selecting a unit as a reference according to the contents.

Next, in the present embodiment, the type of event and the candidate of metadata become the metadata finally attached to each data and are utilized as index information at the time of retrieval. In this case, in order to improve the efficiency and accuracy of the data search, it is possible to predefine the relationship between the events or the candidates of the metadata, and this information can be stored in the DISK or the like in FIG.

FIG. 5 is a diagram showing an example of a template defining the correlation between such events and metadata candidates. One template shows one event, individual events included in that event,
Objects as candidates for metadata and the relationships between them are shown.

There are a plurality of events E3-X (X is a number) in the event template T3-1. This is 1
This is because one event consists of small events that have time and causal relationships. In addition, there is an object such as a person or thing related to the event, and in the figure, Obj
It is represented by 3-X (X is a number).

There are objects that appear only in individual events and objects that commonly appear in all events. In the figure, only Obj3-1 and Obj3-2 commonly appear. , Not in a specific event, but only in the frame of the event template T3-1.

On the other hand, the other objects are
It is surrounded by individual events and is shown to occur only within that event. There is some kind of relationship between each object and some kind of relationship between events. This is represented by R3-X (X is a number). Events, objects and their relationships in this event template T3-1 can have various attributes.

For example, if this is an event template for a wedding reception, Obj3-1 having the attribute of groom and Obj having the attribute of bride.
3-2 is an object that appears in most of the events.

Groom's friendship Ob3-11 with the attribute of moderator in R3-7 declares the beginning of the reception, and groom Obj3-1 and bride Obj3-2 come into the venue. For example, the event is E3-1.

In this way, the relationship between each event and each metadata candidate is defined in advance, and the information is used at the time of data search.

<Metadata Addition> Next, the metadata addition processing performed by the data management system in this embodiment will be described based on such a basic concept.

In the present embodiment, when managing various data such as multimedia information such as moving images, still images, voices, and the like, metadata is added in order to facilitate the search of these data. When assigning data, the user must select the target data, and then the user must be presented with the metadata input sheet in which the metadata candidates for each event described above were presented to the user. Done by. The user can select one or a plurality of candidates composed of a predetermined single or a plurality of sets of metadata presented on the input sheet. The selection of the input sheet for each event may be automatic selection or may be selected by the user.

As described above, by providing the metadata candidates, even a general user who does not have specialized knowledge about metadata addition or data search technology can easily and efficiently add metadata. Become. Further, even a user who has specialized knowledge can realize the efficiency of metadata addition, further achieve the unification and rationalization of the metadata expression format, and the efficiency at the time of search.

The input sheet is, for example, DISK of FIG.
It can be stored in a storage device such as ROM, RAM, MemCard. Also, obtain it from a LAN, etc.
It can also be used by storing it in a storage device.

The metadata adding process in this data management system will be described below with reference to a specific example. Figure 6
3 is a flowchart showing a flow of such processing.

In step S1, the data management system accepts selection of data and selection of input items on the input sheet. The data and the input items of the input sheet are displayed on the CRT, for example, in the form of a list, and the user can select this from the KB. When the input sheet and data are selected, the data management system displays them on the CRT.

The data can be selected in units of some data (referred to as sub-data in the present embodiment) which constitutes each data. As the sub data, for example, when the data is moving image data,
The image of each frame constituting the moving image or the data of a continuous frame image group can be cited. In short, the sub data can correspond to the individual events that make up each event.

In step S2, the data management system receives the input of the metadata for the input sheet. The metadata can be input by the user from the KB according to the input items listed on the input sheet. Input here includes not only inputting character strings such as YES / NO or specific keywords in the presented input items, but also inputting one of the given options. It is a concept.

The processes of steps S1 and S2 will be described below with reference to specific examples. Figure 7 shows C when inputting metadata
It is the figure which showed the example of a display of RT. This is an example of a screen when a window system is used. Of the moving image data obtained by shooting a wedding reception, some of the frame images or sub-data indicating a continuous frame image group is displayed. On the other hand, it is supposed to add metadata.

Reference numeral 71 denotes a title bar, which is a portion for operating the entire window, for example, moving and changing the size. Reference numeral 72 is a portion called a toolbar, which allows the main functions in this window to be easily called. "Open" in the toolbar 72
When the button is selected, a list of managed data appears, and the user can select the data to which the metadata is added.

Next, a rectangular area 73 surrounded by a thick line is an area for displaying selected data, and an inverted L-shaped area 74 surrounded by a thick line is an area showing an input sheet. In this embodiment, the image related to the data to which the metadata is added and the input sheet are displayed together, and the image and the candidate of the metadata are displayed together.

The upper part 731 of the area 73 shows the file name of the selected moving image data, and work on one scene in the moving image file "Yamada's wedding.mpg" recorded by digital video. I'm going.

By operating the slider 731a using the KB shown in FIG. 1, it is possible to select an arbitrary scene in the moving image, and the area 431b is divided into five frames included in the currently selected scene. The image is shown. The user may be allowed to select one frame to which the metadata is added or a series of frames within a certain time period.

In the present embodiment, metadata is added to search for data such as multimedia data. However, a part of a device for editing, processing or reproducing multimedia data may be used. It may exist as a function, and depending on the result of editing, etc., there may be a case where the function of this part is unnecessary or a more advanced function is provided.

The portion 732 below the region 73 is the region 73.
This is a portion for reproducing and viewing the scene selected from the frame image of 1b, and is an image (bold line frame) of sub data to which metadata is to be added. Also, if there is information such as the time when this scene was shot, it will be displayed in this (right part). Using this part, the user can grasp the contents of the scene selected by the user. As the sub-data, the data of a single frame image can be used as one unit, and the data of a group of frame images forming one scene can also be used as one unit.

Next, an area 74 indicates an input sheet, and is an area for adding metadata to an image (sub-data) shown in a portion 732 below the area 73.

The combo box 741 is for selecting the event type of the highest concept, and can be selected from a plurality of types of events in a pull-down menu format. By selecting this event, the area 74
Other input items can be selected. In FIG. 7, the event "wedding reception" is being selected.

The area 742 is for selecting an event which is a subordinate concept of the event selected in the combo box 741, and a plurality of events set in advance as an event constituting the event of "wedding reception" are selected. It is listed. The user can arbitrarily select an event corresponding to the content of the selected sub data, and by selecting the event, other input items in the area 74 will be further selected. In FIG. 7, the event “bride and groom to be lighted up” is being selected.

The area 743 is an area in which the input items determined in accordance with the metadata candidates set for the event selected in the area 742 are displayed. In Figure 7,
In the area 742, the names of the objects that are considered to appear in the selected event are listed. By selecting / deselecting these, the user can instruct whether or not the object with that name appears in the image of the currently selected sub-data, and the user can perform a simple operation to display the person or object that appears. Metadata can be added. Note that in FIG. 7, “groom” and “bride” have appeared in the area 732, and the user has shown a state in which “cake” is not shown in particular.

When the "detailed information" button in the area 743 is pressed, more detailed information can be added to each object. The contents that can be given as this detailed information include addition / deletion / modification of various attributes to individual objects, addition / deletion / modification of characters and objects that were not specified as input items, etc. The user can arbitrarily edit and set data candidates.

Similarly to the area 743, the area 744 also includes the area 7
This is an area in which the input items determined in accordance with the metadata candidates set for the event selected in 42 are displayed. In FIG. 7, items are listed for relationships between objects, situations, and the like. By selecting / deselecting these, the user can indicate whether the relationship or situation exists in the currently selected scene, which allows the user to perform simple operations.
It is possible to add metadata such as the relationship and state of the people and things that have appeared. In FIG. 7, "laughter" is not included, but the user has instructed that "face up", "camera flash", or "applause" has occurred. Further, by pressing the "detailed information" button in this, more detailed contents can be added to this relationship and state.

The contents that can be given as this detailed information are
Addition / deletion / modification of attributes of relationships and states, setting / modification / deletion of objects such as actors (objects) and target persons (objects) of relationships, relationships and states not found in events and event templates Addition / deletion / change etc. can be mentioned, and the user can arbitrarily edit and set the candidate of the metadata.

The area 745 is an input item to which a keyword or description is arbitrarily added to the currently selected sub data. In the figure, some explanatory texts have been added, and the explanatory text "Bride about to fall" is selected. Here, by pressing the "Delete" button, the selected explanation can be deleted. You can also add a new keyword or description by pressing the "Add" button. It should be noted that these keywords and explanations may be arbitrarily input by the user, or appropriate keywords and explanations are set in advance as metadata candidates on the system side and the user is allowed to select them. You may do it.

The area 75 is a portion for finally instructing the addition of the metadata input in the area 74 to the currently selected sub data. When the "register" button is pressed here, The contents input in the area 74 indicating the input sheet are determined as the metadata, and the process proceeds to a process of storing the metadata. If the "Cancel" button is pressed, the contents input in the area 44 showing the input sheet will be discarded.

In this way, the user can input the metadata for each scene of the moving image.
Note that the example of FIG. 7 is an example in which metadata is added to one scene of a moving image, but the data to be added is not limited to this, and metadata can be added to various multimedia contents by the same procedure. It is possible to

In the present embodiment, the user is allowed to select metadata in principle. However, for some metadata, the system side is set at the stage when the type of input sheet (type of event) is specified. May be automatically given in. The metadata that is automatically given can include, for example, the one that indicates the content that always appears in the event. Specifically, for example, at a wedding, the groom and the bride always appear, so the metadata indicating such contents is automatically added by the system side when the type of the input sheet is specified.

Next, returning to FIG. 3, in step S3, it is determined whether or not metadata is already added to the selected data, and if it is added, the metadata (hereinafter, existing data) is added. The metadata (also referred to as metadata) and the metadata given this time (hereinafter, also referred to as new metadata) are combined. If existing metadata is not added,
Through step S3, the process proceeds to step S4.

The synthesis of metadata in step S3 is to unify existing metadata and new metadata, for example, to eliminate duplication of contents. As the existing metadata, for example, the metadata previously added by the method of the present embodiment or the data such as the multimedia content to which the metadata is added has been previously added by the author or the like. Data,
Alternatively, the metadata or the like automatically added by the system side as described above in the description of step S2 may be used.

When new metadata is added to the existing metadata as described above, the amount of metadata may unnecessarily increase and the contents may be duplicated. If this is left unattended, the storage capacity of the storage medium such as the memory is wastefully consumed,
In addition, it may affect the search described later. Therefore, in the present embodiment, existing metadata and new metadata are combined to unify the metadata.

Although various methods are conceivable for synthesizing metadata, the following method is adopted in this embodiment. First,
The data format of metadata is divided into predetermined items,
It consists of an item section and a content section. Such a classification for each item may be automatically performed on the system side after the addition of metadata, or when the above-mentioned metadata is added,
For each item, the content that the user can select may be presented, and the item may be described in the item section and the content selected by the user may be described in the content section.

At the time of synthesizing the metadata, the logical operation of the contents is performed between the same items to synthesize the metadata. In the present embodiment, unification is adopted as the logical operation, but the present invention is not limited to this, and other logical operations may be performed.

For example, if the metadata of each item is X {a,
b, c}. Here, X indicates the item part,
{A, b, c} indicates the content part corresponding to the item part X. Then, for example, if the existing metadata is X {a, b,
c} and the new metadata includes X {a, c, d}, unifying these results in metadata of X {a, b, c, d}. As a result, the existing metadata and the new metadata are unified, and the metadata whose contents are duplicated can be unified.

Next, a more specific description will be given with reference to FIG. FIG. 14 (a) is new metadata, FIG. 14 (b).
And (c) are assumed to be existing metadata. The result of unifying these metadata is shown in FIG. In Fig. 14 (a), "clothes"
The content of the item is "groom (tailcoat)" and "bride (wedding dress)", while in Fig. 14 (b), the content of the item "clothing" is "groom (tailcoat (black), tie)". There is. Also, FIG. 14 (c)
In, there is no item called "clothes".

Then, as a result of synthesizing the contents of the item "clothes" by unifying, as shown in FIG. 14D, the items "clothes" are "groom (tailcoat (black), tie)" and The result is "Bride (wedding dress)". That is, with respect to “groom (tailcoat)”, duplication of metadata has been eliminated, and “(black), tie” and “bride (wedding dress)” have been unified within one item of “clothing”. The same applies to other items.

By synthesizing the metadata in this way, it is possible to prevent the metadata amount from unnecessarily increasing, and also in the search described later, it is possible to prevent the search from being duplicated and to obtain a good result. Get it.

Next, returning to FIG. 6, in step S4, the selected data and the inputted metadata are stored in association with each other. In the example of FIG. 4 described above, for the sub data of the image shown in the area 732, the area 7 showing the input sheet is displayed.
The information input in 4 is added as metadata, and the metadata is associated with the sub-data and stored in a recording medium such as DISK in FIG. In addition,
When the metadata is combined in step S3, the selected data and the combined metadata are stored in association with each other.

FIG. 8 is a table showing the relationship between the data and the metadata attached to the data.
It is shown that the type of event and the input candidate of the metadata are added as the metadata for each of the plurality of groups of frames included in. Here, although character information and the like that can be handled as simple keywords are mainly stored, information having a certain structure such as a relationship between characters and information other than characters can be added.

In addition, the index information indicating the data related to each metadata is created or updated for use at the time of search. FIG. 9 shows an example of index information, in which data related to each metadata is shown in a table format. This index information will be stored in a recording medium such as DISK in FIG.

<Data Retrieval Process> Next, an example of a process for retrieving data based on the metadata added to each data in the data management system of this embodiment will be described.

In the search, the user gives a search condition, the search condition is compared with the metadata given to each data in the above-described procedure, and the search is performed by using the index information as shown in FIG. The data or sub data to which the metadata matching the conditions is added is picked up.

Hereinafter, a specific example of data retrieval will be described.
explain. FIG. 10 is a flowchart of the data search process.

In step S11, data such as index information for retrieval is operated on the RAM of FIG. 1 so that the data management system can use it. The data to be searched, its metadata, and the like can be held on a computer such as a PC on the LAN via the NCU in FIG. 1 or on a computer on an external network via the DTU.

Step S12 shows a process in which the user inputs a sentence or word as a search condition. Here, the user can input a natural sentence expressing an event to be searched as a search condition from the KB of FIG. Further, this search condition may be accepted by the PC of FIG. 1 and input via the LAN.

Here, as an example, the case where the sentence “Tanaka and his wife's up” is input as a search condition will be described below.

In step S13, the natural language processing technique is applied to the sentence or word input as the search condition,
Performs analysis such as morphological analysis and syntactic analysis. Although various known methods can be used as techniques and methods for each of these analyses, here, a sentence is divided into a group of semantic words, and the parts of speech and meanings of the words formed by the division are defined. Anything that can retrieve information and the relationship between those words in the sentence may be used.

FIG. 11 shows an example of the result of morphological analysis and syntactic analysis of the search condition "Tanaka and wife's up". Here, this example sentence is decomposed into words such as "Tanaka", "wife" and "to", and information such as part of speech such as "noun" and "connective" is assigned to each. Then, it is possible to analyze that each of the words has a relation such as "modification" or "parallel".

In step S14, a data search is performed based on the analysis result obtained in step S13 and the correlation between the event, the metadata candidate, etc. in the template shown in FIG.

Here, in the analysis result as shown in FIG. 11, an object or event indicated by an important word having a part of speech such as a noun and the relationship between them are searched, and the word and each corresponding word are searched. Similarity is calculated for combinations of objects and events.

In the case of this example, for example, it is assumed that the groom has the attribute "Tanaka" as the name and the bride has the attribute "wife" as the alias. In addition, it is assumed that there are some events that have the attribute “up” itself or the word of the same meaning as the state of the event.

Then, these are the search results, and the similarity is calculated from the degree of similarity in word sense, what kind of attributes are related to events, objects and relationships. The degree of similarity referred to here is a value indicating how appropriate a particular word in a sentence input by the user is as an expression indicating each object, event, or relationship.

In step S15, the process branches depending on whether there is a candidate for the object, event, relationship, etc. retrieved in step S14. If no candidate is found, the search ends with no corresponding candidate. If there is even one candidate, the process proceeds to step S16.

In step S16, it is checked whether or not there is a structure similar to the relation of words as shown in FIG. 11 with respect to the candidate object, event or relation, and the structure is added as a priority candidate. is there. Figure 11
In the example, "Tanaka" and "wife" are connected in parallel, and "wife" and "up" are connected in a modified relationship, so "groom"
-"Face" (part / whole relationship), "Bride"-"Face" (part / whole relationship), in the scene where the two faces are up,
There is a situation of "up", and this subject is the "face".

As a result, although indirectly, an event having an object searched for by "Tanaka" and a situation having a state of "up" and an event having an object searched by "wife" and a situation of "up" are indirectly performed. , A similar structure is found. Therefore, data or sub data related to such an event will be added as a candidate.

In step S17, based on the similarity of the structure found in step S16 and the information of the similarity of the word calculated in step S14 for each searched data,
This is a part in which the similarity between each data and the search condition is calculated and arranged in descending order of similarity.

As a result, the search is completed, and one or more data or sub-data as a final search candidate and the similarity of how close each of them is to the sentence input by the user. Will be obtained. This completes the search.

Next, the processing for making the search result obtained by the processing of FIG. 10 visible to the user will be described. FIG. 13 is a flowchart showing a search result display process.

In step S21, the information such as the template showing the correlation between the retrieval result obtained by the processing of FIG. 10 and the event related to the data to be retrieved is available on the DISK or RAM. , Is the part that performs various pretreatments.

Step S22 is a part for presenting to the user the candidates of the data retrieved by the processing of FIG. 10. However, there are cases where only some of the data and images are the parts that the user wants, so I asked them to actually inspect the searched data candidates in some form,
It is desirable to have them choose what they want. Therefore, here, each data candidate is shown by, for example, a still image or a description of a typical portion of the scene in the data. This allows
The user can obtain information such as the number of candidates and simple characteristics of each candidate.

A step S23 asks the user to actually select the data candidate shown in the step S22. Further, step S24 is a process of determining the result of this selection. If a candidate is selected, the process proceeds to step S25, and if no candidate is selected, the process ends.

In step S25, the user checks whether or not the selected data is the desired data, and in order that the user can select the data closer to the desired one, the contents of the selected candidate are checked. Get them to go. Therefore, not only the search candidate data but also the event strongly related to the event to which the data belongs are checked based on the relationship between the events in the template of FIG. 5, and if there is data belonging to the event, That is also a part to be added to the display of candidates.

However, if the user has designated in advance,
The candidates are added only when the contents are applicable to the constraints and conditions of the user's environment and situation. The constraints and conditions regarding the user's environment and situation here are:
Due to restrictions imposed by the user's device and network environment, time restrictions imposed by whether or not the user can see the search results carefully, restrictions and conditions imposed by billing, copyright and ethical restrictions. However, it does not matter if it is a limitation or restriction.

In step S26, preparations are made to provide the user with the contents of the data selected by the user and the data added in step S25. For example, when each piece of data is a moving image, these pieces of data are arranged based on the relationship between events and are configured so that they can be reproduced as a series of one moving image.

In step S27, the data prepared in step S26 is provided to the user. For example, when each data is a moving image, processing such as reproducing and displaying them is performed. After the data is provided, the process returns to step S24 so that another candidate can be seen.

With such a processing order, when the user selects a part of the desired multimedia content, he or she also looks at the part closely related to the part so that the user can understand the part that seems to be the desired content. It becomes easier and you can search the multimedia contents necessary to understand the contents, and know the parts other than the relevant part.

As described above, in the present embodiment, it is possible to easily attach metadata to various data such as multimedia contents, and it is easy to perform a search and an understanding of a search result using the metadata. At the same time, since a large amount of metadata is added due to the ease of adding metadata, it is easy to improve the search accuracy, and the user can easily and effectively use the data search. It was

In the template shown in FIG. 5, the events having similar contents and structure in the template are created by deriving from the specific event class by utilizing the object-oriented feature.
The template creator and the user can easily create a template for each event.

FIG. 12 illustrates this. In the figure,
Similar to FIG. 5, the template is T8-X (X is a number),
Events are represented by E8-X (X is a number), Obj8-X (X is a number), and relationships are represented by R8-X (X is a number).

In this case, by declaring a common part of a plurality of events as an event class, describing only the difference from this class in its own event, and deriving the common part from the above-mentioned event class, The description of the common part can be omitted. In FIG. 12, event E8-2
The event E8-4 is derived from the event class C8-1 and inherits the objects and relationships in the event class C8-1.

Although the preferred embodiment of the present invention has been described above, the object of the present invention is to supply the program code of software that realizes the functions of the above-described embodiment to a system or apparatus and to execute the system or apparatus's operation. It is needless to say that it is also achieved by the computer (or the CPU or MPU) reading and executing the program code. In this case, the read program code itself realizes the functions of the above-described embodiment, and the program code or a program product such as a storage medium storing the program code constitutes the present invention. Further, by executing the program code read by the computer, not only the functions of the above-described embodiment are realized, but also an operating system (OS) running on the computer is executed based on the instruction of the program code. It goes without saying that a case where some or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing is also included.

Further, after the program code is written in the memory provided in the function expansion card inserted in the computer or the function expansion unit connected to the computer,
A case where a CPU or the like included in the function expansion card or the function expansion unit performs a part or all of the actual processing based on the instruction of the program code and the processing realizes the functions of the above-described embodiments is also included. Needless to say.

[0126]

As described above, according to the present invention, information for retrieval can be easily and efficiently added to the managed data and the amount of metadata is increased unnecessarily. Can be prevented.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration example of a computer device in which a data management system according to an embodiment of the present invention is constructed.

FIG. 2 is a diagram showing a mode in which an external recording medium storing a data management program and the like of the present embodiment is read into a computer device.

FIG. 3 is an example of information defining relationships between events,
Individual events X-A to I included in a certain event X
It is a figure showing the inclusion relation of.

FIG. 4 is a diagram showing an example of a type of each event and a table of metadata candidates set for each event.

FIG. 5 is a diagram showing an example of a template in which correlations such as events and metadata candidates are defined.

FIG. 6 is a flowchart showing a flow of a metadata adding process.

FIG. 7 is a diagram showing a display example of a CRT at the time of inputting metadata.

FIG. 8 is a diagram showing a table showing a relationship between data and metadata attached to the data.

FIG. 9 is a diagram showing an example of index information used at the time of data search.

FIG. 10 is a flowchart of a data search process.

FIG. 11 shows an example of a result of morphological analysis and syntactic analysis of “up of Tanaka-kun and wife”, which is an example of search conditions.

FIG. 12 is a diagram showing another example of a template defining a correlation between events, metadata candidates, and the like.

FIG. 13 is a flowchart showing search result display processing.

14A is a diagram showing the contents of new metadata, FIGS. 14B and 14C are existing metadata, and FIG. 14D is a diagram showing the contents of synthesized metadata.

Claims (6)

[Claims] 1. A data management system for individually managing data by adding metadata for data search to data to be managed, and means for accepting selection of the data to which the metadata is to be added in advance. Means for providing one or more candidates consisting of a set of defined single or multiple metadata, and selection of metadata to be added to the selected data from the provided metadata candidates. Means for accepting, combining means for combining the metadata already added to the selected data and the selected metadata, the selected data, and the metadata combined by the combining means And a means for storing them in association with each other, and a data management system. 2. The synthesizing unit sets one metadata whose content overlaps between the metadata already added to the selected data and the selected metadata. The data management system according to claim 1, wherein: 3. The metadata is divided into predetermined items, and the synthesizing unit performs a logical operation or an arithmetic operation of the contents between the same items. Data management system. 4. The data management system according to claim 1, wherein the logical operation is unification. 5. A data management method of individually adding metadata for data search to data to be managed for data management, the step of accepting a selection of the data to which the metadata is added, A step of providing one or a plurality of candidates composed of a defined single or a plurality of sets of metadata, and a step of selecting metadata to be added to the selected data from the provided metadata candidates. A step of receiving, a synthesizing step of synthesizing the metadata already added to the selected data and the selected metadata, the selected data, and the metadata synthesized in the synthesizing step And a step of storing them in association with each other, and a data management method. 6. In order to manage data by individually adding metadata for data search to managed data,
A step of receiving a selection of the data to which the metadata is given to a computer; a step of providing one or a plurality of candidates composed of a predetermined single or a plurality of sets of metadata; and the provided metadata A step of receiving a selection of metadata to be added to the selected data from among the candidates, and a combination of combining the metadata already added to the selected data and the selected metadata. A program that executes a step, and a step of storing the selected data and the metadata synthesized in the synthesis step in association with each other.