JP2006133939A - Content data retrieval device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently extract content which a user wants from a content database, in which a large amount of content data are stored, with less processes. <P>SOLUTION: This is a content retrieval device which stores content data in association with multiple kinds of attribute values and concentrates content with an attribute value, which a user inputs, as the retrieval conditions. In this case, the content retrieval device determines whether content data discrimination efficiency is high or low for each attribute item and prompts a user to input retrieval conditions from an attribute item of higher discrimination efficiency. More specifically, the content retrieval device computes the distribution values of content data for each attribute item and determines that an attribute item of large distribution value is an attribute item of high discrimination efficiency. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a database search apparatus for searching content data desired by a user from a large amount of content data.

  2. Description of the Related Art In recent years, a consumer content storage server called a home server or an in-vehicle server is equipped with a large-capacity storage device such as a hard disk and can store a large amount of music content and video content. As a result, users can record and record music and video as much as they like, and can select and enjoy their desired content from a large amount of content stored at any time.

  Here, as a method for selecting the content desired by the user, a method in which accumulated content is displayed as a list in a list format and the user selects a desired content from the list is widely used. However, in this method, the procedure for selecting from the list becomes complicated as the amount of content accumulated increases, and therefore, retrieval based on content attributes may be used in combination. That is, at the timing when content is accumulated, the system automatically obtains the attribute of each content (for example, title name, creation date, genre, impression when viewing the content, etc.) using a known technique such as CDDB and associates it with the content data. When the user inputs an attribute condition desired by the user, a method is used in which the system extracts only content that matches the attribute condition and presents it to the user.

  Specifically, a music content search system (for example, Patent Document 1) that can extract only music content that matches a user-specified condition from a large amount of stored music content, A method for improving the condition input interface and facilitating the narrowing operation (for example, Patent Document 2) has been proposed.

  In the music content search system disclosed in Patent Document 1, it is possible to input conditions regarding a plurality of song attribute information such as song title, singer name, tempo, genre, age, etc. When a condition regarding content is input in a range, the system extracts music that matches the condition from a plurality of music stored in the database. According to this system, if the user specifies only the conditions that he or she remembers regarding the desired content, the music contents that match the conditions are automatically narrowed down. For example, even if the user forgets the song name, the narrowing is performed. It is possible to reduce the complexity of the operation because it is only necessary to search for the music again.

In addition, in the database search method disclosed in Patent Document 2, the user is allowed to select two attribute items from among a plurality of attribute information set for a song, and a coordinate system is set with this as an axis. Based on the attribute value corresponding to the attribute item, the music content is indexed and arranged on the coordinate system, and the screen is presented to the user. FIG. 1A shows an example of a display screen presented to the user by this database search method. In this example, the user has a desire to listen to “a song released after 1990 and a song with a fast tempo”, and coordinates that specify “age” and “tempo” on the vertical and horizontal axes as attribute values. An example of screen display when content is arranged on the system is shown. As shown in FIG. 1 (a), in this search method, the music corresponding to the wish is mapped within a specific range of the coordinate system, so the user can intuitively identify the desired music and the music that is not so out of a large amount of music content. Can be sorted out visually and visually. At this time, by designating an arbitrary closed area with respect to the coordinate system, it is possible to narrow down to only the music included in the area, so that it is possible to reduce the complexity of the subsequent music selection operation. .
JP 2002-73041 A JP 2001-134586 A

  However, in the music content search system disclosed in Patent Document 1, it is assumed that there is almost infinite music content stored in the database, and there is always music that matches the conditions set by the user. ing. In general, the music content that can be played back by the music playback device often occupies most of the music content purchased by the user with money. For this reason, except for the case where it is used for business purposes, the number of accumulated contents is limited especially in content servers used in homes and automobiles, and the distribution of contents is biased depending on user preferences. There are many cases. Under such circumstances, in the music content search system disclosed in Patent Document 1, the user must specify the desired condition without knowing the distribution of viewable content, and as a result, there is no music that matches the condition. There is a possibility that an attribute condition that does not have an effect on narrowing down the content may be specified, which requires troublesome work such as resetting the condition. More specifically, for example, even if a user who accumulates only music of the '90s sets the condition for the content attribute of the age, there is almost no narrowing effect, and the condition setting operation is wasted. And so on.

  In addition, in the database search method disclosed in Patent Document 2, a narrowing-down condition desired by the user can be specified by specifying a closed region on the coordinate system while confirming the distribution of content with the attribute specified in advance as the coordinate axis. Input. However, as described above, depending on the attribute specified as the axis, there is a possibility that the distribution will be concentrated in a certain place, in this case, specify a closed region with a width in the attribute axis direction However, since there is no content, there is almost no effect, and the condition setting operation is wasted. More specifically, as shown in FIG. 1 (b), when a user who has accumulated only music of the '90s takes the attributes of “age” and “tempo” on the coordinate axis, Since the content distribution is small, the user has to specify an area only in the “tempo” axis direction, and the operation specifying “age” is wasted.

  Therefore, an object of the present invention is to focus attention on the distribution of accumulated content, thereby eliminating unnecessary setting operations of conditions and efficiently narrowing down the content in order to quickly reach the content desired by the user It is to provide a search device.

  In order to achieve the above object, a content data search device according to the present invention is a content search device for searching content data desired by a user from a plurality of content data, and the content data and each content data A content data storage unit that stores a plurality of types of attribute values, a search condition input unit that inputs a user-desired attribute value as a search condition for searching for content data, and a user input from stored content data For each of the search unit that extracts only the content data that matches the search condition, the search result presentation unit that presents the extracted content data to the user, and the attribute item for searching the content data, the attribute value is the search condition. The content extracted by the search unit when it is assumed that A search condition presentation order determining unit that determines the order of presenting attribute items to the user according to the degree of decrease in the number of data, and the attribute items that can be input by the user as the search conditions A search condition presenting unit for presenting to the user according to the order determined by the presenting order determining unit, thereby achieving the object.

  Here, for example, the search condition presentation order determination unit calculates the distribution value of the content data for each attribute item, and presents the attribute items to the user in descending order of the value.

  Further, the search condition presenting unit may present the attribute items to the user by displaying a plurality of attribute items in order on the screen. At this time, it is arranged and displayed on the screen according to the order determined by the search condition presentation order determining unit, but the upper-level fixed number of attribute items increase the display size, change the display color to a conspicuous color, symbols, marks, etc. Highlighting may be performed so that the user can more easily recognize than other attribute items. Conversely, a fixed number of attribute items in the lower order may be hidden. In addition, the search condition input unit may specify an attribute item for the search item displayed on the screen using an input device such as a touch panel or a mouse.

  In addition, the search condition presenting unit may present the attribute item to the user by voice so that the user can recognize the attribute item without looking at the screen, for example, while driving a car. Further, the search condition input unit may be able to input the attribute item by the voice uttered by the user.

  In addition, when the number of content data extracted by the search unit based on the search conditions input up to that point exceeds a certain number, the search condition presenting unit inputs other conditions to the user to further narrow down the content data You may make it perform the presentation which prompts.

  In addition, the search condition presenting unit presents the number of content data extracted by the search unit based on the search conditions input up to that point and presents to the user how much the data has been narrowed down. Also good.

  In addition, as a specific example of the attribute item stored in the content data storage unit, any of song name, artist name, genre, release date, song playback time, song tempo, song rhythm, song impression information May be included.

  According to the present invention, since the user can know which attribute item can be efficiently narrowed down from among a plurality of attribute items for narrowing down the content, the user can efficiently perform with fewer condition setting operations. Desired content can be narrowed down. More specifically, even when searching for music content whose song title is not known, it is possible to narrow down to candidate music content with a few steps, so that the operation procedure for searching for the target music content can be simplified.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(Embodiment 1)
FIG. 2 is a block diagram showing a functional configuration of the content data search apparatus according to the embodiment of the present invention. The content data search device 200 includes a search condition input unit 201, a search unit 202, a search result presentation unit 203, a content data storage unit 204, a search condition presentation order determination unit 205, and a search condition presentation unit 206. Is done.

  FIG. 3 shows a specific configuration of content data search apparatus 200 in the embodiment of the present invention. The content data search apparatus 200 includes a microphone 301 for acquiring voice uttered by the user, a storage device 303 such as an HDD for storing content data and attribute information related to each content data, and a search usable by the user. Various information such as conditions and search results are presented to the user as visual information, and the touch panel monitor 304 can input the search conditions by touching the screen, as well as search conditions and search results that can be used by the user. Content stored in the storage device 303 by converting voice data acquired via the microphone 301 and the speaker 305 that presents various information as audio information to the user into search conditions for searching for content by voice recognition processing Match the search conditions entered by the user from the data Extracts only the content data, instructs the touch panel monitor 304 or the speaker 305 to present the extraction result, calculates the variance value for each attribute item for the content data stored in the storage device 303, The CPU 302 calculates the discrimination efficiency of each attribute item based on it.

  In the configuration as described above, the search condition input unit 201 inputs an attribute condition for content desired by the user, and corresponds to, for example, a voice input device such as the microphone 301 and a voice recognition program processed by the CPU 302. Specifically, when the user utters a genre of music such as “genre is J-POP” or impression information such as “music with a bright impression”, the search condition input unit 201 determines “ Search conditions for narrowing down contents such as “genre = J-POP” and “impression = bright” are acquired.

  The content data storage unit 204 stores content data and a plurality of attribute values associated with each content data, and corresponds to a storage device such as the HDD 303, for example. The content data mentioned here may be music data, video data, text data such as novels, and various types of content data. For example, in the case of music data, the song title, artist, In addition to unique information for identifying a song such as name, genre, release date, playback time of the song, etc., information on music feature quantities such as tempo and rhythm is stored. Furthermore, the impression received when listening to the music from the music feature amount using a known technique may be analyzed in advance, and impression information such as the brightness, intensity, and freshness of the music content may be stored. . FIG. 10 shows an example of music attribute data stored in the content data storage unit 204.

  The search unit 202 extracts only the content data desired by the user from the stored content data based on the conditions regarding the content data input by the user, and corresponds to the CPU 302 and a search program operating there.

  The search result presentation unit 203 presents information on the content data extracted by the search unit 202 to the user. For example, the search result presentation unit 203 provides the search result as audio information and the touch panel monitor 304 that provides the extraction result on a screen display. This corresponds to the speaker 305 to be used. In this embodiment, the visual information and the audio information can be provided to the user. However, the visual information alone or the audio information alone may be used.

  The search condition presentation order determination unit 205 calculates the discrimination efficiency for each condition among a plurality of conditions for narrowing down the contents, and determines the order of search conditions to be presented to the user. This corresponds to an order determination program.

  The search condition presentation unit 206 sequentially presents the search conditions in accordance with the order determined by the search condition presentation order determination unit 205 in order to present information for prompting the user to input the search conditions. A touch panel monitor 304 that provides search conditions on a screen display and a speaker 305 that provides search conditions as audio information correspond to the display. In this embodiment, the visual information and the audio information can be provided to the user. However, the visual information alone or the audio information alone may be used.

  Next, the operation of the content data search apparatus according to the first embodiment having the above configuration will be described. FIG. 4 is a flowchart showing the overall flow of the operation of the content data search apparatus 200.

  First, the content data search device calculates the discrimination efficiency for each condition among a plurality of conditions for narrowing down the content, determines the order of search condition candidates to be presented to the user (S41), and the result The user is prompted to input content search conditions by presenting search condition candidates to the user by screen display and voice based on the above. (S42). Next, the user waits for the user to input a content search condition by looking at the presented search condition candidates (S43). When it is detected that the user has input a content search condition (Yes in S43), content search processing is performed to extract content that matches the condition from the accumulated content data (S44), and the search has been narrowed down to the present time. A process of presenting content data to the user is performed (S45). Next, it is determined whether or not the setting of the content search condition has been completed (S46). If it has not been completed (No in S46), an order determination process of search condition candidates to be presented to the user is performed again (S41). If the setting of the content search condition is completed (Yes in S46), the processing of the content data search device is terminated.

  FIG. 5 is a flowchart showing a subroutine of search condition presentation order determination processing (S41) shown in FIG.

  In this search condition presentation order determination process, first, one attribute item of a content to be a key for searching for content is selected (S51). Content attribute items are stored in the content data storage unit 204. Next, it is determined whether or not the user has already set the condition for the selected attribute item, that is, whether or not the user has already narrowed down the content by using the attribute condition so far (S52). If the selected attribute item has already been set with conditions (Yes in S52), the next attribute item is selected (S51). On the other hand, if the attribute item has not yet been subjected to condition setting (No in S52), the content data that has been narrowed down to the present time is plotted in a coordinate system with the attribute item as an axis, and the variance value is calculated. For example, FIG. 6A shows an example in which music content data is plotted with respect to the attribute item “age”. In this example, content data is accumulated almost uniformly from music data released in the 1960s to music data released in the 2000s, indicating that the variance value is high. On the other hand, FIG. 6B shows an example in which music content data is plotted with respect to the attribute item “genre”. In this example, the content data is almost biased to the genre “J-POP”, indicating that the variance value is low. The attribute items are rearranged so that the calculated variance values are arranged in descending order from the largest value (S54). In general, in order to narrow down content data efficiently with a small number of condition settings, the number of contents to be narrowed down by adding a new condition among a plurality of attribute conditions can be reliably narrowed down to a certain fraction. It is important to use the conditions first. In the example of FIG. 6, when the user next specifies the attribute item “age” as a search condition, the number of contents to be narrowed is always reduced no matter how the user specifies the attribute value. When the attribute item “genre” is designated as the search condition, the number of contents to be narrowed down is hardly reduced when the user designates “J-POP” as the attribute value. Focusing on this, the content data search apparatus according to the present embodiment realizes efficient content data narrowing by prompting the user to set conditions in descending order of content distribution by attribute item. Finally, it is determined whether or not the sorting process has been completed for all the attribute items (S55). If there is an unprocessed attribute item (No in S55), the process is repeated for the next attribute item. If the processing is completed for all the attribute items (Yes in S55), the search condition presentation order determination processing is terminated.

  FIG. 7 is a diagram showing an example of a screen presented to the user in the search condition presentation process (S42) shown in FIG. In the search condition presentation process, first, the user presents the search condition 701 that has been set up to the present time, and prompts the user not to use the same search condition again. In addition, according to the order determined by the search condition presentation order determination process, attribute items as search candidates are presented in a list format as in 702 to 705. In this example, it is determined that the attribute item “age” is the most efficient when narrowing down the content, and then the attribute items “brightness”, “genre”, and “intensity” continue. If the number of attribute items is too large to be drawn on one screen, other attribute items can be presented by displaying the scroll bar 706 and scrolling the screen. The user can specify an attribute item that he / she wants to set as a narrowing-down condition among the plurality of attribute items displayed on the screen, and can specify an attribute value on an attribute value input screen (not shown). Alternatively, the attribute value can be input by voice input. The search condition presentation order determination process obtains a variance value for each attribute item, but an attribute item having a large value has a high filtering efficiency and displays an emphasis mark 707 for prompting the user to set a condition relating to the attribute item. You may do it. Also, in the search condition presentation by voice, user input may be prompted only for the attribute item selected as the first in the search condition presentation order determination process, such as “Please specify the age of content”. In this search condition presenting process, attribute items with high discrimination efficiency are emphasized and presented, so that the content can be efficiently narrowed down to the user with few procedures.

  FIG. 8 is a diagram showing an example of a screen presented by the content search result presentation process (S45) shown in FIG. In the content search result presentation processing, first, the search condition 801 set by the user up to the present time and the content data number 802 narrowed down by that are presented. In this example, two attribute conditions of “tempo” and “age” have been set up to now, indicating that the content data has been narrowed down to 24 songs. The narrowed-down content data is displayed in a list format (803 to 805). If the content data does not fit within one screen, scroll bar 806 is displayed, and other content data can be displayed by scrolling the screen. In addition, switches 807 and 808 that allow the user to specify whether or not to perform a new condition setting process for further narrowing down the content data are displayed. On this screen, the user can confirm the display of the number of content data 802 and, if the number is small, can select a procedure for extracting one desired content data from the list-displayed content data list. If it is troublesome to extract content data from the list display with a large number of narrowing down, it is possible to select a procedure for further narrowing down the content data by specifying other attribute conditions.

  FIG. 7 is an example of a screen presented to the user in the search condition presenting process (S42), assuming that the user mainly inputs a search condition by touching the screen and operating the touch panel. When considering application to a terminal, an operation while driving is assumed, and it is considered that an operation mainly using voice is also necessary without paying attention to the screen. In consideration of this, FIG. 9 shows an example of information that the search condition presenting process (S42) presents to the user in order for the user to operate without paying attention to the screen. Here, these two operation systems shown in FIG. 7 and FIG. 9 may allow the user to freely select either, for example, in conjunction with a sensor that detects whether the automobile is running, You may make it switch automatically, such as voice operation in the middle, and touch panel operation in the stop.

  If the user cannot watch the screen, it is difficult for the user to recognize the attribute items that are search candidates in the list format, so only the search conditions ranked in the highest order by the search condition presentation order determination process are spoken. To the user. In the example of FIG. 9, when “age” is ranked at the top as an attribute item, “Please tell me the release date of the music you want to extract” to the user by voice, so that “age” is shown to the user. Encourage the setting of conditions. At this time, along with the voice, attribute items to be input by the user such as 902a and 902b are also displayed on the screen, and specific attribute values such as 903a and 903b are displayed on the screen. The user can easily input the search condition even when the system voice notification is missed. Also, the number of songs that have been narrowed down at the present time is presented on the screen (901a to 901c), so that the user can confirm the narrowing effect by inputting the search conditions. Furthermore, when the number of narrowed down songs is large, the user is prompted to input other search conditions for further narrowing down. For example, when narrowed down to 10 or less, the narrowed content list is displayed as a list 904. The format is displayed.

  As described above, in the content data search apparatus according to the present embodiment, the system can efficiently narrow down content data with a small number of condition settings by presenting attribute items with a high degree of discrimination to the user. For example, when the user cannot remember the singer name or the song title, the target content data can be efficiently extracted even when the user finally selects the content from the list.

  Since the content data search device according to the present invention can efficiently narrow down content data desired by the user from a large amount of content with a small number of procedures, the music playback device / Applications to various content playback devices such as video playback devices are useful.

(A) In the music content search interface in the prior art, a diagram showing a display example when the content distribution is uniform. (B) In the music content search interface in the conventional technology, there is a bias in the content distribution. Of display example The block diagram which shows the structure of the content data search apparatus of embodiment of this invention The block diagram which showed the specific structure of the content-data search apparatus of embodiment of this invention The flowchart which shows the whole operation | movement of the content data search device of embodiment of this invention. The flowchart which shows the subroutine of search condition presentation order determination processing (S41) shown by FIG. (A) A diagram showing an example of content mapping when “age” is the coordinate axis. (B) A diagram showing an example of content mapping when “genre” is the coordinate axis. The figure which shows the example of a screen shown to a user in order to prompt search condition input when assuming touch panel operation The figure which shows the example of a user presentation screen which shows a content search result The figure which showed the example of the sequence of the screen, the example of a voice, and the user input which are presented to the user in order to prompt the search condition dormitory when the voice operation is assumed The figure which shows the example of the content attribute data which is accumulated in the database

Explanation of symbols

DESCRIPTION OF SYMBOLS 200 Content data search device 201 Search condition input part 202 Search part 203 Search result presentation part 204 Content data storage part 205 Search condition presentation order determination part 206 Search condition presentation part 301 Microphone 302 CPU
303 Storage Device 304 Touch Panel Monitor 305 Speaker 701 Area for Displaying Search Conditions Already Set 702 to 705 Search Condition Candidates 706 Scroll Bar for Displaying Search Condition Candidates Not Fit on One Screen 707 Large or Low of Search Condition Discrimination Efficiency 801 A region for displaying a search condition that has already been set 802 The number of content that matches the search condition and is narrowed 803 to 805 The narrowed content list 806 A scroll bar for displaying content that does not fit on one screen 807 Button for continuing content narrowing 808 Button for finishing content narrowing 901a-c Number of content that matches the search condition and 902a-b prompts input of search condition 903a-b Specific attribute values to be entered as search conditions 904 List of narrowed contents

Claims (12)

A content search device for searching content data desired by a user from a plurality of content data,
A content data storage unit that stores content data and a plurality of types of attribute values associated with each of the content data, and a search condition input for inputting an attribute value desired by the user as a search condition for searching the content data A search unit that extracts only content data that matches the search condition from the content data;
A search result presentation unit for presenting content data extracted by the search unit to a user;
For each of the attribute items for searching the content data, when it is assumed that the attribute value is input as a search condition, the degree of decrease in the number of content data extracted by the search unit is estimated, and the degree of decrease is increased or decreased. Based on the order determined by the search condition presentation order determination unit, the search condition presentation order determination unit for determining the order in which the attribute items are presented to the user and the search condition presentation order determination unit to prompt the user to input the search conditions A content data search apparatus comprising: a search condition presenting unit to be presented to a user. The said search condition presentation order determination part calculates the dispersion | distribution value of content data for every attribute item, and sets the order which shows an attribute item to a user in order with a high said dispersion | distribution value. Content data search device. The content data search device according to claim 1, wherein the search condition presentation unit displays a plurality of the search condition display units simultaneously on the screen in the order determined by the search condition presentation order determination unit. The search condition presenting unit changes a display size, changes a display color, or adds a specific symbol to the display of the upper fixed number of attribute items in the order determined by the search condition presenting order determination unit. The content data search apparatus according to claim 3, wherein the content data is highlighted from other attribute items by the above-described method. 5. The content data search device according to claim 3, wherein the search condition presentation unit hides a fixed number of attribute items in a lower order of the order determined by the search condition presentation order determination unit. The content data search device according to any one of claims 3 to 5, wherein the search condition input unit specifies an attribute item by selecting the attribute item displayed on the screen with a touch panel or a mouse. . The content data search device according to claim 1, wherein the search condition presenting unit presents only the highest-order attribute item in the order determined by the search condition presenting order determining unit. The content data search device according to claim 6, wherein the search condition presenting unit presents the attribute item to the user by voice. The content data search apparatus according to claim 8, wherein the search condition input unit inputs an attribute item by a voice uttered by a user. The said search condition presentation part performs the presentation which urges | urges another condition input with respect to a user, when the number of the content data extracted by the said search part exceeds a fixed number, The one of Claims 1-9 characterized by the above-mentioned. The content data search device according to claim 1. The said search result presentation part presents the number of content data extracted by one or more conditions which the user has input so far to the user. 11. Content data search device. The attribute item stored in the attribute data storage unit includes any of a song title, artist name, genre, released date, song playback time, song tempo, song rhythm, song impression information. The content data search device according to claim 1, wherein the content data search device is a content data search device.

Images