JP4774087B2 - Movie evaluation method, apparatus and program - Google Patents

Description

The present invention relates to a moving image evaluation method , apparatus, and program . In particular, the present invention relates to a method , an apparatus, and a program for automatically evaluating a moving image published on the Internet.

In recent years, video sites on the Internet such as YouTube (registered trademark) have become widely known. A plurality of videos can be viewed on the video site, and a viewer who views the video can give an evaluation to each video.
“Comment & rating on videos”, [online], Youtube Fan, [Search July 25, 2008], Internet <URL: http: // www. youtube-fan. com / 5/23 />

  However, in order to give a rating to a video, a viewer accesses the video site, searches for the video to watch, and inputs the rating by performing a predetermined operation on the rating input field attached to the viewing interface for that video. Had to do. Moreover, the evaluation was determined only by the quantity such as the number of stars, and it was impossible to know what aspect of the video was evaluated only by the information. Also, viewers may post comments separately from ratings, and it is possible to know what aspect of the video was evaluated by reading the comments, but read many comment sentences There is a need.

  On the other hand, blogs are widely used as information transmission media on the Internet. A video viewer may describe an evaluation of a specific video as a blog article.

  An object of the present invention is to make it possible to grasp at a glance what aspect of a moving image is evaluated by using a blog article.

  The present invention provides the following solutions.

  (1) In the method of giving an evaluation to the moving image M, the association of the scene Ms of the moving image M and the keyword Ks related to the scene Ms is stored in the storage unit, and the computer has a plurality of blog articles with hyperlinks to the moving image M A blog article search step for searching for, an evaluation word extraction step for extracting a keyword K related to the moving image M and an evaluation word H related to the keyword K from the plurality of searched blog articles, a keyword Ks related to the scene Ms, When the keyword K related to the moving image M matches, a scene matching step for associating the scene Ms with the evaluation word H related to the keyword K, and a scene stored in the storage unit in association with the related scene Ms and evaluation word H An evaluation word storage step and the evaluation word associated with the scene Ms Video evaluation method statistics reading the evaluation word H from the storage unit to and executes the evaluation calculating step of calculating for each scene Ms, the a.

  In the present invention, the evaluation word extraction step and the scene evaluation word storage step are repeated to collect a plurality of evaluation words related to the scene Ms of the moving image M from the blog article. The statistics of the plurality of evaluation words are output as evaluation regarding the scene Ms of the moving image M.

According to this, video evaluation words obtained from blog posts are aggregated and statistics of video evaluation words are provided to viewers for each video scene. There is no need to register an evaluation. In addition, since the evaluation word is presented, it is possible to know at a glance from the evaluation word what aspect of the moving image is being evaluated. In addition, since the number of the same evaluation words is displayed, it is possible to know at a glance what aspect of the moving image is evaluated.
Furthermore, in the present invention, evaluation words are associated with each scene, and evaluation word statistics can be obtained for each scene. Since the evaluation word is presented for each scene of the moving image, only the desired scene can be selected for viewing based on the evaluation word.

(2) In the blog article search step, when the viewer of the video site selects a video on the video site, the computer sets the selected video as the video M. (1) Video evaluation method.

  In the present invention, when a viewer accessing the video site selects an arbitrary video, information on the selected video is sent to the computer of the present invention, and in the blog article search step, the selected video is the video. Set as M.

  According to this, since processing is performed only for the moving image selected by the viewer of the moving image site, it is possible to omit processing for other moving images.

  (3) In an apparatus that gives an evaluation to the moving image M, a moving image scene information storage unit that stores the association between the scene Ms of the moving image M and the keyword Ks related to the scene Ms, and a plurality of blog articles with hyperlinks to the moving image M Blog article search means for searching for, a video evaluation word extraction unit for extracting a keyword K related to the video M and an evaluation word H related to the keyword K from the plurality of searched blog articles, and a keyword Ks related to the scene Ms When the keyword K related to the moving image M matches, a moving image scene matching unit that associates the scene Ms with the evaluation word H related to the keyword K, and a moving image evaluation that stores the associated scene Ms and the evaluation word H in association with each other. A word storage unit and the evaluation word H associated with the scene Ms as the moving image evaluation word storage unit Video evaluation term extraction device characterized by comprising a statistical Luo read the assessment words H and video evaluation word output means for calculating for each scene Ms, the.

  In the present invention, a plurality of evaluation words related to the moving image M are collected from the blog article. The statistics of the plurality of evaluation words are output as evaluation regarding the scene Ms of the moving image M.

According to this, the evaluation word of the video acquired from one or more blog articles is aggregated and the statistics of the evaluation word about the video are provided to the viewer. There is no need to register. In addition, since the evaluation word is presented, it is possible to know at a glance from the evaluation word what aspect of the moving image is being evaluated. In addition, since the number of the same evaluation words is displayed, it is possible to know at a glance what aspect of the moving image is evaluated.
Furthermore, in the present invention, evaluation words are associated with each scene, and evaluation word statistics can be obtained for each scene. Since the evaluation word is presented for each scene of the moving image, only the desired scene can be selected for viewing based on the evaluation word.

  (4) A moving picture evaluation program for causing a computer to execute the method according to (1) or (2).

  According to this, the function of the method of (1) or (2) is realizable by installing said program in a computer and making it run.

  According to the present invention, a plurality of evaluation words relating to the moving image M are collected from the blog article, and statistics of the plurality of evaluation words are output as evaluations relating to the scene Ms of the moving image M. By counting and providing statistics of evaluation words regarding the moving image to the viewer, it is possible to know at a glance for each scene what level of the moving image is evaluated.

[1. First Embodiment]
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings.

[1-1. Overall structure of video evaluation system]
FIG. 1 is an overall configuration diagram of a moving image evaluation system. First, the outline of the present embodiment will be described. When a viewer who operates the browsing terminal 20 accesses a video site and selects a specific video, the video evaluation word extraction device 10 blog articles that have been posted on the blog server so far. The blog article relating to the selected video is searched from, and the evaluation word relating to the video is extracted from the searched blog article. Then, the moving image evaluation word extraction device 10 calculates the statistics of the extracted evaluation words and provides them to the moving image site. The statistics of evaluation words provided to the video site are provided to the viewing terminal 20 of the viewer who has selected the video. The viewer can automatically obtain the statistics of the evaluation words posted on the blog, and can determine the viewing of the video based on the statistics of the evaluation words.

  In FIG. 1, moving image sites A and B distribute moving images to the Internet. The moving image sites A and B distribute and reproduce the moving image selected by the viewer to the viewing terminal 20 of the viewer. In addition, the video sites A and B provide evaluation information given to the video to the viewer's browsing terminal 20 to help the viewer select a video.

  The blog servers A and B accept posting of blog articles related to the viewed videos from viewers who viewed the videos on the video sites A and B.

  The browsing terminal 20 includes a web browser, and is capable of viewing videos distributed by the video sites A and B, posting blog articles to the blog servers A and B, and the like.

  The moving image evaluation word extraction device 10 extracts evaluation words related to moving images from blog articles posted to the blog servers A and B. Also, the moving image evaluation word extraction device 10 aggregates the statistics of the extracted evaluation words and provides the result to the moving image site.

In the above description, the symbols A and B are added to indicate that there are a plurality of blog servers and moving image sites, and have no further meaning. Further, although only one browsing terminal 20 is illustrated, there are actually a plurality of browsing terminals 20.

[1-2. Configuration of video evaluation word extraction device]
FIG. 2 is a configuration diagram of the moving image evaluation word extraction device 10. The blog article acquisition unit 1 acquires a blog article about a video from a blog server. As a blog article related to a moving image, a blog article including a description of a moving image identifier (including a moving image site domain and a moving image file identifier) is assumed. In the present embodiment, the blog article acquisition unit 1 searches the blog server for a blog article related to the video selected by the viewer. When the viewer connects to the video site from the viewing terminal 20 and selects a video being distributed on the video site, the video identifier of the selected video is notified from the video site to the blog article acquisition means 1. Yes. It is conceivable that the search destination blog server accessed by the blog article acquisition unit 1 is sequentially selected from a list of blog servers registered in the storage unit in advance. The blog article acquisition unit 1 acquires a plurality of blog articles related to the selected video from one or more blog servers.

  The blog article storage unit 2 stores a plurality of blog articles acquired by the blog article acquisition unit 1 in association with video identifiers.

  The video evaluation word extraction means 3 uses a morphological analysis (natural sentence analysis) method from a blog article stored in the blog article storage unit 2 and a keyword (noun) and an evaluation word (adjective, exclamation word, negative word) related to the keyword. ) And extract. The configuration of the moving image evaluation word extraction unit 3 will be described later.

  The moving image evaluation word storage unit 4 stores the evaluation word extracted by the moving image evaluation word extraction unit 3 in association with the moving image identifier of the moving image related to the blog article as the extraction source.

  The moving image evaluation word output means 5 provides evaluation information obtained by tabulating the evaluation words stored in the moving image evaluation word storage unit 4 to a moving image site that distributes the moving image subjected to the evaluation.

  FIG. 3 is a configuration diagram of the moving image evaluation word extraction unit 3. The moving image evaluation word extraction unit 3 associates an evaluation word for each moving image scene.

  The video evaluation word extraction unit 31 uses a morphological analysis (natural sentence analysis) method from a blog article stored in the blog article storage unit 2 and an evaluation word (adjective, exclamation, negation) regarding the keyword. Word). For example, it is possible to assume that a subject of a single sentence is a keyword, and an adjective, exclamation word, or negative word included in the same sentence is an evaluation word related to the keyword.

  The moving image evaluation word holding unit (storage unit) 32 stores the keyword extracted by the moving image evaluation word extraction unit 31 and the evaluation word related to the keyword in association with the moving image identifier of the moving image related to the blog article as the extraction source.

  The moving image scene information storage unit 33 associates a moving image identifier of the moving image, a scene identifier of the scene obtained by dividing the moving image on the time axis, and a keyword related to the scene (a location or a person name appearing in the scene). Stored. The generation of the moving image scene information will be described later.

The moving image scene matching unit 34 extracts a record in which the moving image identifier and the keyword match by matching the record stored in the moving image evaluation word holding unit 32 with the record stored in the moving image scene information storage unit 33. The extracted moving image identifier, scene identifier, and evaluation word of each record are stored in the moving image evaluation word storage unit 4 in association with each other.

  FIG. 4 shows the structure of data stored in the blog article storage unit 2. The blog article storage unit 2 stores the text of the blog article and the moving image identifier associated with the blog article in association with each other. The moving image identifier is constituted by a combination of a moving image site URL and a moving image file identifier. The article body is stored as text data.

FIG. 5 shows the structure of data stored in the moving image evaluation word holding unit 32. The moving image evaluation word holding unit 32 stores a moving image identifier, a keyword K extracted from a blog article related to the moving image, and an evaluation word H related to the keyword K in association with each other. The moving image identifier is constituted by a combination of a moving image site URL and a moving image file identifier.

FIG. 6 shows the structure of data stored in the moving image scene information storage unit 33. The moving picture scene information storage unit 33 stores a moving picture identifier, a scene identifier for identifying the scene of the moving picture, and a keyword Ks related to the scene in association with each other. The moving image identifier is constituted by a combination of a moving image site URL and a moving image file identifier. The scene identifier includes the serial number of the scene and the start time and end time of the scene. The keyword Ks is not limited to one for each scene, and a plurality of keywords may be associated with one scene.

FIG. 7 shows the structure of data stored in the moving image evaluation word storage unit 4. The moving image evaluation word storage unit 4 stores a moving image identifier, a scene identifier indicating a scene of the moving image, and an evaluation word H related to the scene in association with each other. The moving image identifier is constituted by a combination of a moving image site URL and a moving image file identifier. The scene identifier includes the serial number of the scene and the start time and end time of the scene.

[1-3. Operation of the video evaluation word extraction device]
Next, the operation of this embodiment will be described. 8 and 9 are flowcharts of processing executed by the moving image evaluation word extraction device 10. FIG. 9 shows details of step S4 in FIG.

First, it is assumed that the viewer of the video site operates the viewing terminal 20 and selects the video M being distributed on the video site A. Then, the computer of the moving image site A notifies the moving image evaluation word extracting device 10 of the URL of the moving image site A and the moving image identifier of the moving image M (referred to as moving image 002) as the moving image identifier of the moving image M selected by the viewer.

In the video evaluation word extraction device 10, the blog article acquisition unit 1 acquires the video identifier of the video M notified from the video site A, and searches the blog servers A and B for a plurality of blog articles associated with the video identifier. (S1). The blog article acquisition means 1 stores the searched blog article in the blog article storage means 2 in association with the video identifier of the video M (S2). The blog article acquisition unit 1 acquires a plurality of blog articles related to the moving image M by repeating the steps S1 and S2. Here, as shown in FIG. 4, it is assumed that it blog SL related to movie M is retrieved number, it each blog SL is assumed to have been stored in association with the video identifier videos M.

Then, the video evaluation word extraction unit 31, read a blog Symbol things from a blog post storage unit 2, this blog Symbol thing or et al., By the method of morphological analysis (natural sentence analysis), and the keyword K (noun), the keyword An evaluation word H (adjective, exclamation or negative word) relating to K is extracted (S3). The moving image evaluation word extraction unit 3 stores the extracted keyword K and evaluation word H in the moving image evaluation word holding unit 32 in association with the moving image identifier of the moving image M. Here, as shown in FIG. 5, the keyword K and the evaluation word such as “real” as the evaluation word H of the keyword K “XXX” and “interesting” as the evaluation word H of the keyword K “ΔΔΔ”. Suppose that many records associated with H are obtained.

Subsequently, the moving image scene matching unit 34 matches the record stored in the moving image evaluation word holding unit 32 with the record stored in the moving image scene information storage unit 33, and the keyword K stored in the moving image evaluation word holding unit 32. And a record that matches the keyword Ks stored in the moving image scene information storage unit 33 is selected (S4, S41). The video scene matching unit 34, the record including the keyword K and Ks are matched, and video identifier videos M, and scene identifier, in association with evaluation word H is stored in the video evaluation word storing unit 4 (S4 , S42).

  Here, it is assumed that the moving image scene information storage unit 33 stores the moving image scene information shown in FIG. That is, the keyword “□□□” is associated with the scene 01 of the moving image M, the keyword “ΔΔΔ” is associated with the scene 02 of the moving image M, and the keyword “XXX” is associated with the scene 03 of the moving image M. It is assumed that “XXX” is associated and a large number of associations are stored in the same manner.

  In this case, as shown in FIG. 7, the moving image evaluation word storage unit 4 has 10 evaluations of “interesting” for the scene 02 of the moving image M and “real” for the scene 03 of the moving image M. Assume that three evaluations, for example, five evaluations of “existence” are associated with the scene 03 of the moving image M.

More specifically, referring to FIG. 5 and FIG. 6, for example, the first record of the moving image evaluation word holding unit 32 shown in FIG. 5 has the keyword K “XX”, which is shown in FIG. The keyword K “◯◯◯” matches the keyword Ks of the scene 03. Therefore, it is predicted that “XXX” obtained from the blog article is “real” and that the evaluation given to “scene 03” of the video M is particularly “scene 03” and the evaluation word “real”. Associate with. It is assumed that the information shown in FIG. 7 is stored in the moving image evaluation word storage unit 4 by such processing.

  Subsequently, the moving image evaluation word output unit 5 totals the evaluation words stored in the moving image evaluation word storage unit 4 for each scene (S5). The moving image evaluation word output means 5 totals the number of the same evaluation words associated with the same scene of the moving image M. In the example of FIG. 7, since ten evaluation words “interesting” associated with “scene 02” are stored, the evaluation word “interesting” is counted as ten for “scene 02”. Similarly, for “scene 03”, three evaluation words “real” and five evaluation words “realism” are tabulated. The moving image evaluation word output means 5 associates the moving image identifier of the moving image M with the total number of evaluation words for each scene, and outputs it to the moving image site (here, the moving image site A) where the viewer selects the moving image M. (S6). The destination video site can be known from the video identifier of video M.

  When the computer of the moving image site A receives the moving image identifier of the moving image M and the total number of evaluation words for each scene of the moving image M from the moving image evaluation word extracting device 10, the evaluation word is counted for each scene of the moving image M. Display control information for displaying the number of cases is generated and transmitted to the viewing terminal 20 of the viewer who has selected the moving image M. The computer of the browsing terminal 20 displays the page of the video site A shown in FIG. 10, for example, by receiving this display control information and executing it on the browser. In FIG. 10, the boxes described as moving images, scene 01, scene 02, and scene 03 are display columns for jackets or screen shot still images corresponding to the respective moving images. Below the screen shot of each scene, the evaluation word associated with each scene and the total number thereof are displayed. Displays that there were 10 “interesting” evaluation words for “scene 02”, and that 5 “realism” and 3 “real” were displayed as evaluation words for “scene 03”. is doing. The character font of each evaluation word is displayed with a larger number of points as the number of cases increases.

  According to the first embodiment described above, the evaluation words of the video acquired from the blog article are aggregated and statistics of the evaluation words related to the video are provided to the viewer. There is no need to register an assessment on the site. In addition, since the evaluation word is presented, it is possible to know at a glance from the evaluation word what aspect of the moving image is being evaluated. In addition, since the number of the same evaluation words is displayed, it is possible to know at a glance what aspect of the moving image is evaluated. In addition, since the evaluation word is presented for each scene of the moving image, only a desired scene can be selected for viewing based on the evaluation word.

[2. Second Embodiment]
Next, a second embodiment of the present invention will be described. In the first embodiment, when the viewer selects a video, the blog article is searched for the selected video and the evaluation words are collected. According to this, each time a viewer selects a moving image, evaluation words related to the moving image are collected, so that the viewer can always know the latest evaluation word statistics. On the other hand, since the evaluation word acquisition process is started after the viewer selects the moving image, the response from the viewer selecting the moving image to the viewer showing the evaluation word related to the moving image becomes relatively slow. Therefore, in the second embodiment to be described, an example will be described in which evaluation word collection related to a moving image and evaluation word provision to a viewer are performed asynchronously.

[2-1. Overall structure of video evaluation system]
FIG. 1 is an overall configuration diagram of a moving image evaluation system according to the present embodiment. Since this configuration is the same as that of the first embodiment, description thereof is omitted.

[2-2. Configuration of video evaluation word extraction device]
FIG. 11 is a configuration diagram of the moving image evaluation word extraction device 10 according to the second embodiment. Regardless of whether or not the viewer selects a video on the viewing terminal 20, the blog post acquisition unit 1 searches for a blog post related to the video from the blog server at an independent timing. In order to search for blog articles related to videos, it is necessary to specify a video identifier. By registering the main video site domain in the storage unit in advance, a hyperlink containing the video site domain is included. It can be assumed that the blog article is a blog article related to a moving image, and the URL of the hyperlink can be acquired as a moving image identifier.

The moving image evaluation word extraction unit 3 also operates independently of the operation timing of the blog article acquisition unit 1. In this case, in the example of information shown in FIGS. 4 to 7, records relating to a plurality of moving image sites and a plurality of moving images (that is, a plurality of different moving image identifiers) are mixed.

  In this embodiment, the moving image evaluation word output unit 5 operates when the viewer operates the viewing terminal 20 to select a moving image on the moving image site. When the viewer operates the viewing terminal 20 to select a video on the video site, the video site computer notifies the video evaluation word output means 5 of the video identifier of the video selected by the viewer. The moving image evaluation word output means 5 operates in response to this notification. When the record associated with the notified moving image identifier exists in the moving image evaluation word storage unit 4, the evaluation for each scene is performed as in the first embodiment. Count words and send the results to the video site.

  Other configurations of the moving image evaluation word extraction device 10 are the same as those in the first embodiment.

[2-3. Operation of the video evaluation word extraction device]
In the flowchart of FIG. 8, in this embodiment, the blog article acquisition unit 1 repeatedly executes the steps S1 and S2. In the repetition, the moving image M may be changed after the end of the step S2.

In addition, the moving image evaluation word extraction unit 3 repeatedly executes the steps S3 and S4 at a timing independent of the execution of the steps S1 and S2. Steps S5 and S6 are executed by the moving image evaluation word output unit 5 when the viewer operates the viewing terminal 20 and selects any moving image on the moving image site, as described above.

  Other operations are the same as those in the first embodiment.

According to the present embodiment, since the viewer selects a video on the video site and obtains an evaluation word related to the video, the video evaluation word extraction device 10 only needs to execute the processes of S5 and S6. Compared to the first embodiment, a quicker response can be realized.

  Moreover, since evaluation words relating to a moving image are collected at independent timings, it is possible to collect a large amount of evaluation words in advance compared to the first embodiment. For this reason, it is possible to know the statistics of the evaluation word with high accuracy.

[3. Video scene information generator]
In the first embodiment and the second embodiment described above, the discussion has been made on the assumption that the moving image scene information is stored in advance in the moving image scene information storage unit 33. Next, generation of moving image scene information will be described. .

[3-1. Configuration of Movie Scene Information Generation Device]
FIG. 12 is a configuration diagram of the moving image scene information generation device. The moving image acquisition means 51 acquires moving image data from a moving image site. The moving image storage unit 52 stores the moving image data acquired by the moving image acquisition unit 51 in association with the moving image identifier. The scene dividing unit 53 divides the moving image data stored in the moving image storage unit 52 into a plurality of scenes in the direction of the time axis, and associates the scene identifier of each divided scene with the moving image identifier of the original moving image to store moving image scene information. Registered in the unit 33. The scene moving image storage unit 54 stores the moving image data of each scene divided by the scene dividing unit 53 in association with the scene identifier and the moving image identifier. The keyword setting unit 55 sets a keyword related to each scene stored in the scene moving image storage unit 54 and registers the keyword in the moving image scene information storage unit 33 in association with the moving image identifier and the scene identifier of each scene.

[3-2. Scene division method]
There are several methods by which the scene dividing unit 53 divides moving image data into scenes. First, a method in which an editor operates a computer and manually divides moving image data into scenes can be considered. Also, a method of dividing the moving image data into a plurality of scenes by automatically dividing the moving image data at a preset time can be considered. In addition, a method of determining a scene division point by analyzing moving image audio is conceivable. Another possible method is to divide a scene when the attribute value of successive frame images of a moving image greatly changes beyond a certain threshold. As a technique for dividing the scene according to the time of the video, for example, and the like Japanese Open No. 2 005-210252. Moreover, as a technique for dividing a scene according to the sound of a moving image, for example , there is JP-A- 2005-260391.

[3-3. Keyword setting method]
There are several methods by which the keyword setting means 55 sets keywords according to the scene. First, a method in which an editor operates a computer and manually associates keywords with a scene can be considered. In addition, a method is conceivable in which the speech of each scene video is recognized and converted into text, frequent keywords (nouns) are extracted from the speech text by a morphological analysis (natural sentence analysis) method, and associated with the scene. In addition, a method of extracting a subtitle character string of each scene moving image, extracting frequent keywords (nouns) from the subtitle text by a morpheme analysis (natural sentence analysis) method, and associating it with the scene is conceivable.

[3-4. Application to each embodiment]
In the first embodiment, the moving image evaluation word extraction device 10 is activated when the viewer selects a moving image. For this reason, it is preferable that the moving image scene information generating device also generates moving image scene information for the selected moving image when the viewer selects the moving image. In this case, among the above-described scene dividing method and keyword setting method, a method that is manually performed by the editor is not suitable, and it is preferable to employ another automatic processing method. In this case, the moving image acquisition unit 51 may acquire moving image data from the moving image site using the moving image identifier notified to the blog article acquisition unit 1 from the moving image site.

On the other hand, in the second embodiment, since the evaluation word extraction process and the evaluation word provision process to the video site are executed asynchronously, the above-described scene division method and keyword setting method are manually performed by the editor. It can be adopted including methods. In this case, the moving image scene information generation device may process a moving image having the same moving image identifier as the evaluation word extraction processing performed by the moving image evaluation word extraction device 10, or evaluation word extraction performed by the moving image evaluation word extraction device 10. You may process the moving image of a moving image identifier different from a process. In the former case, it is conceivable that the video acquisition unit 51 receives a notification of the video identifier to be processed from the blog article acquisition unit 1 and acquires a video corresponding to the video identifier from the video site. In the latter case, it is conceivable that domains of main video sites are registered in the storage unit, and videos are selected and processed while visiting the registered domains. Since there are many popular videos listed on the top page of the video site, processing may be performed on the videos listed on the top page.

[4. Hardware configuration]
Next, a hardware configuration of a computer that operates as the moving image evaluation word extraction device 10 and the moving image scene information generation device in each of the above embodiments will be described.

FIG. 13 is a diagram illustrating a hardware configuration of a computer that operates as the moving image evaluation word extraction device 10 and the moving image scene information generation device according to the present embodiment.
This computer includes a CPU (Central Processing Unit) 310 (a plurality of CPUs such as a CPU 320 may be added in a multiprocessor configuration), a bus line 200, and a communication I / F (I / F: interface) that constitute the control unit 300. 330, a main memory 340, a BIOS (Basic Input Output System) 350, an I / O controller 360, a hard disk 370, an optical disk drive 380, and a semiconductor memory 390. The hard disk 370, the optical disk drive 380, and the semiconductor memory 390 are collectively referred to as a storage device 410.

The control unit 300 is a part that exercises control over the moving image evaluation word extraction device 10 and the moving image scene information generation device, and appropriately reads and executes various programs stored in the hard disk 370 (described later), thereby executing the above-described hardware. In cooperation with the wearer, various functions according to the above embodiment are realized.

The communication I / F 330 is a network adapter when the moving image evaluation word extraction device 10 and the moving image scene information generation device transmit / receive information to / from a blog server or a moving image site (FIG. 1) via the Internet (FIG. 1). . The communication I / F 330 may include a modem, a cable modem, and an Ethernet (registered trademark) adapter.

The BIOS 350 records a boot program executed by the CPU 310 when the moving image evaluation word extraction device 10 and the moving image scene information generation device are started, a program that the moving image evaluation word extraction device 10 and the moving image scene information generation device depend on hardware, and the like.

  A storage device 410 such as a hard disk 370, an optical disk drive 380, and a semiconductor memory 390 can be connected to the I / O controller 360.

The hard disk 370 stores various programs for causing the hardware to function as the moving image evaluation word extraction device 10 and the moving image scene information generation device, programs for executing the functions of the present invention, and data. Note that the moving image evaluation word extraction device 10 and the moving image scene information generation device can use a hard disk (not shown) separately provided as an external storage device.

As the optical disk drive 380, for example, a DVD-ROM drive, a CD-ROM drive, a DVD-RAM drive, and a CD-RAM drive can be used. In this case, the optical disk 400 corresponding to each drive is used. A program or data can be read from the optical disk 400 by the optical disk drive 380 and provided to the main memory 340 or the hard disk 370 via the I / O controller 360.

Note that the computer in the present invention refers to an information processing device including a storage device, a control unit, and the like, and the moving image evaluation word extraction device 10 and the moving image scene information generation device include a storage device 410, a control unit 300, and the like. An information processing apparatus is included, and this information processing apparatus is included in the concept of the computer of the present invention.

In the moving image evaluation word extraction device 10 and the moving image scene information generation device, the control unit 300 corresponds to each unit, and the storage area of the storage device 410 corresponds to each storage unit.

The computer of the blog server, the computer of the moving image site, and the computer of the browsing terminal shown in FIG. 1 have the same configuration as the computer hardware described above.

  As mentioned above, although embodiment of this invention was described, this invention is not restricted to embodiment mentioned above. For example, in each of the above embodiments, the video identifier of the video selected by the viewer is acquired from the video site, but may be acquired from the viewer's viewing terminal. The effects described in the embodiments of the present invention are only the most preferable effects resulting from the present invention, and the effects of the present invention are limited to those described in the embodiments of the present invention. is not.

It is a figure which shows the whole structure of the moving image evaluation system containing the moving image evaluation word extraction device of this invention. It is a block diagram of the moving image evaluation word extraction apparatus in 1st Embodiment of this invention. Is a block diagram of a video evaluation word extraction means shown in FIGS. 2 and 12. It is a data structure figure of the blog article storage part shown in FIG. It is a data structure figure of the moving image evaluation word holding | maintenance part shown in FIG. It is a data structure figure of the moving image scene information storage part shown in FIG. It is a data structure figure of the moving image evaluation word memory | storage part shown in FIG. It is a flowchart of the process which a moving image evaluation word extraction device performs. It is a flowchart of S4 step shown in FIG. It is a display block diagram which shows the example of a display of the moving image evaluation in the browsing terminal of FIG. It is a block diagram of the moving image evaluation word extraction apparatus in 2nd Embodiment of this invention. It is a block diagram of a moving image scene information generation device. It is a hardware block diagram of the computer which operate | moves as the moving image evaluation word extraction apparatus 10 and a moving image scene information generation apparatus.

DESCRIPTION OF SYMBOLS 1 Blog article acquisition means 2 Blog article storage part 3 Video evaluation word extraction means 4 Video evaluation word storage part 5 Video evaluation word output means 10 Video evaluation word extraction device 20 Viewing terminal 31 Video evaluation word extraction part 32 Video evaluation word holding part 33 Movie scene information storage unit 34 Movie scene matching unit

Claims (4)

In the method of giving an evaluation to the video M,
Storing the association between the scene Ms of the video M and the keyword Ks related to the scene Ms in the storage unit;
Computer
A blog article search step for searching a plurality of blog articles with hyperlinks to the video M;
An evaluation word extraction step of extracting a keyword K related to the moving image M and an evaluation word H related to the keyword K from the plurality of searched blog articles;
A scene matching step for associating the scene Ms with the evaluation word H for the keyword K when the keyword Ks for the scene Ms and the keyword K for the moving image M match;
A scene evaluation word storage step of associating the associated scene Ms and the evaluation word H and storing them in the storage unit;
An evaluation calculation step of reading the evaluation word H associated with the scene Ms from the storage unit and calculating statistics of the evaluation word H for each scene Ms;
A video evaluation method characterized by executing In the blog article search step, the computer
The video evaluation method according to claim 1, wherein when a viewer of the video site selects a video on the video site, the selected video is set as the video M. In the device that gives evaluation to the video M,
A moving image scene information storage unit that stores the association between the scene Ms of the moving image M and the keyword Ks related to the scene Ms;
Blog article search means for searching a plurality of blog articles with hyperlinks to the video M,
A moving image evaluation word extraction unit that extracts a keyword K related to the moving image M and an evaluation word H related to the keyword K from the plurality of searched blog articles;
A moving image scene matching unit that associates the scene Ms and the evaluation word H related to the keyword K when the keyword Ks related to the scene Ms matches the keyword K related to the moving image M;
A moving image evaluation word storage unit that stores the associated scene Ms and the evaluation word H in association with each other;
Video evaluation word output means for reading the evaluation word H associated with the scene Ms from the video evaluation word storage unit and calculating statistics of the evaluation word H for each scene Ms;
A video evaluation word extraction device characterized by comprising:   A moving image evaluation program for causing a computer to execute the method according to claim 1.

Images