JP5916790B2 - Video playback device, candidate extraction method and program - Google Patents

Description

  The present invention relates to a moving image reproducing apparatus, a candidate extracting method, and a program for extracting comment input candidates for moving image data.

  In recent years, moving image content is distributed not only to PCs (Personal Computers) but also to mobile phone terminals, televisions, and the like. There has also been proposed a technique for accepting posting of a comment corresponding to the playback time of the moving image and displaying the moving image and the comment in accordance with the moving image playback time. Patent Document 1 discloses a technique for presenting input candidates to a user who inputs a comment according to a moving image to be reproduced. Specifically, Patent Document 1 discloses a technology in which comments for moving image data are clustered, and comments included in a cluster having a large number of comments belong to the input candidates.

JP2013-229873A

However, with the technique described in Patent Document 1, if the number of comments for moving image data is small, there is a possibility that an appropriate input candidate cannot be presented.
An object of the present invention is to provide a moving image reproducing apparatus, a candidate extracting method, and a program for presenting appropriate input candidates even when the number of comments for moving image data is small.

In the first aspect, the moving image storage unit stores the moving image data selected for viewing by the user in the past and the comment on the moving image data in association with each other, and the moving image reproduction for reproducing the moving image data stored in the moving image storage unit. And a candidate extraction unit that extracts comment input candidates for moving image data reproduced by the moving image reproduction unit from comments for other moving image data stored in the moving image storage unit .

In addition, the third mode is a moving image reproducing unit that reproduces moving image data, and for each comment on other moving image data selected for viewing by the user in the past, the moving image data in which a comment similar to the comment appears. A score calculation unit that calculates a score according to the number, and a comment for the video data that the video playback unit plays back a comment that is given by the score calculation unit out of comments for the other video data It is a moving image reproducing apparatus provided with the candidate extraction part extracted as an input candidate.

Moreover, a 4th aspect is a 3rd aspect. WHEREIN: The said score calculation part appears the comment similar to the said comment about each comment with respect to the said other moving image data selected for viewing by the user in the past. This is a moving image reproducing apparatus that calculates a score obtained by multiplying the number of moving image data by a weight corresponding to the viewing time of the moving image data.

Further, a fifth aspect is the third or fourth aspect, in which the score calculation unit is similar to the comment for each comment on the other video data selected for viewing by the user in the past. Is a moving image reproduction device that adds a monotonically increasing score to the number of comments when a plurality of characters appear in the same moving image data.

Further, the sixth aspect is similar to the comment on the moving image data reproduced by the moving image reproducing unit among the input candidates extracted by the candidate extracting unit in the first aspect or any of the third to fifth aspects. This is a moving image playback apparatus including a candidate specifying unit that specifies a candidate to be input as a comment input candidate for the moving image data.

According to a seventh aspect, there is provided a moving image storage unit that stores the moving image data selected for viewing by the user in the past and a comment on the moving image data in association with the moving image data stored in the moving image storage unit. This is a program for functioning as a candidate extracting unit for extracting a moving image reproducing unit to be reproduced and a comment input candidate for moving image data to be reproduced by the moving image reproducing unit from comments for other moving image data stored in the moving image storage unit .

In addition, according to the eighth aspect, a video playback unit that plays back video data, a video in which a comment similar to the comment appears for each comment on other video data selected for viewing by the user in the past. A score calculation unit that calculates a score according to the number of data, and a comment for video data that is reproduced by the video reproduction unit when a score given by the score calculation unit is greater than or equal to a predetermined value among comments for the other video data It is a program for functioning as a candidate extraction part which extracts as an input candidate.

  According to at least one of the above aspects, the moving image reproduction device generates input candidates based on comments on moving image data reproduced in the past. Thereby, the moving image reproducing apparatus can present an appropriate input candidate even when the number of comments for the moving image data to be reproduced is small.

It is a schematic block diagram which shows the structure of the comment delivery system which concerns on one Embodiment. It is a schematic block diagram which shows the structure of the terminal device which concerns on one Embodiment. It is a flowchart which shows the operation | movement at the time of the moving image data acquisition of the terminal device which concerns on one Embodiment. It is a flowchart which shows the operation | movement at the time of the moving image data reproduction | regeneration of the terminal device which concerns on one Embodiment. It is a figure which shows an example of an input interface. It is a schematic block diagram which shows the structure of the computer which concerns on at least 1 embodiment.

Hereinafter, embodiments will be described in detail with reference to the drawings.
FIG. 1 is a schematic block diagram illustrating a configuration of a comment delivery system 1 according to an embodiment.
The comment distribution system 1 includes a web server 100, a moving image distribution server 200, a comment distribution server 300, and a plurality of terminal devices 500. The moving image distribution server 200, the comment distribution server 300, and the plurality of terminal devices 500 are connected via a network N, respectively.

  The web server 100 provides a web page for displaying moving image data. The web server 100 stores metadata of each moving image data. In the present embodiment, the metadata of moving image data is a character string (moving image tag) indicating the classification of the moving image data. The web page provided by the web server 100 includes a URL (Uniform Resource Locater) for acquiring moving image data, a URL for acquiring a comment for the moving image data, a URL for acquiring a comment input candidate, Contains metadata for video data.

  The moving image distribution server 200 stores a plurality of moving image data. The moving image distribution server 200 distributes moving image data to the requesting terminal device 500 in response to a distribution request from the terminal device 500. This distribution is performed by streaming distribution, for example.

  The comment distribution server 300 receives a comment on the moving image data distributed by the moving image distribution server 200 from the terminal device 500. The comment distribution server 300 stores the comment in association with the ID of the video data, the time when the comment is input, the playback time for displaying the comment, the user ID of the user who input the comment, and the display method of the comment. To do. The comment distribution server 300 distributes the comment associated with the ID of the moving image data to each terminal device 500 browsing the moving image data.

  The terminal device 500 displays a web page provided from the web server 100. Thereby, the terminal device 500 displays the comment distributed from the comment distribution server 300 by superimposing it on the moving image data distributed from the moving image distribution server 200. Further, the terminal device 500 stores dictionary data for storing input candidates for moving image data, and displays comment input candidates for moving image data based on the dictionary data. The terminal device 500 is realized by an information processing device such as a television, a PC, a mobile terminal, or a game machine. The terminal device 500 is an example of a moving image playback device.

FIG. 2 is a schematic block diagram illustrating a configuration of the terminal device 500 according to an embodiment.
The terminal device 500 includes an input unit 501, a communication unit 502, a moving image recording unit 503, a moving image storage unit 504, a candidate extraction unit 505, a score calculation unit 506, a dictionary recording unit 507, a dictionary storage unit 508, a candidate specifying unit 509, and display control. Unit 510, display unit 511, comment recording unit 512, and comment storage unit 513.
The input unit 501 receives input from the user via the input device. Examples of the input device include a mouse and keyboard of a PC, a key and touch panel of a mobile phone terminal, an infrared remote control of a television, and the like.
The communication unit 502 communicates with the web server 100, the moving image distribution server 200, and the comment distribution server 300 via the network N.
The moving image recording unit 503 records the moving image data received from the moving image distribution server 200 in association with the comment received from the comment distribution server 300 and the playback time of the comment in the moving image storage unit 504. Note that the moving image recording unit 503 records the moving image data selected for viewing by the user in the moving image storage unit 504.
The moving image storage unit 504 stores moving image data reproduced in the past, comments, and the reproduction time of the comments.
The candidate extraction unit 505 extracts comment input candidates based on the comments stored in the moving image storage unit 504.
The score calculation unit 506 calculates a score indicating the priority of presentation for each input candidate extracted by the candidate extraction unit 505.
The dictionary recording unit 507 records each input candidate extracted by the candidate extraction unit 505 in the dictionary storage unit 508 in association with the priority calculated by the score calculation unit 506.
The dictionary storage unit 508 stores dictionary data in which input candidates are associated with scores.
The candidate specifying unit 509 specifies a comment input candidate for the reproduced moving image data, based on the comment received from the comment distribution server 300 and the dictionary data stored in the dictionary storage unit 508.
Based on the web page received from the web server 100, the display control unit 510 generates display data for displaying the moving image data, the comments, and the input candidates in a superimposed manner. The display control unit 510 is an example of a moving image reproducing unit that reproduces moving image data.
The display unit 511 displays the display data generated by the display control unit 510 on the display.
The comment recording unit 512 records the comment input by the user in the comment storage unit 513.

  Here, the operation of the terminal device 500 will be described. The terminal device 500 according to the present embodiment records moving image data and comments to be viewed in advance in the moving image storage unit 504, and reproduces the moving image data stored in the moving image storage unit 504 at an arbitrary timing thereafter. .

FIG. 3 is a flowchart illustrating an operation of the terminal device 500 according to an embodiment when acquiring moving image data.
When the terminal device 500 is connected to the network N, the user operates the terminal device 500 and records moving image data in the terminal device 500.
First, the input unit 501 receives an input of a moving image list display instruction from the user (step S101). When a moving image list display instruction is input to the input unit 501, the communication unit 502 transmits a moving image list transmission request to the web server 100. When the communication unit 502 receives the moving image list from the web server 100 (step S102), the display unit 511 displays the moving image list on the display (step S103). In the moving image list, hyperlinks to URLs of web pages for displaying the moving image data are provided in association with information of the plurality of moving image data.

  Next, the input unit 501 receives an input of a URL of a web page for displaying moving image data from the user (step S104). The user inputs a URL by selecting a hyperlink to the URL of a web page for displaying desired video data from the video list displayed on the display unit 511. That is, the user selects moving image data for viewing. When a URL is input to the input unit 501, the communication unit 502 receives a web page for displaying moving image data from the web server 100 according to the URL (step S105). The web page includes a URL for acquiring moving image data, a URL for acquiring a comment, and a URL for acquiring a comment input candidate.

Next, the moving image recording unit 503 acquires moving image data, comments, and dictionary data via the communication unit 502 in accordance with the URL included in the web page (step S106). That is, the communication unit 502 receives moving image data from the moving image distribution server 200 according to the URL. The communication unit 502 receives a comment from the comment distribution server 300 according to the URL. Then, the moving image recording unit 503 records the moving image data and comments received by the communication unit 502 in the moving image storage unit 504 in association with each other (step S107).
As a result, the terminal device 500 allows the moving image data and comments selected by the user to be reproduced even when not connected to the network N.

  Next, the candidate extraction unit 505 reads comments and reproduction times for all the moving image data stored in the moving image storage unit 504 (step S108). Here, all the moving image data stored in the moving image storage unit 504 is recorded by a user operation. Therefore, the moving image data stored in the moving image storage unit 504 reflects the user's preference. Next, the candidate extraction unit 505 analyzes each comment and generates a parameter capable of calculating the similarity between the comments (step S109). For example, the candidate extraction unit 505 divides each comment in units of a predetermined number of characters (for example, two characters) (N-Gram), and a parameter that can calculate a similarity between a character string of the predetermined number of characters and a vector indicating the number of appearances. Generate as

  Specifically, the candidate extraction unit 505 uses the character string “No” once, the character string “Hi” once, the character string “Ah” three times, A vector indicating that the string “Aw” appears once and the character string “ww” appears five times is generated. Thereby, the similarity between comments is calculated by calculating the cosine similarity between vectors, for example. When calculating the similarity between vectors, it is possible to calculate the similarity more appropriately by weighting the number of occurrences of each character string so as to monotonously decrease with respect to the total appearance frequency of the character string. (For example, TF / IDF method). Specifically, since the character string “ww” and the character string “Oh” are frequent, the weight is relatively small with respect to other character strings. When calculating the degree of similarity using such weighting, the degree of similarity between the character strings “suee wwwwwwwwwww” and “Yamero wwwwwwwwwww” is lower than when no weighting is performed. On the other hand, when calculating the degree of similarity using such weighting, the degree of similarity between the character strings “Yamero wwwwwwwwww” and “Yamero w” is higher than when weighting is not performed.

  Next, the candidate extraction unit 505 generates a group including a plurality of similar comments from the comments acquired by the comment acquisition unit (step S110). For example, the candidate extraction unit 505 extracts a combination of comments whose vector cosine similarity is a predetermined threshold (for example, 0.9) or more as a combination of similar comments, and identifies a group of a plurality of similar comments. To do. At this time, the candidate extraction unit 505 determines that when two similar comments exist at a predetermined time (for example, 4 seconds) or more apart and there is no other similar comment between the two comments, the two comments As belonging to another group. As a result, the candidate extraction unit 505 can handle comments that are composed of the same character string and appear concentrated at different times (for example, the first and second lyrics) as separate groups. . The candidate extraction unit 505 assigns a unique ID for each group.

  The candidate extraction unit 505 extracts representative comments included in each identified group (step S111). Typical comments include, for example, those having the highest average cosine similarity with vectors related to other comments in the same group. Next, the score calculation unit 506 calculates a score indicating the priority of presentation for each group extracted by the candidate extraction unit 505 (step S112).

Specifically, when the score calculation target group includes comments for different video data, the score calculation unit 506 calculates a score obtained by multiplying the basic point (for example, 100 points) by the number of the video data, Add to the score of the group. That is, the score calculated by the score calculation unit 506 monotonously increases with respect to the number of moving image data in which similar comments appear.
When the score calculation target group includes a plurality of comments for the same video data, the score calculation unit 506 calculates a score obtained by multiplying the basic point (for example, 10 points) by the number of comments for the same video data. , Add to the score of the group. That is, the score calculation unit 506 adds a monotonically increasing score to the number of comments when a plurality of similar comments appear in the same video data.
On the other hand, the score calculation unit 506 does not give a score for other comments. In this embodiment, the basic point of the score added to a group including a comment that appears in common in different video data is added to a group including a comment that appears in the same video data. It should be higher than the basic score. The score calculation unit 506 normalizes the score of each group.
As a result, comments that tend to appear in common with videos viewed by the user have a higher score (closer to 1), and comments that are less likely to appear have a lower score (closer to 0).

  Note that in another embodiment, the score calculated by the score calculation unit 506 may be reduced as time passes for viewing the moving image. Specifically, the score calculation unit 506 calculates, for each moving image data in which a similar comment appears, a score that increases as the posting time of the similar comment in the moving image data becomes new, and each similar comment appears. The sum of moving image data scores may be calculated as a comment score. Thereby, the score is reduced as the comment is on the moving image data with the oldest viewing time. That is, it is possible to promptly comment on moving image data of a genre in which the user has been enthusiastic in the past and then the user's interest has diminished from among the comment candidates.

Next, the candidate extraction unit 505 extracts a predetermined number (for example, 20) of comments with the highest score of the group to which the comment belongs from among the extracted representative comments as input candidates (Steps). S113). That is, the dictionary recording unit 507 extracts, as input candidates, a score assigned by the score calculation unit that is equal to or higher than the 20th score (predetermined value) from the top. Then, the dictionary recording unit 507 records the input candidates extracted by the candidate extraction unit 505 in the dictionary storage unit 508 in association with the score (step S114).
Through the above-described processing, the terminal device 500 records dictionary data of input candidates in the dictionary storage unit 508. Also, as a result, a comment having little commonness with other moving image data is deceived from the dictionary data.

When the terminal device 500 records the moving image data and the comments in the moving image storage unit 504, the terminal device 500 reproduces the moving image data stored in the moving image storage unit 504 by a user operation. At this time, the terminal device 500 may not be connected to the network N.
FIG. 4 is a flowchart showing an operation of the terminal device 500 according to an embodiment when reproducing moving image data.
The display control unit 510 generates display data including a list of moving image data stored in the moving image storage unit 504. The display unit 511 displays the display data generated by the display control unit 510 on the display (step S201). Next, the input unit 501 receives an input for selecting moving image data to be reproduced from the user (step S202). When selection of moving image data to be reproduced is input to the input unit 501, the display control unit 510 reads the selected moving image data and a comment for the moving image data from the moving image storage unit 504 (step S203).

  Next, the display control unit 510 generates display data for displaying a comment superimposed on the read moving image data. The comment received from the comment distribution server 300 is associated with the reproduction time of the moving image data. The display control unit 510 starts displaying the comment when the reproduction time of the moving image data matches the reproduction time associated with the comment. Then, the display control unit 510 hides the comment when a predetermined time has elapsed from the playback time when the display of the comment is started. The display unit 511 displays the display data generated by the display control unit 510 on the display (step S204).

  Here, the input unit 501 receives a comment input instruction from the user (step S205). Specifically, when a predetermined key (for example, a determination key) of the input device is pressed, the input unit 501 determines that the pressing is a comment input instruction. When a comment input instruction is input to the input unit 501 (step S205: YES), the candidate specifying unit 509 analyzes each comment on the moving image data to be reproduced and stored in the moving image storage unit 504, and between the comments. A parameter capable of calculating the similarity is generated (step S206). For example, the candidate specifying unit 509 divides each comment in units of a predetermined number of characters (for example, two characters) (N-Gram), and a parameter that can calculate a similarity between a character string of the predetermined number of characters and a vector indicating the number of appearances. Generate as

  Next, the candidate specifying unit 509 generates a group including a plurality of similar comments from the comments received by the communication unit 502 (step S207). At this time, if the two similar comments exist at a predetermined time (for example, 4 seconds) or more apart and there is no other similar comment between the two comments, the candidate specifying unit 509 As belonging to another group.

  The candidate specifying unit 509 specifies a generated group in which the number of comments included in the group is a predetermined number (for example, 5) or more (step S208). Next, the candidate specifying unit 509 calculates a presentation period when a comment related to each group specified in step S211 is set as an input candidate (step S209). Specifically, the candidate specifying unit 509 calculates the presentation period based on the average value of the reproduction times in which comments included in each group appear and the degree of distribution.

  The candidate specifying unit 509 specifies a group associated with the presentation period including the current playback time of the moving image data (step S210). The candidate specifying unit 509 specifies a representative comment included in each group specified in step S210 as an input candidate (step S211). As a result, the candidate specifying unit 509 can list comments that are frequently input in the moving image data currently being reproduced as input candidates.

The candidate identifying unit 509 determines whether or not the number of identified input candidates is equal to or greater than a predetermined minimum number of presentations (step S212). That is, the candidate specifying unit 509 determines whether or not the number of input candidates presented to the user satisfies the minimum number of presentations.
If the candidate specifying unit 509 determines that the number of input candidates is less than the minimum number of presentations (step S212: NO), the candidate specifying unit 509 reproduces the input candidates included in the dictionary data stored in the dictionary storage unit 508. Those having high similarity to the comment on the moving image data are identified as input candidates to be presented to the user (step S213). That is, the candidate specifying unit 509 lists, as input candidates, comments that frequently appear in the moving image data stored in the moving image storage unit, among the comments that are not frequent with respect to the moving image data being reproduced. Accordingly, the candidate specifying unit 509 can present an appropriate input candidate to the user even when the number of comments for the moving image data being reproduced is small.

  When the candidate identification unit 509 determines that the number of input candidates has reached the minimum number of presentations (step S212: YES), or when the input candidate is identified based on the dictionary data in step S213, the display control unit 510 generates display data including an input interface of comment input candidates superimposed on the moving image data and the comment (step S214). Then, the display unit 511 displays the display data on the display. Next, the input unit 501 receives an input of a comment from the user (step S215).

FIG. 5 is a diagram illustrating an example of the input interface.
The input interface according to the present embodiment includes an image of an input device (in FIG. 4, a determination key arranged at the center of the direction key and the direction key), input candidates corresponding to the input device key, and comments being input. And a preview area for displaying a preview. In the example illustrated in FIG. 4, the input candidate “w” is associated with the upper key, the input candidate “Nuko” is associated with the left key, and the input candidate “Upotsu” is associated with the right key. The lower key is associated with a “character change” function for switching input candidates to be presented. In addition, the determination key is associated with a function of inputting a character string that is a combination of input candidates input so far as a comment. As a result, the user can input the comment “Uptsu www” by pressing the upper key three times after pressing the right key and then pressing the enter key.

  When the input unit 501 receives an input of a comment from the user, the comment recording unit 512 records the comment in the comment storage unit 513 in association with the reproduction time of the moving image data when input (step S216). Note that when the terminal device 500 is connected to the network N, the communication unit 502 may transmit the comment stored in the comment storage unit 513 to the comment distribution server 300. Further, the display control unit 510 generates display data that does not include an input interface.

  When there is no input of a comment input instruction from the user (step S205: NO) or when a comment is transmitted in step S216, the display control unit 510 determines whether or not the reproduction of the moving image data has ended (step S217). ). The display control unit 510 determines that the reproduction of the moving image data has ended when the moving image data has been reproduced to the end or when the stop button has been pressed. If the display control unit 510 determines that the reproduction of the moving image data has not ended (step S217: NO), the display control unit 510 returns to step S205 and continues to accept the comment input instruction. On the other hand, when it is determined that the reproduction of the moving image data has ended (step S217: YES), the display control unit 510 ends the process. Thereafter, when the terminal device 500 is connected to the network N, the communication unit 502 transmits the comment stored in the comment storage unit 513 to the comment distribution server 300.

  Thereby, the terminal device 500 can easily input a comment to the moving image data. In particular, according to this embodiment, even when the terminal device 500 is difficult to input a character string such as a television, the user can easily input a comment. In addition, the comment distribution system 1 according to the present embodiment generates comment input candidates for moving image data based on comments for other moving image data. Thereby, the comment delivery system 1 which concerns on this embodiment can present an appropriate input candidate, when there are few comments with respect to moving image data.

As described above, the embodiment has been described in detail with reference to the drawings. However, the specific configuration is not limited to that described above, and various design changes and the like can be made.
For example, in the present embodiment, a case has been described in which the terminal device 500 generates dictionary data of input candidates based on comments stored in the moving image storage unit 504, but the present invention is not limited to this. For example, in another embodiment, the terminal device 500 reads the user's viewing history and favorite list from the web server 100 via the communication unit 502, and generates dictionary data based on comments on the moving image data related to the viewing history and favorite list. You may do it. The moving image data included in the user's viewing history and favorite list is moving image data selected for viewing by the user in the past. Further, the terminal device 500 may generate dictionary data based on a viewing history recorded locally and a comment on moving image data included in the favorite list.

  Further, the terminal device 500 may extract a comment input candidate from a comment on the moving image data selected for viewing by another user who has a high commonality with the user who owns the terminal device 500. Other users with high commonality of preference include other users who belong to the same community as the user who owns the terminal device 500, other users who share attributes such as age and gender with the user who owns the terminal device 500, and accuracy Other users included in the combination of user clusters generated for the purpose of improvement are listed. Similarly, in another embodiment, the terminal device 500 is viewed by another user who has a high commonality between the comment on the moving image data selected for viewing by the user who owns the terminal device 500 and the user. Comment input candidates may be extracted from the comment on the moving image data selected for the purpose. That is, the “user” is not limited to “a user who owns the terminal device 500”, but “a user who owns the terminal device 500” such as “another user who has a high commonality with the user who owns the terminal device 500”. It is a concept including other users.

  In this embodiment, the moving image recording unit 503 records moving image data and comments to be reproduced in the moving image storage unit 504, and the display control unit 510 reproduces the moving image data and comments stored in the moving image storage unit 504. explained. Thereby, the terminal device 500 can generate dictionary data of input candidates even when the terminal device 500 is not connected to the network N. On the other hand, the configuration of the terminal device 500 is not limited to this. For example, in another embodiment, the display control unit 510 directly acquires a moving image and a comment via the communication unit 502 without performing buffering by the moving image storage unit 504. Also good.

  Moreover, although the terminal device 500 demonstrated the case where the terminal device 500 extracted an input candidate based on the comment with respect to all the moving image data which the moving image memory | storage part 504 memorize | stores, it is not restricted to this. For example, in another embodiment, input candidates may be extracted based on comments on moving image data similar to moving image data to be reproduced among moving image data stored in the moving image storage unit 504. The terminal device 500 can specify similar moving image data based on the metadata of the moving image data. Examples of metadata include a moving image tag, a title, and a moving image description. In addition, the terminal device 500 analyzes the moving image data, specifies other moving image data using the same music as the moving image data to be reproduced, and other moving image data including a similar image as similar moving image data. Also good.

  Moreover, although this embodiment demonstrated the case where the terminal device 500 extracted the comment contained in moving image data as an input candidate, it is not restricted to this. For example, in another embodiment, the terminal device 500 may extract a part of comments included in the moving image data as input candidates. For example, the letter “w” may be used at the end of other words to express laughing. The large number of letters “w” expresses the degree of laughter. That is, the letter “w” is included in the comment together with other words, and the number of the letters differs for each comment. Therefore, in other embodiments, such characters and character strings may be extracted as input candidates. Specifically, the candidate extraction unit 505 may perform morphological analysis on the comments stored in the moving image storage unit 504 and extract each morpheme acquired by morphological analysis as an input candidate. In this case, the candidate extraction unit 505 can extract a known word as an input candidate. In another embodiment, the candidate extraction unit 505 extracts a character string from each comment for each of a plurality of different numbers of characters (N-Gram), and extracts input candidates based on the number of appearances and the number of characters of the character string. Also good. For example, the candidate extraction unit 505 calculates a parameter that monotonously increases with respect to the number of appearances and the number of characters, and extracts characters or character strings whose parameter values are greater than a predetermined value as input candidates. In this case, the candidate extraction unit 505 can extract not only known words but also idioms and unknown words as input candidates.

  Moreover, although the candidate extraction part 505 demonstrated the case where the similarity of a comment was calculated by N-Gram analysis in this embodiment, it is not restricted to this. For example, in another embodiment, the candidate extraction unit 505 may calculate the similarity using a vector generated based on a morpheme extracted by morpheme analysis. In another embodiment, the candidate extraction unit 505 may calculate the similarity based on the Levenshtein distance between comments.

FIG. 6 is a schematic block diagram illustrating a configuration of a computer 900 according to at least one embodiment.
The computer 900 includes a CPU 901, a main storage device 902, an auxiliary storage device 903, and an interface 904.
Each of the terminal devices 500 described above is mounted on a computer 900. The operation of each processing unit described above is stored in the auxiliary storage device 903 in the form of a program. The CPU 901 reads a program from the auxiliary storage device 903, develops it in the main storage device 902, and executes the above processing according to the program. In addition, the CPU 901 ensures a storage area corresponding to the above-described storage unit in the main storage device 902 according to the program.

  In at least one embodiment, the auxiliary storage device 903 is an example of a tangible medium that is not temporary. Other examples of the tangible medium that is not temporary include a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, and a semiconductor memory connected via the interface 904. When this program is distributed to the computer 900 via the network N, the computer 900 that has received the distribution may develop the program in the main storage device 902 and execute the above processing.

  The program may be for realizing a part of the functions described above. Further, the program may be a so-called difference file (difference program) that realizes the above-described function in combination with another program already stored in the auxiliary storage device 903.

  DESCRIPTION OF SYMBOLS 1 ... Comment delivery system 100 ... Web server 200 ... Movie delivery server 300 ... Comment delivery server 500 ... Terminal device 501 ... Input part 502 ... Communication part 503 ... Movie recording part 504 ... Movie storage part 505 ... Candidate extraction part 506 ... Score calculation 507 ... Dictionary recording unit 508 ... Dictionary storage unit 509 ... Candidate specifying unit 510 ... Display control unit 511 ... Display unit 512 ... Comment recording unit 513 ... Comment storage unit 900 ... Computer 901 ... CPU 902 ... Main storage device 903 ... Auxiliary storage Device 904 ... Interface N ... Network

Claims (7)

A moving image storage unit that stores the moving image data selected for viewing by the user in the past and a comment on the moving image data in association with each other;
A moving image reproducing unit for reproducing the moving image data stored in the moving image storage unit ;
A moving image reproducing apparatus comprising: a candidate extracting unit that extracts comment input candidates for moving image data reproduced by the moving image reproducing unit from comments for other moving image data stored in the moving image storage unit . A video playback unit for playing back video data;
For each comment on other video data selected for viewing by the user in the past, a score calculation unit that calculates a score according to the number of video data in which a comment similar to the comment appears ,
A candidate extracting unit that extracts a comment given to the other moving image data by a score given by the score calculating unit equal to or greater than a predetermined value as a comment input candidate for the moving image data reproduced by the moving image reproducing unit;
A video playback device comprising: For each comment on the other video data selected for viewing by the user in the past, the score calculation unit responds to the number of video data in which a comment similar to the comment appears according to the viewing time of the video data. The moving image reproducing apparatus according to claim 2 , wherein a score obtained by multiplying the weight is calculated. For each comment on the other video data selected for viewing by the user in the past, the score calculation unit determines the number of comments when a plurality of comments similar to the comment appear in the same video data. video playback apparatus according to claim 2 or claim 3 adds a score monotonically increasing for. The candidate specifying part which specifies the thing similar to the comment with respect to the moving image data which the said moving image reproducing part reproduce | regenerate among the input candidates extracted by the said candidate extraction part as an input candidate of the comment with respect to the said moving image data is provided. Item 5. The moving image playback device according to any one of Items 4 to 4 .   Computer
  A moving image storage unit that stores the moving image data selected for viewing by the user in the past and a comment on the moving image data in association with each other;
  A moving image reproducing unit for reproducing the moving image data stored in the moving image storage unit;
  Candidate extraction unit for extracting comment input candidates for moving image data reproduced by the moving image reproduction unit from comments for other moving image data stored in the moving image storage unit
  Program to function as. Computer
A video playback unit that plays back video data,
For each comment on other video data selected for viewing by the user in the past, a score calculation unit that calculates a score according to the number of video data in which a comment similar to the comment appears,
For functioning as a candidate extraction unit that extracts a comment given to the other video data as a candidate for inputting a comment for video data reproduced by the video reproduction unit , with a score given by the score calculation unit equal to or greater than a predetermined value program.

Images