JP5608570B2 - Related content search apparatus and program - Google Patents

Description

  The present invention relates to a related content search device and a program.

  With the widespread use of VOD (video on demand), users can view content such as television programs without being limited by time. Many VOD sites allow content searches by genre or keyword. For example, on a video site such as YouTube (http://www.youtube.com/, YouTube, LLC), a user can search for and view a target video using a keyword or the like.

  On the other hand, there is a technique for searching for content using text information written on a bulletin board or blog on the Internet. For example, in the technique of Patent Document 1, user-written information is associated with a recorded moving image, and a cueing signal or the like is automatically recorded in the content using the appearance frequency of a character string included in the written information. To do. In the technique of Patent Document 2, the content of a user's utterance on a bulletin board or the like is stored together with reference information for specifying a program or a scene, and a corresponding scene is searched from recorded content based on the content of the utterance. .

JP 2008-244657 A JP 2007-193827 A

  When a content is searched for by a specific keyword, the content is deeply related to the keyword because the keyword is emphasized. If the user's purpose is clear and the desired content image is also clear, such a keyword search is effective. However, the content obtained as a search result of content search using keywords is limited to information related to information that the user has in advance. The keywords used by the user for the search are words within the range that the user can conceive, and content search using such keywords rarely yields “unexpected content” for the user. Therefore, it can be said that it is insufficient from the viewpoint of expanding the user's viewpoint.

  The technique of Patent Document 1 described above performs a search using information written by a plurality of users, but adds index information for searching a certain scene in the content, and does not search the content itself. . Moreover, the technique of patent document 2 searches the user's utterance linked | related with the program scene using keyword search etc., and does not provide the result of having searched the content from the new viewpoint for the user.

  The present invention has been made in consideration of such circumstances, and provides a related content search apparatus and program capable of searching content that matches user preferences from a new viewpoint.

[1] According to one aspect of the present invention, an information text including any one or more of the content summary sentence, subtitles, and manuscript representing the content is stored in association with the content information. A content storage unit, a write storage unit that stores identification information of the subject , and writing content data indicating an impression or an opinion on the subject, and a subject designation reception that receives information specifying the subject from a plurality of user terminals A writing content data received from the user terminal, and the received writing content data and the subject identification information specified by the information received from the user terminal by the subject designation receiving unit A write receiving unit for writing to the write storage unit, and write content data corresponding to the subject identification information are read from the write storage unit. A key text generation unit that generates a key text in which the written content data that has been read and read is set, and a similarity between the key text generated by the key text generation unit and the information text stored in the content storage unit A search unit that calculates a degree and selects information text similar to the key text based on the calculated similarity; and the content storage unit stores information on the content corresponding to the information text selected by the search unit A related content search device comprising: a distribution unit that reads and outputs the content from the content distribution unit.
According to this aspect, the related content search device searches for related content using the content written on the electronic bulletin board by a plurality of users for a specific subject and outputs the content to the user terminal.
As a result, users browsing content sites such as video sites and community sites provided on the network will be picked up on the bulletin board of the content provider site without requiring keyword input or other operations. It is possible to present content related to a topic. This related content was searched using written content such as comments of multiple users written on the electronic bulletin board about the video being distributed and the theme of the community, discussion between users, etc. , You can learn content that deepens discussions and gives users new perspectives.

[2] One aspect of the present invention is the related content search device described above, wherein the storage unit stores subject identification information, review content data, and comment content data for the review content data in association with each other. The key text generation unit includes a part of review content data corresponding to the subject identification information, a part of review content data corresponding to the subject identification information, and a comment content corresponding to the part of the review content data. A key text in which all the review content data and comment content data corresponding to the data or the subject identification information is set is generated.
According to this aspect, the related content search device can include all of the written contents written by a plurality of users on a specific subject, or representative opinions and impressions, or representative opinions and impressions, and responses and responses to the opinions. Search for related content using as the key text.
As a result, related content that matches the claims of all users who wrote about the video being distributed and the theme of the community, related content that matches the subjectivity of the user who wrote a certain opinion or impression, or a certain opinion or impression The related content in which different viewpoints are taken into consideration can be presented to the user.

[3] One aspect of the present invention is the related content search device described above, wherein the distribution unit includes the review content data or the review content data and the comment content data that are the generation source of the key text, The content information corresponding to the information text selected by the search unit is output in association with the information text.
According to this aspect, the correspondence between the range of the written content that has generated the key text and the content information obtained as a result of the search using the key text is transmitted to the user terminal.
Thereby, the user can grasp | ascertain the content relevant to which range of writing content.

[4] One aspect of the present invention is the related content search apparatus described above, wherein the key text generation unit sets the write content data designated by the user among the write content data corresponding to the identification information of the subject. The key text is generated.
According to this aspect, the related content is searched by using the written content designated by the user as the key text among the written content written on a certain subject such as the moving image being distributed or the theme of the community.
As a result, it is possible to search and present content related to the written content that the user is interested in.

[5] According to one aspect of the present invention, there is provided an information text including any one or more of an outline sentence, a subtitle, and a manuscript of the content that represents a content , A content storage unit that stores content information in association with each other, a write storage unit that stores identification information on the subject and written content data indicating an impression or opinion on the subject in association with each other, and identifies a subject from a plurality of user terminals A subject designation receiving unit for receiving information to be received, receiving the written content data from the user terminal, and receiving the written content data and the subject specified by the information received from the user terminal by the subject designation receiving unit. Corresponding to the identification information of the subject, the writing receiver that writes the identification information in association with the writing storage unit The written content data is read from the write storage unit, the key text generating unit for generating the key text in which the read writing content data is set, the key text generated by the key text generating unit, and the content stored in the content storage unit. A search unit that calculates a similarity to the information text and selects an information text similar to the key text based on the calculated similarity, and the content corresponding to the information text selected by the search unit The program is made to function as a distribution unit that reads and outputs the information from the content storage unit.

  ADVANTAGE OF THE INVENTION According to this invention, the content relevant to a theme can be searched and shown to a user using the content which the other user wrote in the electronic bulletin board about the specific theme which the user is browsing on a network. Therefore, it is possible to perform a search considering a deep purpose that cannot be expressed explicitly as a keyword by the user himself, or to perform a wide search that is not captured by a specific keyword, and matches a user's preference and is obtained by a keyword search. Unexpected content that has not been made can be presented to the user. In other words, it is possible to present to the user content that cannot be obtained only by a simple search using a keyword specified by the user, but that matches the user's purpose. Thereby, for example, it is possible to present content that deepens the discussion on the electronic bulletin board or content that has a viewpoint that the user did not expect.

It is a functional block diagram of the related content search apparatus by the 1st Embodiment of this invention. It is a figure showing an example of data which a contents storage part by the embodiment memorizes. It is a figure which shows the example of data which the write memory | storage part by the same embodiment memorize | stores. It is a figure which shows the screen image by the embodiment. It is a flowchart which shows the processing flow of the related content search device by the embodiment. It is a figure which shows the example of the key text by the same embodiment. It is a figure which shows the specific data flow when the processing flow by the embodiment is performed. It is a figure which shows the other screen image by the same embodiment. It is a functional block diagram of the related content search device by the 2nd Embodiment of this invention. It is a figure showing an example of data which a contents storage part by the embodiment memorizes. It is a figure which shows the example of data which the write memory | storage part by the same embodiment memorize | stores. It is a figure which shows the screen image by the embodiment. It is a figure which shows the processing flow of the related content search device by the embodiment.

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

[First embodiment]
FIG. 1 is a block diagram showing a configuration of a related content search apparatus 1 according to the first embodiment of the present invention. The related content search device 1 is connected to a plurality of user terminals 2 via a network 3. The network 3 is a public network such as the Internet or a private network such as an intranet. The related content search apparatus 1 can be realized by one or a plurality of server computers or the like, and includes a content storage unit 11, a write storage unit 12, a distribution request reception unit 13 (subject designation reception unit), a content search unit 14, A screen generation unit 17, a distribution unit 18, and a write reception unit 19 are provided.

  The content storage unit 11 and the write storage unit 12 are realized by a hard disk device, a semiconductor memory, or the like. The content storage unit 11 stores content data that is digital data of the content main body and information text that is text data describing a description of the content. Here, the content is, for example, a broadcast program, a moving image, a still image, sound, text, or the like. The writing storage unit 12 stores writing contents that are text data of sentences written by the user regarding the contents.

  The distribution request receiving unit 13 receives a content distribution request from the user terminal 2. The content search unit 14 includes a key text generation unit 15 and a search unit 16. The key text generation unit 15 reads out the written content of the content to be distributed from the write storage unit 12 and generates key text that is text data used for searching for related content. The search unit 16 calculates the similarity between the key text generated by the key text generation unit 15 and the information text stored in the content storage unit 11, and relates the content corresponding to the information text having a similarity higher than a predetermined value. The screen generation unit 17 is notified as content.

  The screen generation unit 17 reads the content data of the content to be distributed read from the content storage unit 11, the written content of the content read from the write storage unit 12, and the information related to the related content obtained as a result of the search by the search unit 16. Generate screen data to display. The distribution unit 18 distributes the screen data generated by the screen generation unit 17 to the user terminal 2 via the network 3. The write receiving unit 19 receives the write content from the user terminal 2 via the network 3 and writes it in the write storage unit 12.

  The user terminal 2 is a computer terminal such as a personal computer or a mobile phone. A browser or the like operating on the user terminal 2 displays the screen data received from the related content search device 1.

  Hereinafter, a case where the related content search apparatus 1 according to the present embodiment provides a VOD (video on demand) service for distributing a broadcast program requested by a user via a website will be described as an example. That is, the content distributed by the related content search device 1 is a broadcast program. The website provided by the related content search apparatus 1 includes a page of each broadcast program (hereinafter also simply referred to as “program”). By accessing each page from the user terminal 2, the user can view a program corresponding to the page, and can freely input reviews and comments on the program, such as comments and opinions. it can. The written reviews and comments can be freely shared between users who are viewing the same program on the same site. Hereinafter, each program page is referred to as a program distribution site. The program distributed on the program distribution site is the subject, and the contents of the review and comment are the contents written on the subject.

  FIG. 2 shows an example of data stored in the content storage unit 11. The content storage unit 11 stores content data, representative image data, and content additional data in association with the content ID. The content ID uniquely identifies each content. The representative image data is data of thumbnail images of content. The content additional data is text data of text describing information related to the content, and includes a program title, a URL (Universal Resource Locator) of a program distribution site, and an information text. The information text is, for example, text data indicating a summary sentence indicating a program outline, a program performer, a program subtitle, a program manuscript, and the like.

FIG. 3 shows a data configuration example of write data stored in the write storage unit 12. The write data is configured by associating a content ID, review data, and comment data.
The review data includes a review ID, a user ID of the user who wrote the review, and review content data (write content data). A review is an impression or opinion about a program. The review ID uniquely identifies a review within the same content. The user ID uniquely identifies the user. Review content data (hereinafter simply referred to as “review content”) is text data of a review written by a user regarding a comment.
The comment data includes a comment ID, a user ID of the user who wrote the comment, and comment content data (write content data). The comment is an opinion or opinion of another user regarding the review. The comment ID uniquely identifies a comment within the same review. Comment content data (hereinafter simply referred to as “comment content”) is text data of a comment written by the user with respect to the review content.

  FIG. 4 is a diagram showing a display image of the program distribution site display screen in the user terminal 2. As shown in the figure, a representative image of a program, a program title, and an outline of the program are displayed at the top of the program distribution site display screen. The outline of the program includes a synopsis of the program included in the information text, performer information, and the like. A program is reproduced by clicking the representative image. Below the representative image of the program, a review input field for writing the review content for the program and a review posting button for instructing to transmit the review content written in the review input field to the related content search device 1 are displayed. Is done.

  In the review display area under the review input field and the review posting button, the review content of the posted review, an image representing the posted user, and the like are displayed. In the comment display area below the review display area, the comment content of the comment posted for the review is displayed. The comment display areas are arranged in the order in which the comments are written, and a comment input field for writing the comment contents and a comment written in the comment input field to the related content search device 1 below the last comment display area. A comment posting button for instructing transmission is displayed.

In the display area of “recommended related programs” located on the right side of the program distribution site display screen, representative images of related programs searched based on written contents such as review contents and comment contents are displayed. Representative images of related programs are displayed in descending order of relevance with the written content.
In the figure, only one review display area is displayed, but a plurality of review display areas may be displayed. Similarly, a comment display area, a comment input field, and a comment posting button are displayed below each review display area.

FIG. 5 is a flowchart showing a processing flow of the related content search apparatus 1 according to the present embodiment.
The user inputs a user ID and an instruction to access the program distribution site to the user terminal 2. When the user terminal 2 receives the input of the user ID and the access instruction to the program distribution site from the user, the user terminal 2 transmits the user ID and the access request to the related content search device 1. In the access request as the content distribution request, the URL (information specifying the subject) of the program distribution site of the program that the user wants to watch is set. The distribution request receiving unit 13 of the related content search device 1 receives the user ID and the access request from the user terminal 2 (step S105: access request). Hereinafter, a program distributed on the program distribution site of the URL set in the access request is referred to as key content.

  The key text generation unit 15 of the content search unit 14 stores the review stored in the write storage unit 12 corresponding to the content ID of the key content (hereinafter referred to as “key content ID”) (subject identification information). From the data and comment data, review data and comment data in a range to be used as key texts for searching for related contents are specified. The key text generation unit 15 reads the review content from the specified review data, reads the comment content from the specified comment data, and generates the key text by combining the read review content and the comment content (step S110).

  The search unit 16 calculates the similarity between the key text generated by the key text generation unit 15 and the information texts of all other contents stored in the content storage unit 11 (step S115). For the calculation of the similarity, the similarity calculation method between documents described in Reference Documents 1 to 3 can be applied.

[Reference 1] Goto et al., “Searching for related programs starting from the program being watched”, Proc. Of the 8th Information Science and Technology Forum (FIT2009), no.3, D-006, 2009, p.127- 128
[Reference 2] Goto et al., “Searching for related programs based on Web content being browsed”, 2009 Video Information Media Society Winter Conference Proceedings, 2-1, 2009
[Reference 3] Goto et al., “Program Search Using Wikipedia's Update History”, Proceedings of the 2010 Annual Conference of the Institute of Image Information and Television Engineers, 3-4, 2010

In References 1 and 3, the score of each word n-gram (engram) appearing in the document is calculated by the Okapi BM 25, and the calculated score is weighted by the extended specific expression and the n-garam. Then, the similarity is calculated from the sum of the weighted scores of the word n-gram included in the two documents in common.
When applying the similarity calculation method of References 1 and 3, the content storage unit 11 stores in advance the weighted score of the word n-gram included in each information text. The search unit 16 extracts the word n-gram that co-occurs with the key text for each information text, and calculates the similarity by summing the weighted scores of the extracted word n-gram.

  Further, in Reference Document 2, the similarity between two texts is calculated from a word vector weighted by TF / IDF (word importance evaluation method) of each word. In the appearance frequency method (TF), when there is a document group composed of a plurality of (DN) documents, a TF (w) value indicating the number of occurrences of a word w (keyword) in a document, And the TF-IDF value of the word w in the document is TF (w) × log (DN / DF (w)) using the DF (w) value indicating the number of documents in which the word w is included. calculate. The word vector of each document is expressed by using the TF-IDF value of each word as an element.

When applying the similarity calculation method of Reference Document 2, the word vector of each information text is stored in the content storage unit 11 in advance. The search unit 16 calculates the TF-IDF value of each word included in the key text from the DF value of each word and the TF value of each word included in the key text, and calculates the calculated TF-IDF value of each word. Generate a word vector whose elements are The search unit 16 calculates a cosine between the word vector of the key text and the word vector of each information text read from the content storage unit 11 as the similarity. The larger the cosine, the closer the angle between the word vectors is, indicating that the documents are similar.
Note that the similarity calculation method between documents is not limited to the above method, and other similarity calculation methods may be used.

  The search unit 16 selects an information text whose degree of similarity with the key text satisfies a predetermined condition as the information text of the related content (step S120). The predetermined condition is, for example, that the similarity with the key text is within a predetermined order from the higher side in the information text, or that the similarity with the key text is greater than or equal to a predetermined value set. is there. The search unit 16 reads the URL of the program distribution site read from the content additional data including the information text of the related content and the representative image data corresponding to the content additional data of the related content from the content storage unit 11, and Are output to the screen generator 17 together with the similarity calculated for the information text (step S125).

  The screen generation unit 17 reads representative image data and content additional data associated with the key content ID from the content storage unit 11 (step S130). Subsequently, the screen generation unit 17 reads the review content and user ID in the review data corresponding to the key content ID and the comment content in the comment data corresponding to the review data from the writing storage unit 12 (step S135). Further, the screen generation unit 17 reads the user image stored in advance in the writing storage unit 12 in correspondence with the user ID read from the review data.

The screen generation unit 17 receives the representative image data, the program title and information text in the content additional data, the review display region for displaying the review content and the user image, the comment display region for displaying the comment content, and the search unit 16. Screen data of a program distribution site display screen including a display area of a recommended related program for displaying representative image data of related content is generated (step S140). The representative image data of the related content is arranged in descending order of similarity and is associated with the URL of the related content.
The distribution unit 18 transmits the program distribution site display screen data generated by the screen generation unit 17 to the user terminal 2 that is the access request transmission source (step S145).

  The user terminal 2 of the access request transmission source displays the program distribution site display screen data received from the related content search device 1 on the display. When the user clicks the representative image of the key content with a mouse or the like, the user terminal 2 transmits a content reproduction request to the related content search device 1. When the distribution request receiving unit 13 of the related content search device 1 receives the content reproduction request, the screen generation unit 17 reads content data corresponding to the key content ID from the content storage unit 11 and instructs the distribution unit 18 to distribute. The user terminal 2 reproduces the content data distributed from the related content search device 1.

  At any of the user terminals 2 that are currently accessing the key content program distribution site, the user writes the review content in the review input field of the program distribution site display screen and clicks the review posting button. When the user terminal 2 receives the comment content written in the review input field and the input of the click of the review posting button, the user terminal 2 sends the review write data in which the input review content and the user ID are set to the related content search device 1. Send. When receiving the review writing data (step S105: review), the writing receiving unit 19 of the related content search device 1 assigns a review ID. The write receiving unit 19 writes the review data in which the assigned review ID, the review content and the user ID in the received review write data are set in association with the key content ID in the write storage unit 12 (step S150).

  Alternatively, in any of the user terminals 2 that are currently accessing the key content program distribution site, the user writes the comment content in the comment input field of the program distribution site display screen and clicks the comment posting button. When the user terminal 2 receives the comment content written in the comment input field and the input of the comment post button click, the input comment content, the review ID of the review corresponding to the comment, and the user ID are set. The comment writing data is transmitted to the related content search device 1. When receiving the comment writing data (step S105: comment), the writing receiving unit 19 of the related content search device 1 assigns a comment ID. The writing receiving unit 19 writes and stores comment data in which the assigned comment ID, the comment content in the received comment writing data, and the user ID are set in association with the review data specified by the key content ID and the review ID. The data is written in the unit 12 (step S155).

  After the process of step S150 or step S155, the related content search device 1 performs the same process as that after step S110. However, in step S145, the distribution unit 18 distributes to all user terminals 2 that are currently accessing the site that distributes the key content. Therefore, in step S140, the screen generation unit 17 generates program distribution site display screen data for displaying the content data being distributed to the user terminal 2 as the distribution destination instead of the representative image.

  When the user clicks on the representative image of the related program displayed on the user terminal 2, the user terminal 2 transmits the user ID and an access request in which the URL associated with the representative image is set to the related content search device 1. To do. The related content search device 1 performs the processing from step S105 using the related content for which the representative image is clicked as the key content. Thereby, the user can access the page of the program distribution site of the selected related program and can view the related program.

FIG. 6 is a diagram showing an example of key text generation in step S110 of FIG.
Two review data are written in the write storage unit 12 corresponding to the content ID (key content ID) of the key content S1. Review contents in the two review data are review contents R1 and R2, respectively. Further, three comment data are associated with each of the review data including the review content R1 and the review data including the review content R2, and are written in the write storage unit 12. The comment contents of the comment data corresponding to the review data including the review contents R1 are comment contents C1-1, C1-2, and C1-3, respectively, and the comment contents of the comment data corresponding to the review data including the review contents R2 are respectively , Comment contents C2-1, C2-2, and C2-3.

  The range of the written content used as the key content is, for example, the review content, the review content and the comment content for the review content, all the review content and the comment content corresponding to the key content.

  When generating the key text for each review content, the key text generating unit 15 sets each of the review content R1 and the review content R2 as the key text. Of the review contents R1 and R2, only the review contents of the review data in which the user ID of the program distribution destination user is set may be used as the key text, and the review ID or user ID designated by the user is received from the user terminal 2 and received. Only the review content of the review data in which the review ID or the user ID is set may be used as the key text.

  For example, by using the key text KA including only the review content R1, it is possible to obtain related content that greatly reflects the opinion of the person who has written the review content R1 according to the subjectivity of the user. That is, the related content corresponding to the claim of the specific user with respect to the key content S1 is searched. For example, when the user who wrote the review content R1 writes “there is an aspect of X” to the key content S1, a related content search using text information of X is performed, and as a result Is presented. In this way, a “content group related to a certain user's assertion” can be presented.

  When generating the key text using the review content and the comment content for the review content, the key text generation unit 15 combines the review content R1 and the comment content C1-1, C1-2, C1-3, the review text, A key text is generated by combining the content R2 and the comment content C2-1, C2-2, C2-3. The review contents of the review data in which the user ID of the program distribution destination user is set and the comment contents for the review contents may be used as the key text, and the review ID or user ID designated by the user is received from the user terminal 2 and received. The review content of the review data in which the ID or user ID is set and the comment content for the review content may be used as key text.

  For example, a topic raised by the user who wrote the review content R1 by using the key text KB that combines the review content R1 and the comment content C1-1, C1-2, C1-3 for the review content R1. And a search for content related to all the claims felt by other users. The user who wrote the review content R1 writes “there is an aspect of X” on the key content S1, and other users write “there is an aspect of Y”, “Z” on the review content R1. In the case of an assertion such as “I can't overlook this aspect”, contents related to all aspects of X, Y, and Z are searched. Therefore, the related content search device 1 can present a “content group related to one topic related to the content S1”. Each user can know the content produced from a viewpoint different from the viewpoint that he claimed, and can enjoy the merit of expanding the range of thinking.

  When generating the key text using all the review contents and the comment contents, the key text generating unit 15 reviews the review contents R1, R2 and the comment contents C1-1, C1-2, C1-3, C2-1, C2-. 2 and key text K-C are generated by combining C2-3. By using the key text K-C generated from all writing related to the key content S1, it is possible to perform a search in consideration of the claims of all users who have written regarding the key content S1. That is, the related content search device 1 can present a “content group relating to a general claim for the content S1”.

  The key text generation unit 15 of the related content search device 1 receives the user ID designated by the user from the user terminal 2, and the review content of the review data in which the received user ID is set and the received user ID are set. The comment content of the comment data that has been made may be combined into key text. Further, the key text generation unit 15 of the related content search device 1 receives the review ID and the comment ID designated by the user from the user terminal 2, the review content of the review data in which the received review ID is set, and the received comment The comment contents of the comment data in which the ID is set may be combined into key text.

  The range used as the key content may be determined in common for all program distribution sites or for each program distribution site, for example, or may be designated by the user. By selecting a range of written contents to be used as key content or using a plurality of ranges, the related content search device 1 can realize a related content presentation function suitable for the purpose of the site operator. It becomes.

FIG. 7 is a diagram showing a specific data flow when the processing flow shown in FIG. 5 is executed.
The related content search device 1 distributes the content data of the key content S1 to the user terminal 2 accessing the program distribution site. The user terminal 2 displays the key content S1 on the program distribution site display screen, and each user views the key content S1 displayed on the user terminal 2. A user posts the review content R1 of the key content S1 being viewed on the program distribution site through the user terminal 2. When receiving the input of the review content R1 of the key content S1, the user terminal 2 transmits the input review content R1 to the related content search device 1. The write receiving unit 19 of the related content search device 1 writes the review data including the review content R1 in the write storage unit 12 in association with the content ID of the key content S1.

  Thereafter, other users write the comment contents C1-1 to C1-n to the review contents R1 using their own user terminals 2. When the user terminal 2 of another user receives the input of the comment contents C1-1 to C1-n for the review contents R1, the input comment contents C1-1 to C1-n are transmitted to the related content search device 1. The write receiving unit 19 of the related content search device 1 associates the comment data including the comment contents C1-1 to C1-n with the content ID of the key content S1 and the review ID of the review content R1 in the write storage unit 12. Write. The written reviews and comments are displayed on other user terminals 2 that are currently accessing the program distribution site of the key content S1.

  The key text generation unit 15 of the related content search device 1 generates the key text K1 from the review content R1 and the comment content C1-1 to C1-n. The key text K1 is a text data of the review content R1 and the comment content C1-1 to C1-n that are simply connected and combined, but only the review content R1 may be used as the key text K1.

  The content storage unit 11 stores content data, representative image data, and content additional data of the contents P1 to Pn. The search unit 16 calculates the similarity between the key text K1 and the information texts T1 to Tn of the contents P1 to Pn stored in the content storage unit 11. The search unit 16 generates program distribution site display screen data that displays a related program list that is content corresponding to the information text selected based on the similarity to the key text K 1, and distributes it to the user terminal 2. This is displayed as “recommended related program” on the content distribution screen shown in FIG.

When the user specifies a range to be used as key content, the user terminal 2 transmits the user ID and the key content range input by the user to the related content search device 1. The key text generation unit 15 of the related content search device 1 stores the key content range in association with the user ID.
In step S <b> 110, the key text generation unit 15 generates key text according to the key content range corresponding to the user ID of each user terminal 2. In steps S <b> 115 to S <b> 120, for each key text generated for each user ID, the search unit 16 calculates a similarity with each of the information texts of all the contents other than the key contents and selects related contents. In step S125, the search unit 16 outputs the user ID, the URL of the related content program distribution site selected based on the key text generated for the user ID, and the representative image data to the screen generation unit 17 together with the similarity. In step S140, the screen generation unit 17 generates program distribution site display screen data for each user ID. That is, the screen generation unit 17 generates program distribution site display screen data that displays representative image data of related content corresponding to the user ID in descending order of similarity. In step S145, the distribution unit 18 transmits the program distribution site display screen data generated corresponding to the user ID of the user terminal 2 to the user terminal 2 currently accessing the program distribution site of the key content.

  FIG. 8 is a diagram showing a display image of a program distribution site display screen displayed on the user terminal 2 when a plurality of ranges used as key contents are applied. As shown in the figure, a review display area and a comment display area are displayed below the review input field and the review posting button on the program distribution site display screen. Then, in correspondence with the range of the written content used as the key content, a representative image of the related content as a result of searching using the key content is displayed. For example, a representative image D1 of related content obtained as a result of a search using the review content R1 as key content is also displayed in the review display area that displays the review content R1. Also, a representative image D2 of related content obtained as a result of using the review content R1 and the comment content C1 to Cn as key content is represented in a frame including these written content. Furthermore, in the display area of “recommended related programs” located on the right side of the program distribution site display screen, representative images D3, D4, and related images of related contents obtained as a result of searching all the review contents and comment contents as key contents. D5 is displayed.

  As described above, when the key text generation unit 15 generates key texts having different writing content ranges, the screen generation unit 17 uses the writing content range used as the key text and the writing content as the key text. Screen data to be displayed in association with a representative image of related content obtained as a result of the search is generated.

  As described above, in this embodiment, it is possible to provide a user with a related program of a program by using user-written information related to the program.

  The related content presented in the present embodiment is not content searched by a specific user using a specific keyword, but content that matches the content of a discussion that is raised about a certain content. As a result, it is possible to support user activities such as activating discussions thereafter and deepening the discussions.

Information in content additional data such as information text may be displayed on the program distribution site display screen in addition to the representative image of the related content or instead of the representative image of the related content.
Further, the content storage unit 11 may not store content data. Further, the content storage unit 11 may store only the content additional data without storing the content data and the representative image data. In this case, the program distribution site display screen displays information in the content additional data such as information text instead of the representative image.
Further, the content searched by the related content search device 1 may be information provided on an electronic bulletin board related to a certain topic or a site of a community service such as SNS (social networking service). In this case, the content storage unit 11 of the related content search device 1 stores the content additional data in which the URL of the community site and the information text of the community site are set.

  In addition, as information text of this embodiment, a program summary sentence, a program performer, a program subtitle, a program manuscript (hereinafter referred to as “program summary sentence etc.”) described in advance is shown as a program text. Text data to which the written content data, which is the content of the review and the content of the comment written, is added may be used. Thereby, the related content search device 1 calculates the similarity with the key text for the information text including both the program summary sentence and the written content data. Therefore, the related content search device 1 is not only content related to a program summary sentence similar to the key text, but also content whose written content data (current topic) written by the user for the program is similar to the key text. Can also be presented as a search result.

  Further, only the written content data for the program may be used as the information text of the present embodiment. In this case, the information text does not include a program summary sentence or the like, but content additional data other than the information text may include a program summary sentence or the like. By using only written content data (so-called word-of-mouth) for a program as information text, the related content search device 1 can perform a search using implicit information that is not explicitly included in the program. It becomes. Therefore, the related content search device 1 can present the result of searching for content with an emphasis on unexpectedness to the user. For example, neither the program A nor the program B appear in the person C, and the information of the synopsis or performer does not include the person C, but the contents written in the program A and the program B are related to those programs. A person C may be taken up as a person. In such a case, the program B can be presented as a search result to the user viewing the program A.

  As described above, when using the information text to which the write content data is added or the information text in which only the write content data is set, all the write content data may be targeted and written within a predetermined time from the current time. The written content data may be the target. By targeting all written content data, the related content search device 1 can present content obtained as a result of performing a search in consideration of the overall topic and evaluation of the program to the user. Further, by targeting written content data within a predetermined time from the current time, the related content search device 1 presents to the user content obtained as a result of performing a search with a bias due to the latest topicality. Can do.

  When all the write content data is targeted, in step S150 of FIG. 5, the write receiving unit 19 adds the review content in the received review write data to the information text in the content additional data corresponding to the key content ID. To do. Further, in step S155 of FIG. 5, the write receiving unit 19 adds the comment content in the received comment write data to the information text in the content additional data corresponding to the key content ID.

When writing content data within a predetermined time from the current time is targeted, in step S150 in FIG. 5, the write receiving unit 19 associates the assigned review ID with the key content ID and the received review write data. The review data in which the review contents and user ID and the writing time indicating the current time are set are written in the writing storage unit 12. Further, in step S155 of FIG. 5, the write receiver 19 associates the key content ID with the assigned comment ID, the comment content and user ID in the received comment write data, and the write time indicating the current time. Is written in the write storage unit 12.
Then, after step S150 or step S155, the write receiving unit 19 reads the review contents by specifying the review data indicating the write time within the predetermined time from the current time among the review data corresponding to the key content ID. . Further, the writing receiving unit 19 reads out the comment content by specifying the comment data indicating the writing time within the predetermined time from the current time among the comment data corresponding to the key content ID. The write receiving unit 19 updates the review content and comment content currently described in the information text in the content additional data corresponding to the key content ID with the read review content and comment content. And the process after step S110 is performed.

[Second Embodiment]
In the first embodiment, the case where the related content search device 1 provides a content distribution service such as the VOD service has been described. However, in this embodiment, the related content search device is an electronic bulletin board, an SNS (social networking service), or the like. A case where the community service is provided will be described.

  The community service is a function that enables users who are interested in a specific theme to form a community and exchange information between users belonging to the community. A large-scale SNS site manages a huge number of communities and provides a community search function for the convenience of the user. Generally, a user searches for a target community by performing a search using a community title representing a community theme as a keyword by using such a community search function. However, there are many communities whose actual contents match the user's interest even though the community title is not likely to be related to the user's interest. It is difficult for a user to reach such a community only with a community search function using a conventional keyword. Therefore, the related content search apparatus of the present embodiment uses the comparison method using the similarity between texts shown in the first embodiment, and presents other related communities when browsing a certain community. To do. That is, in this embodiment, the theme of a community is a theme, and the information exchanged between users is the content written to the theme.

  FIG. 9 is a block diagram showing the configuration of the related content search device 5 according to the second embodiment of the present invention. The related content search device 5 can be realized by one or a plurality of servers or the like, and includes a content storage unit 51, a write storage unit 52, a distribution request reception unit 53 (subject designation reception unit), a content search unit 54, a screen A generation unit 57, a distribution unit 58, and a write reception unit 59 are provided.

  The content storage unit 51 and the write storage unit 52 are realized by a hard disk device, a semiconductor memory, or the like. The content storage unit 51 stores community information that describes what kind of community it is and information text that is text data that combines the contents written to the community. The write storage unit 52 stores the write contents written by the user for each community.

  The distribution request receiving unit 53 receives an access request to the community site from the user terminal 2. The content search unit 54 includes a key text generation unit 55 and a search unit 56. The key text generation unit 55 reads the content written for the community accessed by the user from the write storage unit 52 and generates the key text. The search unit 56 calculates the similarity between the key text generated by the key text generation unit 55 and the information text stored in the content storage unit 51, and associates the community corresponding to the information text with the similarity higher than a predetermined value. The screen generation unit 57 is notified as a community.

  The screen generation unit 57 generates screen data for displaying the community information read from the content storage unit 51, the written content read from the write storage unit 52, and related community information obtained as a result of the search by the search unit 56. The distribution unit 58 distributes the screen data generated by the screen generation unit 57 to the user terminal 2 via the network 3. The write receiving unit 59 receives the write content from the user terminal 2 via the network 3 and writes it in the write storage unit 52.

  FIG. 10 shows an example of data stored in the content storage unit 51. The content storage unit 11 stores community information, representative image data, and information text in association with the community ID. The community ID uniquely identifies each community. The community information indicates a community title and a community description. The information text is text data of contents written to the community.

  FIG. 11 shows a data configuration example of write data stored in the write storage unit 52. The write data indicates the association between the community ID and the comment data. The comment data includes the comment ID, the user ID of the user who wrote the comment, and the comment content (written content data).

  FIG. 12 is a diagram showing a display image of the community site display screen displayed on the user terminal 2 in the present embodiment. As shown in the figure, the community ID, community title, and community description are displayed from the top of the community site display screen. Below this description, a comment input field for writing the comment content and a post button for instructing to transmit the comment content written in the comment input field to the related content search device 5 are included.

  Below the comment input field and the comment posting button, a comment display area for comments posted to the community bulletin board is displayed. In the comment display area, the comment content of the posted comment is displayed. In addition, when a comment that the user likes is specified with a mouse or the like, a check (selection flag) “Like” is displayed in the comment area of the comment. In the display area of “related community” located on the right side of the community site display screen, a representative image of the related community searched based on the written content is displayed. The representative images of the related communities are displayed in the order of related programs that are highly related to the written content.

FIG. 13 is a diagram showing a processing flow of the related content search device 5 according to the present embodiment.
The user inputs a user ID and an access instruction to the community site to the user terminal 2. When the user terminal 2 receives the input of the user ID and the access instruction to the community site, the user terminal 2 transmits the user ID and the access request to the related content search device 5. In the access instruction, the URL (information specifying the subject) of the community site to which the user has instructed access is set. The distribution request receiving unit 53 of the related content search device 5 receives the user ID and the access request from the user terminal 2 (step S505: access request). Hereinafter, a community provided on the community site of the URL set in the access request is referred to as a key community.

  The key text generation unit 55 reads the comment data corresponding to the community ID of the key community (hereinafter referred to as “key community ID”) (the identification information of the subject) from the writing storage unit 52 and combines the read comment contents. Then, key text is generated (step S510). The search unit 56 uses the same similarity calculation method as in the first embodiment to compare the key text generated by the key text generation unit 55 with the information texts of other communities stored in the content storage unit 51. The degree is calculated (step S515).

  The search unit 56 selects an information text whose similarity with the key text is higher than a predetermined condition as the information text of the related community (step S520). The predetermined condition may be a predetermined number in descending order of similarity, a predetermined similarity or higher. The search unit 56 reads the URL of the community site in the community information corresponding to the information text of the related community and the representative image data from the content storage unit 51, and the screen generation unit 57 together with the similarity calculated for the information text of the related community. (Step S525).

  The screen generation unit 57 reads community information associated with the key community ID from the content storage unit 51 (step S530). Furthermore, the screen generation unit 57 reads comment data corresponding to the key community ID from the write storage unit 52 (step S535). The screen generation unit 57 displays the community title and description in the community information read in step S530, the comment content in the comment data read in step S535, and the representative image data of the related community received from the search unit 56 in step S525. Community site display screen data is generated (step S540). At this time, the representative image data of the related community is arranged in descending order of similarity and is associated with the URL of the related community. The distribution unit 58 transmits the community site display screen data generated by the screen generation unit 57 to the user terminal 2 that is the access request transmission source (step S545). The user terminal 2 of the access request transmission source displays the community site display screen data received from the related content search device 5 on the display.

  In any user terminal 2 currently accessing the community site of the key community, the user writes the comment content in the comment input field of the community site display screen, and clicks the posting button. When the user terminal 2 receives the input of the comment content and the click of the posting button, the user terminal 2 transmits the comment writing data in which the input comment content and the user ID are set to the related content search device 5. When receiving the comment writing data (Step S505: Comment), the writing receiving unit 59 of the related content search device 5 assigns a comment ID. The write receiving unit 59 writes the assigned comment ID, the comment content in the received comment write data, and the user ID in the write storage unit 52 in association with the key community ID (step S550). Further, the write receiving unit 59 reads the comment data corresponding to the key community ID from the write storage unit 52 and combines the read comment contents to generate an information text. The write receiving unit 59 writes the information text generated in association with the key community ID in the content storage unit 51 (step S555).

  Thereafter, the related content search device 5 performs the processing after step S510. However, in step S545, the distribution unit 58 distributes to all user terminals 2 currently accessing the community site of the key community.

  When the user clicks the representative image of the related community displayed on the user terminal 2 using a mouse or the like, the user terminal 2 transmits a user ID and an access request to the related community to the related content search device 5. The related content search device 5 performs the processing from step S505 using the related community as a key community. Thereby, the user can access the community site of the selected related community.

  In addition to the representative image of the related community or instead of the representative image of the related community, the community title and description in the community information may be displayed on the community site display screen.

  According to the present embodiment, it is possible to present the presence of another community related to the community in which the user is currently participating (the subject that the user is interested in) without performing a keyword search or the like.

  In addition, you may make it search a related community using only the comment content which a user likes as a key text. In this case, the related content search device 5 generates a community site display screen that displays a button together with the comment content in the comment display area of the community site display screen. Then, the user accessing the community site browses the comment content of other users on the display screen of the community site displayed on the user terminal 2, and clicks the button corresponding to the comment content that he / she likes with a mouse or the like. , Check “Like”. The user terminal 2 transmits the comment ID of the checked comment content to the related content search device 5. The key text generation unit 55 of the related content search device 5 reads only the comment data specified by the received comment ID from the comment data corresponding to the key content ID from the write storage unit 52 and generates the key text. The related content search device 5 performs the processing after step S515. Thereby, it is also possible to present related content in consideration of the user's interests and preferences.

[effect]
The related content search device of the present embodiment described above is a content service written for a certain subject, for example, a user's opinion or conversation written on an electronic bulletin board about a distributed program, a community service such as SNS The related content is searched using the user's opinions and conversations regarding the thread written in the above. Therefore, the user can obtain content related to the current browsing content or a subject such as a thread without inputting an explicit search key such as a keyword. That is, by using text data written by other users who have the same interest, related content can be searched and presented even when the user has no specific search intention. In this way, by not requiring the user to perform an operation of explicitly assigning a keyword, it is possible to present content corresponding to a deeply desired request that the user cannot express as a keyword. Therefore, it is possible to provide the user with an unexpected encounter with new content while meeting the user's preference. It is also possible to present content that makes discussions on the electronic bulletin board deeper. As search key texts, electronic bulletin boards on the network, written contents of electronic bulletin boards in content distribution sites (review sites), and short posted texts such as Twitter can be used.

[Other]
In the above embodiment, as the information for specifying the subject, the URL of the site that provides the subject content such as the URL of the program distribution site or the URL of the community site, or the electronic bulletin board about the subject is used. However, the present invention is not limited to this. The information for specifying the subject may be information that can specify the content ID or community ID, and the content ID or community ID itself may be used.
Further, the related content search device 1 and the related content search device 5 may not receive the information for specifying the subject and the written content data at the same time.

  The related content search device 1 and the related content search device 5 described above have a computer system therein. Then, the distribution request receiving unit 13, the content searching unit 14, the screen generation unit 17, the distribution unit 18 and the write receiving unit 19 of the related content search device 1, and the distribution request receiving unit 53 of the related content search device 5, the content search unit 54, the operation process of the screen generation unit 57, the distribution unit 58, and the writing reception unit 59 is stored in a computer-readable recording medium in the form of a program, and the computer system reads and executes this program. The above processing is performed. The computer system here includes a CPU, various memories, an OS, and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a storage unit such as a hard disk built in the computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

1, 5 ... related content search device 11, 51 ... content storage unit 12, 52 ... write storage unit 13, 53 ... distribution request reception unit (subject specification reception unit)
14, 54 ... Content search unit 15, 55 ... Key text generation unit 16, 56 ... Search unit 17, 57 ... Screen generation unit 18, 58 ... Distribution unit 19, 59 ... Write reception unit 2 ... User terminal 3 ... Network

Claims (5)

A content storage unit that stores the information text including one or more of the content summary sentence, the subtitles, and the manuscript, which represents the content, and the content information;
A writing storage unit that stores the identification information of the subject and the writing content data indicating the impression or opinion on the subject in association with each other;
A subject designation receiving unit for receiving information identifying a subject from a plurality of user terminals;
Write content data is received from the user terminal, and the received write content data is associated with identification information of the subject specified by the information received by the subject designation receiving unit from the user terminal, and the write storage. Write receiver to write to
A key text generation unit that reads the write content data corresponding to the identification information of the subject from the write storage unit, and generates a key text in which the read write content data is set;
The similarity between the key text generated by the key text generation unit and the information text stored in the content storage unit is calculated, and an information text similar to the key text is selected based on the calculated similarity A search section to
A distribution unit that reads out and outputs information on the content corresponding to the information text selected by the search unit from the content storage unit;
A related content search apparatus comprising: The storage unit stores subject identification information, review content data, and comment content data for the review content data in association with each other,
The key text generation unit includes a part of review content data corresponding to the subject identification information, a part of review content data corresponding to the subject identification information, and a comment content data corresponding to the part of the review content data. Or generate a key text in which all the review content data and comment content data corresponding to the subject identification information are set,
The related content search apparatus according to claim 1, wherein: The distribution unit associates the review content data or the review content data and the comment content data that are the generation source of the key text with content information corresponding to the information text selected by the search unit. Output,
The related content search device according to claim 2, wherein: The key text generation unit generates key text in which the write content data designated by the user among the write content data corresponding to the subject identification information is set.
The related content search apparatus according to claim 1, wherein: A computer used as a related content search device,
A content storage unit that stores the information text including one or more of the content summary sentence, the subtitles, and the manuscript, which represents the content, and the content information;
A writing storage unit for storing the identification information of the subject and the writing content data indicating the impression or opinion on the subject in association with each other;
A subject designation receiving unit for receiving information identifying a subject from a plurality of user terminals;
Write content data is received from the user terminal, and the received write content data is associated with identification information of the subject specified by the information received by the subject designation receiving unit from the user terminal, and the write storage. Write receiver to write to the part,
A key text generation unit that reads out the write content data corresponding to the identification information of the subject from the write storage unit and generates a key text in which the read write content data is set;
The similarity between the key text generated by the key text generation unit and the information text stored in the content storage unit is calculated, and an information text similar to the key text is selected based on the calculated similarity Search part,
A distribution unit that reads out and outputs information on the content corresponding to the information text selected by the search unit from the content storage unit;
A program characterized by functioning as

Images