JP2013141050A - Content recommendation server, content display terminal, and content recommendation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evaluate scene information analyzed from viewable contents successively by user preference, thereby successively recommending scenes one scene at a time when another content is higher in preference than a currently viewed content.SOLUTION: User preference information on a display terminal indicating the preferences of the user is acquired by a server, while a currently viewed content and another content are received synchronously with the time of day at which they are displayed on the display terminal or separately in advance, and these contents are evaluated for their correlations with the user preference information by a predetermined method of analysis to find the respective degrees of compatibility with preferences. When some other content become higher in the degree of compatibility with preferences than the content currently viewed on the display terminal, recommended content information identifying the other content as a recommended content is sent to the display terminal.

Description

  The present invention relates to a content recommendation server, a content display terminal, and a content recommendation system, and more particularly to a technique for recommending another program from a server to a display terminal that displays content using a network.

  In recent years, with the advancement of compression technology for content such as video and audio, and the digitization of television broadcasting using the content, the use efficiency of radio waves and networks has been improved, and the number of channels has been increased along with higher image quality.

  In addition, the performance of display terminals such as personal computers and portable terminals has been improved, and network lines and storage devices have been increased in speed and capacity. By connecting to a network using these technologies, in addition to receiving television broadcasts by radio waves, a television that can receive video content using a network and receive services connected to a server, a so-called network TV has appeared.

  Furthermore, it will also be applied to so-called digital signage, which is a display device for providing advertisements and information provided in public facilities and companies by using this net TV, and the contents such as a huge number of images and sounds by users. Opportunities to be exposed to increased.

  In this way, TVs, particularly Internet TVs, receive more content and services by connecting to a server, and users can access enormous amounts of content. On the other hand, they can search for desired content and services. It has become difficult.

  Therefore, in the prior art, an electronic program guide (EPG) explaining the contents of a TV broadcast program (content) is distributed, and automatic searching from the program guide using a keyword desired by a user has become widespread. .

  For example, in Patent Document 1, a keyword that a user wants to search is registered in advance, the server searches for program information from the program information, and the result is transmitted to a recording device, thereby automatically recording a program desired by the user. It is disclosed.

  Further, in Patent Document 2, the degree of matching between a user's program viewing history and a program guide is evaluated, and the appearance order of channels is changed from the currently broadcasted program in the order that matches the user's preference. It is disclosed that it is easy to select a program that seems to match.

  In Patent Document 3, a plurality of programs that similarly match user preferences are searched from the program table so that the broadcast times do not overlap, and are arranged in one time series (one channel of the program table) across the broadcasting station. It is disclosed that a program that suits the user's taste can be easily accessed.

  In Patent Document 4, by analyzing the reproduction operation information so far on a video that is subject to the user's preference, it is possible to create a digest video without requiring an explicit digest generation method and parameter input from the user. A digest generation method is disclosed that automatically adjusts the method.

Japanese Patent No. 4320814 Japanese Patent No. 3844901 JP-A-2005-295001 JP 2009-4994 A

  According to the technique described in Patent Document 1 described above, automatic recording based on a search result of a program including a desired keyword from a program guide on a server using a keyword is possible, but compared with a program currently being viewed. It is not a technology to do. In addition, since it is a search by program unit using the program guide, it is difficult to evaluate in the middle of the program, and when a scene that matches the user's preference appears in the middle, it is not described in the program guide. An oversight occurs.

  Further, according to the technique described in Patent Document 2, it is a technique used for so-called zapping that searches for other programs that the user is likely to be interested in by changing the channel from the viewing program. There is no disclosure of automation. Further, since the search is performed in units of programs using the program guide, it is difficult to perform evaluation during the program as in Patent Document 1.

  Further, according to the technique described in Patent Document 3, the purpose is to create a broadcast program viewing plan using a program guide, but the program viewing plan for a time zone in which there is no program that matches the preference is disclosed. It is difficult to use this vacant time zone. Furthermore, similar to Patent Document 1 and Patent Document 2, since it is a technology that uses character information in the program guide, it is possible to detect episodes and characters that occur in a relatively short time such as “corners” and “scenes” in the program. There is no resolution, and overlooks again.

  Further, even the technique described in Patent Document 4 does not disclose recommendation of a more suitable content to the user in real time or prevention of oversight.

  An object of the present invention is to provide a content display terminal, a content recommendation server, and a content recommendation system that can prevent the above-described oversight and can recommend a more suitable content from a plurality of contents to a user in real time.

  A typical example of the invention disclosed in the present application is as follows. That is, the server of the present invention is a server configured to be connectable to a display terminal and a content providing source, and the server acquires user preference information indicating a preference of a user who uses the display terminal, and the display The first content currently being viewed on the terminal and at least one or more other content other than the first content provided from the content providing source are respectively correlated with the user preference information in parallel. The recommended content information that evaluates and obtains the preferred degree of preference and recommends the other content having the higher degree of preferred preference when the other content having the higher preferred degree of preference than the first content being viewed exists. It produces | generates and transmits to the said display terminal, It is characterized by the above-mentioned.

  According to the present invention, the user's preference and multiple contents are evaluated in real time on the server, and more suitable contents are recommended in units of scenes of contents that change from moment to moment. It is possible to contribute to the improvement of the display terminal value by increasing the active time of the display terminal by guiding to another content.

It is a block diagram which shows the structure of the content recommendation system which becomes the 1st Embodiment of this invention. It is a figure which shows an example of the recommendation server function in the content recommendation system of 1st Embodiment. It is explanatory drawing of the operation information in 1st Embodiment. It is a conceptual diagram which extracts subtitle person name information as feature information from the subtitle data in video data. It is explanatory drawing of keyword importance information. It is a conceptual diagram of the scene importance level information creation method in a person name search. It is a processing sequence figure of generation / update of user taste information in a 1st embodiment. It is a figure explaining the data which transmit the content information in view in 1st Embodiment. It is a processing sequence figure which compares and evaluates the viewing-in-contents and another content with user preference of 1st Embodiment, and recommends it to a display terminal. It is a figure explaining the data which transmit recommended content information of 1st Embodiment. It is explanatory drawing which is shown to a user when a content recommendation is made on the display terminal according to the first embodiment and is selected for viewing. It is explanatory drawing which is shown to a user when a content recommendation is made on the display terminal according to the first embodiment and is selected from a plurality. 6 is an explanatory diagram for selecting an ongoing recording stop when recording recommended content in the first embodiment; FIG. FIG. 10 is a processing sequence diagram for performing service recommendation when there is no recommended content according to the second embodiment of this invention. FIG. 10 is a processing sequence diagram for performing service recommendation when there is no recommended content according to the second embodiment of this invention. It is explanatory drawing of the setting screen of this threshold value and service setting information in the display part of the display terminal in 2nd Embodiment. It is explanatory drawing of the detailed setting of a service setting in the display part of the display terminal in 2nd Embodiment. It is a figure explaining the data transmitted as recommendation service information in a 2nd embodiment. It is explanatory drawing which is shown to a user when a service is recommended on the display terminal of 2nd Embodiment, and is made to select viewing. It is a process sequence diagram which compares with the user preference and recommends to the display terminal the content being viewed and another content according to the third embodiment of the present invention.

According to a typical embodiment of the present invention, in a content recommendation system comprising a display terminal that displays content and a server that recommends the content displayed on the display terminal to a user, the server is a user of the display terminal. Obtaining the user preference information indicating the user's preference, sending the viewing content information identifying the content currently being viewed from the display terminal to the server, and the server identifying the content identified from the viewing content information. Synchronized with the display time on the display terminal or separately received in advance, and one or more other contents other than the content being viewed on the display terminal can be displayed in synchronization with the displayable time on the display terminal or can be displayed While receiving sequentially from the time in advance, these contents are determined in advance. A method of evaluating a preference preference degree by evaluating a correlation with the user preference information by a method, and when the preference preference degree of other content is higher than the content being viewed on the display terminal Sending recommended content information to be specified as recommended content to the display terminal, and changing the display on the display terminal to the recommended content.
Hereinafter, embodiments according to the present invention will be described in detail with reference to the drawings.

  FIG. 1A is a block diagram showing a configuration of a content recommendation system according to the first embodiment of this invention. The content recommendation system in the form of the first actual example includes a recommendation server 100, at least one display terminal 120 for users, and at least one external content server 160. The recommendation server 100 and the display terminal 120 each provide a plurality of contents through the network 180 or using the transmission / reception antennas 161, 103-1, 103-2, 103-3, 104-1, 104-2, 104-3. It is configured to be connected to the source at the same time.

  The recommendation server 100 includes a user attribute information database 101 and an authentication unit 102, a reception unit 104, a user preference information database 105 and a preference evaluation unit 106, a calculation unit 107, a communication unit 108, a scene analysis unit 109, a service setting unit 111, and a user service. It consists of a setting information database 110. The calculation unit 107 includes a processor, a memory, and a storage device (such as a magnetic disk), and executes a program for controlling the recommendation server. The recommendation server may be configured by extending a general computer. The communication unit 108 performs communication by transmitting and receiving data using a network. This data may include content using a network. The receiving unit 104 receives the content distributed by radio waves by the receiving antenna 103, demodulates it, and enables it to be taken in as content data. In the example of FIG. 1A, three antennas (103-1, -2, -3) and three receivers (104-1, -2, -3) are shown. In this embodiment, at least two antennas are used. Need one.

  FIG. 1B shows an example of main functions of the recommendation server in the content recommendation system of the first embodiment. The recommendation server 100 has a parallel content acquisition function, a user preference information acquisition function, a content-specific scene analysis function, an inter-content preference appropriateness comparison function, a recommended content information generation function, and an idle time. A recommended content providing function and a display change function for the recommended content are provided. Note that the sharing of functions between the recommendation server 100 and the display terminal 120 in the content recommendation system may be changed as appropriate according to the application. For example, when the display terminal 120 is assumed to be a television and shared with a family, the user attribute information database 101, the authentication unit 102, the user preference information database 105, the preference evaluation unit 106, etc. A part or all of the user preference information acquisition function or the like may be transferred to the display terminal 120 so that a plurality of users in the home each have a user ID and a password and can receive a content recommendation service individually.

  Next, a more specific offense and action of the content recommendation system will be described. First, as described above, the recommendation server 100 receives or receives a plurality of contents in parallel via a reception antenna or a network (hereinafter, reception is performed unless reception and distribution are particularly distinguished), and processes them. It has a function to acquire multiple content information in parallel.

  Based on the operation information of the user who uses the display terminal 120, the authentication unit 102, which is part of the user preference information acquisition function, uses the data of the user attribute information database 101 to determine whether the user can use the system. While authenticating, it determines which user's preference information database 105 should be used.

  A scene analysis unit 109 that realizes a content-specific scene analysis function analyzes content data received using radio waves or a network, and acquires scene information. Here, the scene is a segment of the operation within the content. In other words, in the video, audio, text information, etc. included in the content, logical changes such as physical changes such as sudden changes in the screen and sound due to camera switching, etc., or the appearance and disappearance of people and the insertion of music ) Is defined as a part of the content indicated by the time between each change time. These are often used as an expression method for content creators, and are considered to be units in which user interests change. Scene information refers to physical quantities such as video and audio colors and levels in these scenes, character strings that make up character information, or character strings, meanings, concepts, and labels that describe them. Suppose that The scene information is sequentially analyzed by the scene analysis unit 109 in units of scenes as content is input every moment, and the scene information can be output.

  A preference evaluation unit 106 that bears a part of the user preference information acquisition function acquires user preference information of the user from a user preference information database 105 obtained in advance, and sequentially outputs scenes in units of scenes from the above-described scene information. By sequentially obtaining the correlation with the information, the user's preference for the scene, that is, the degree of “like” or “dislike” is output as the “preference level”. In addition, the preference evaluation unit 106 uses the operations while the user is viewing the content, that is, the content selection / change, the presence / absence of recording, the selection / fast forward / rewind, the pause, etc. for the recorded content. It is assumed that the user's preference level with respect to the content viewing scene can be evaluated in units of scenes, and recorded as user preference information in the user preference information database 105 and updated.

  The service setting unit 111 recommends switching and viewing with the idle recommended content providing function when there is no content with high preference among the currently viewable content specified in advance by the user through the terminal. The service setting information database 110 for each user, which is an option, is managed and presented to the user as recommended content as necessary.

  The display terminal 120 includes a calculation unit 125, a communication unit 121, a reception unit 124, a decoding unit 128, a display unit 126, an input unit 127, and a recording unit 122. The display terminal is, for example, a mobile terminal such as a television, a radio, or a mobile phone, a personal computer, a digital signage terminal, or the like. The calculation unit 125 includes a processor, a memory, and a storage device (such as a magnetic disk), and executes a program for controlling the display terminal. The communication unit 121 performs communication by transmitting and receiving data using a network. This data may include content using a network. The receiving unit 124 receives the content distributed by the radio wave by the antenna 130, demodulates it, and enables content data to be captured. FIG. 1A shows that there are two (130-1, -2) antennas and two (124-1, -2) receiving units, and this embodiment requires at least two antennas. The decryption unit 128 decrypts the content acquired from the communication unit 121 or the reception unit 124. The content thus received is presented to the user 150 from the display unit 126 together with terminal settings and in-terminal applications. In addition, the user operation is acquired from the input unit 127 such as a mouse, a keyboard, or a remote controller as necessary, and if necessary, from the communication unit 121 to the recommendation server 100 or the external content server via the network 180. Send it.

  The external content server 160 distributes content and content information to the display terminal 120 and the recommendation server 100 using a network or radio waves. Content that is distributed together with content information in units of content such as an electronic program guide (EPG) to terminals and recommendation servers, or content that is held in an external content server based on content information from terminals and recommendation servers May be selected and distributed, and the internal configuration is not limited.

  Here, the content is a so-called TV program including video and audio in TV broadcasting, text information such as subtitles, video, music, text, etc. distributed in the form of streaming or file distribution over the network. Suppose that it includes services such as media content, game content, email and web content using web technology, or an application that combines these. The content information includes a channel (frequency) necessary for reception and an identifier such as an identifier URI (Uniform Resource Identifier) indicating a resource (resource) in a certain format, title, distribution time zone, format, program outline, It is assumed that the information describes the content, such as performers and synopses. Regardless of the method, such as digitized compression format or analog modulation, the content and content information are distributed from an external content server using a predetermined format and received on the recommendation server or terminal side configured as a system It suffices if the decoding / demodulation and the like can be performed in the unit.

  When the external content server is a broadcasting station using general radio waves, the content distribution using the transmission antenna 161 is performed, and the server receives the radio waves using the radio wave reception antennas 103-1, 103-2, 103-3, and receives the reception unit 104. -1, 104-2, 104-3 to receive the content. Similarly, the terminal receives radio waves using the radio wave reception antennas 130-1 and 130-2, and can receive content using the reception units 124-1 and 124-2.

  When the external content server is a distribution server using a network, it is distributed through the network 180 and received using the communication unit 108 of the recommendation server 100 or the communication unit of the display terminal 120. The scene analysis unit 109 is used for the recommendation server, and the decoding is performed for the terminal. Decoded by the unit 128. A decoding unit may be provided separately from the scene analysis unit of the recommendation server.

Next, user operation information and user preference information related to the user preference information acquisition function of the recommendation server 100 will be described.
FIG. 2A is an explanatory diagram of user operation information using the display terminal 120. These user operations are input from the input unit 127 such as a remote controller or a touch panel to the calculation unit 125, and the display state is changed. This content is output to be given to the recommendation server 100 as operation information. Here, it is assumed that operation information can be managed for each user, and each operation can be accumulated as a history. Each operation information includes, for example, the date and time when the operation was performed, actions indicating the type of operation, parameters of those actions, and the temporal position of the video displayed when the operation was performed. The user 150 can output a history of viewing while performing a series of operations on the video of each ID.

  Next, a specific example of the user preference information acquisition / generation method will be described with reference to FIGS. 2B to 2D.

  FIG. 2B is a conceptual diagram of extracting caption person name information from caption data of one scene in video data. If there is subtitle data in the video data, for example, subtitle data such as "Sato keeps the ball and feeds to Tanaka" is displayed in the section "0: 00: 05.255 to 0: 00: 08.822" The extraction unit decomposes the character string into words by general morphological analysis. Thereby, it is separated into verbs, particles, nouns, etc., and can be classified into personal names. Based on this, the appearing section and the extracted person name are registered in the caption person name information as one feature of the video data. By accumulating these steps for all the video data, it is possible to use subtitle personal name information for digest creation. As described above, caption personal name feature information obtained by extracting a personal name from caption information in video data is obtained for each video.

  FIG. 2C is an explanatory diagram of the keyword importance information 320 to 321 managed in the user preference information database 105 and evaluated by the preference evaluation unit (106). The keyword importance information is held for each genre for each user, and holds keywords that the user seems to like in each genre. In this case, a genre-independent keyword group may be recorded as common keyword information 3201 or genre-specific keyword information 3202 and 33203. Also, the user may change the contents independently, or add or delete them.

  FIG. 2D is a conceptual diagram of a method for creating scene importance information in person name search. The parameters in the digest creation method according to the present example are the number numKwd from the top of the keyword importance information used for the search, and the start time offset ts and the end time offset te for extending the time segment to be reproduced by the search. . For example, when searching for and finding the keyword `` Sato '' with importance q1 (3410) (3440), (3441), first, (3440) is offset to the start time (3412) and end time (3413) of the personal name feature, respectively s1 (3431) and e1 (3432) are obtained so that ts (3411) and te (3414) are widened. If there is no overlapping existing scene importance in this part, the scene importance is registered with (s1, e1, q1) as elements (3405). Next, regarding the `` Sato '' part of (3441), the same items are prepared as elements (s2, e2, q1), but `` Suzuki '' (3442) of importance q2 (3415) is already as shown in the figure If the start part is registered so as to overlap the new part, the start time of the existing scene importance item (3407) is changed to e2 (3434) to leave the high importance `` Sato ''. Register a new element (s2, e2, q1). In this way, the scene importance level is registered so that a part having a high importance level remains. Finally, normalization is performed so that the importance level becomes 1 at the maximum, and the scene importance level information is updated, and the process ends. In addition, the description of patent document 4 is used for more detail.

  FIG. 3A is a processing sequence diagram for generating / updating user preference information according to the first embodiment of the present invention. Here, based on the viewing behavior of the user 150, a process of creating the user preference information database 105 as preference information of each user is performed. First, the user starts a login operation to the content recommendation system using the display terminal 120 (S201). Through this operation, the display terminal performs user authentication with the recommendation server 100 (S211 and S221). If the recommendation server acquires the data of the user attribute information database 101 and the authentication unit 102 can authenticate the user operation, for example, the user ID and password against the user attribute information, the login is completed. In the following description, the communication path between the display terminal and the recommendation server is assumed to be established. However, for each data exchange by communication, individual communication using a mutual address and user identifier may be specified. After logging in, the server preference evaluation unit 106 acquires the user preference information of the user from the user preference information database 105 (S222). When the user performs an action of viewing content (S202), the display terminal 120 displays the content from the external content server, acquires content information, and records it in the recording unit 122 (S212).

  Next, the content information being viewed is sent from the display terminal to the recommendation server (S213, S223). FIG. 3B is an example of content information to be sent. Here, data for transmitting an event name (301) for indicating the currently viewed content and a URI (302) for identifying the content is configured. The URI in FIG. 3B indicates that “terrestrial D”, that is, terrestrial digital broadcasting, is viewing the channel “091” in the “Tokyo 1” area. This URI may use a format that is unique, such as the above-mentioned notation of digital terrestrial broadcasting, but can be identified in a unified manner, and does not preclude the use of a format that is publicly available other than the URI. For example, when using a URL (Uniform Resource Locator) notation (notation starting from http: // or ftp: ///, file: ///) indicating a web page on the Internet or video content on the web, The external server (160) indicates the content (and resources) on the server uniquely defined from this URL notation, and may receive these distributions.

  The recommendation server 100 records that the program has been viewed (S224), the content information (S225, S231) from the external content server, and the same content that the user is viewing on the display terminal 120, Synchronized with the display on the display terminal or received in advance (S226, S232), the scene analysis unit 109 analyzes and obtains scene information (S227). When the content is changed while the user is viewing, the content information whose viewing is changed is transmitted again (S214). Here, for example, based on character string information using a character recognition technique (OCR) which is a generally known technique for character string information by subtitles, which is often included in TV programs, and telops displayed graphically on the screen, It is assumed that a “keyword” obtained by extracting a proper noun such as a person name or a place name in the character string or a general noun using a generally known technique is used as scene information.

  Next, the preference evaluation unit (106) evaluates the keyword as the scene information (S228). This evaluation method uses the history operated by the user and weights keywords that appear before and after the time on the content where the operation is performed from the state transition table of the previous operation and the current operation. In addition, the preference level which is an evaluation value as a keyword group and their statistical information changes according to the viewing history. For example, when the user loses interest and changes the content according to the content viewed by the user, the scene including the keyword immediately before the change is less interested (the preference is lowered). In the present embodiment, since the content is selected or changed is acquired in S213 and S223, the user's preference can be obtained as a result.

  Note that, as the viewing history disclosed in Patent Document 4, use of the history of operations such as content selection, pause, fast forward, and fast reverse in recorded content is not prevented. In this case, the operation history when viewing the recorded data database (123) of the display terminal (120) is sent to the recommendation server (100) in each case or afterwards. In the recommendation server, the same content is acquired from the external content server (160), the above-described content distribution, scene analysis, and scene evaluation are reproduced in the same manner as the history at the terminal, and finally the user preference information is similarly obtained. You may do it. Further, the content in this case may be recorded in advance so as to be recorded in the recommendation server. In this case, the recording data in the recommendation server is acquired in the same manner as the content distribution.

  In this way, the user preference information database (105) configured by sequentially evaluating the distribution contents is updated (S230), and the processing sequence of the user preference information configuration is ended.

  Next, processing for receiving content recommendation in the present embodiment during actual viewing will be described. FIG. 4 is a processing sequence diagram in which the content being viewed and another content are compared with the user preference and recommended to the display terminal in the first embodiment. This processing sequence includes a parallel acquisition function of multiple content information, a user preference information acquisition function, a content-specific scene analysis function, a content-specific scene analysis function, an inter-content preference suitable degree comparison function, a recommended content shown in FIG. 1B by the recommendation server. This is realized by executing each function of the information generation function and the display change function to the recommended content.

  Similar to the process of generating / updating user preference information in FIG. 3A, first, the user uses the display terminal 120 to start a login operation to the content recommendation system (S201). Through this operation, the display terminal performs user authentication with the recommendation server 100 (S211 and S221). If the recommendation server acquires the data of the user attribute information database 101 and the authentication unit 102 can authenticate the user operation, for example, the user ID and password against the user attribute information, the login is completed. In the following description, the communication path between the display terminal and the recommendation server is assumed to be established. However, for each data exchange by communication, individual communication using a mutual address and user identifier may be specified. After logging in, the preference evaluation unit of the recommendation server acquires the user preference information of the user from the user preference information database 105 (S222). Note that explanation of logout is omitted.

  Next, when the user 150 requests the display terminal 120 to start recommendation (S401), the terminal transmits a recommendation request to the recommendation server (S410). When the recommendation server receives it, processing for content that can be received by the preference evaluation unit (106) for the user, the reception unit (104), and the scene analysis unit (109) in the server is started (S430). . When the user starts viewing the content using the terminal, the display terminal receives the content information (S451, S411) and content distribution (S453, S411) from the external content server (160), and displays the content. The content information currently being viewed is transmitted to the recommendation server 100 (S412 and S431). The content information (S432, S451) and the same content that the user is viewing on the display terminal are received from the external content server 160 in synchronization with the display on the display terminal or distributed in advance ( (S433, S452), the scene analysis unit 109 performs analysis to obtain scene information (S434). Using this scene information and the previously acquired user preference information, the preference level for the content being viewed by the user is evaluated (S435, S436). With regard to this evaluation, in the present embodiment, in accordance with the above-described user preference information configuration processing, for example, a keyword that is a character string generated from subtitles or telops sent in accordance with content distribution is used. That is, the keywords appearing in the content being viewed are sequentially acquired, and if it exists in the user preference information, or if it exists, the preference degree of the keyword is obtained. For example, the sum of the keyword preference degrees is calculated in the scene. Evaluated as user preference. Note that the preference level does not exist only for the duration of the scene, but may be influenced by being added to the preference level of the subsequent scene while being reduced by a time constant.

  On the other hand, the recommendation server acquires content information different from one or more contents being viewed by the user and receives the distribution of the contents according to the number of receiving units 104 and scene analysis units 109 (see FIG. S453, S454, S441, S442) and the preference degree are sequentially evaluated in the same manner as described above (S443, S444).

  In this way, the values at the time are sequentially compared with the respective preference levels of the content being viewed and some other content (S435). When the preference level of the content being viewed is high, the new analysis and the preference level evaluation are continued. However, when the preference level of the other content is high, the following other content is recommended (S446).

  When the user has a higher preference for another content than the content currently being viewed, the recommendation server 100 sends the content information of the other content to the display terminal 120 as recommended content information (S447). The sending of recommended content information is repeated until the end of the content (S448, S449).

  FIG. 5 is an explanatory diagram of data that the recommendation server 100 transmits the recommended content information to the display terminal 120. The recommended content information includes, for example, “recommended content” 501 as the event, the preference level 502 of the currently viewed content, the number 503 of other content that can be recommended and that has a higher preference level, and that can be recommended. Each preference level 504, URI 505, reason for recommendation 506, and its parameters 507 are shown.

  In FIG. 5, since the number of recommended contents is 2, the parameters (508 to 511) of the recommended contents 2 as well as the recommended contents 1 are shown. In this example, the keyword “Kitakata Ramen” appears on the 031 channel of terrestrial digital Tokyo 1 as the recommended content 1, and this preference is 0.78, which is higher than 0.52 of the content currently being viewed. The keyword “goal” in channel 101 of digital terrestrial / Tokyo 1 of content 2 is also high at 0.64, indicating that a scene to be recommended appears in two different contents.

  The display terminal 120 receives the recommended content information (S414). Using the display unit 126 of the display terminal, the user 150 is confirmed whether to view the recommended content (S415). FIG. 6 is an explanatory diagram of a display screen that is presented to the user and is selected for viewing when content is recommended from the recommendation server at the display terminal. Here, the content being viewed on the screen is displayed and the input of whether to view the recommended content 1 using the channel “031ch” keyword “Kitakata ramen” of the recommended content information is requested. (601) For example, when the user presses a determination button on a remote controller connected to the input unit 127 of the terminal, it is determined that viewing is performed (S403). When there are a plurality of recommended contents as in the above example, a plurality of contents are displayed (701) as shown in the explanatory diagram of the display screen for performing viewing selection shown in FIG. Alternatively, it may be selected by using an up / down cursor.

  When the display terminal receives the selection input from the user as described above (S416), the user is asked to determine whether to record the currently viewed content (S417). For example, when there is a remote control operation other than the designated enter button, it may be determined that there is no intention to view the recommended content, and the recommended content display may be terminated. Distribution is continued (S404).

  Next, when recording is already in progress at the recording unit 122 of the terminal and no further recording can be performed at the recording unit (S418), the recommended content is displayed on the display unit 126 at the terminal of FIG. As shown in the explanatory diagram for selecting an ongoing recording stop at the time of recording, a selection input (801) on whether to stop the recording is displayed on the screen (S419). When the user inputs a stop instruction from the remote control or the like (S405) to the display terminal (S420, S421), the ongoing recording is stopped (S422), and the recording of the content being viewed is started (S423). ). When there is no operation for a certain period of time or when there is an input indicating rejection, for example, when there is a remote control operation other than the specified enter button, it is determined that there is no intention to record the content being viewed in preference to ongoing recording. The recording stop selection display may be terminated, and the ongoing content recording on the display terminal is continued (S421). When content recording according to the limitation of the recording unit 122 and the number of receiving units 124 is sufficiently possible, the recording of the content being viewed may be started without displaying the instruction to stop recording. Thereafter, the user continues viewing behavior, but at this time, another content recommended and instructed to change is viewed (S406).

  As described above, when the recommendation server evaluates that the degree of preference is higher than the content currently being viewed by the user at the terminal, and when the user intends to change the recommended content by letting the user select display change The process of displaying the new recommended content to the user at the terminal (S411) is continued, and at the same time as the processing at the display terminal 120 and the recommendation server 100 is changed to the content currently being viewed, the recommended content 1 is changed. The content that has been viewed up to now is recorded as necessary, and can be viewed later.

  According to this embodiment, the user's preference and multiple contents are evaluated in real time on the server, and more suitable contents are recommended in units of scenes of contents that change from moment to moment. Therefore, it is possible to contribute to the improvement of the display terminal value by increasing the active time of the display terminal by guiding to another content.

  The content recommendation system according to the second embodiment of the present invention is the same system as the first embodiment described above, but any content that can be currently distributed by the recommended content providing function of the recommendation server 100 during idle time is used. This is an example of performing an operation of recommending a preset service when the preference level is low. The service here refers to functions provided by the terminal, functions of external devices connected to the terminal, web contents using mail and web technology by an external server, or idle applications such as applications that combine these functions. Is a content information service. Among these services, there may be an application server that performs content distribution.

  FIGS. 9A and 9B are process sequence diagrams for performing service recommendation of idle content information when there is no recommended content in the second exemplary embodiment of the present invention. It is assumed that the processing sequence for generating / updating user preference information according to the first embodiment of FIG. 3A is performed in advance, and the user preference information is configured in the same manner. Further, the processing sequence of FIG. 9A is performed in the same manner as described in FIG. 4 with respect to the first embodiment until the middle of the processing sequence in which the content being viewed and other content are compared and evaluated with the user preference and recommended to the terminal. The description up to the middle is omitted.

  That is, in FIG. 9A, in the middle of the processing sequence of FIG. 4, both the display terminal 120 and the recommendation server set the threshold values set in advance by the recommendation server for all the preference levels of the content being viewed and some other content. Processing (S413, S445) for determining whether or not there is an excess is added. This threshold value may be set in advance by the system administrator on the recommendation server, or the terminal shipment setting or a value set by the user using the display unit 126 and the input unit 127 on the terminal is set in advance as the user attribute. It may be transmitted to the recommendation server and recorded as an element of information in the information database 101. FIG. 10 is an explanatory diagram of a setting screen for the threshold and service setting information on the display unit of the terminal. If the preference value changes from 0 to 1 when the user 150 moves the knob display 1003 on the screen by using the left and right buttons of the remote control or the like, “none” is 0.0 (preference level). For example, “low” may be set to a threshold value that is a slider value that is calculated so that “low” is about 0.6.

  If no recommendation server 100 in the processing sequence of FIG. 9A exceeds the threshold set as described above (S445, FIG. 9 (3)), one piece of information in the user attribute information database 101 on the recommendation server Service setting information to be registered as an element is acquired (S931 in FIG. 9B). It is assumed that this service setting information is set in advance by the user. In FIG. 10, up to five services in the order of priority as recommended services, four in this example, that is, “mail”, “net video”, “recorded program”, and “game” are registered. FIG. 11 is an explanatory diagram of detailed setting of service setting information on the display unit 126 of the display terminal 120. This is a screen for setting details when the setting button in FIG. 10 is pressed. This figure shows the display unit 126 in a state where the setting button 1005 of the mail 1004 in FIG. 10 is pressed by the user's remote control operation. Is displayed. In this example, the mail receiving server address 1103 and the sending server address 1105 are set to “mail.hogehoge.ne.jp”, the user ID 1106 is set to “user-hoge”, and the login password 1109 does not display text. It has been shown. These can be input by a user's remote control operation or the like by pressing a setting button on the right of each display to make another display. These service settings may be services specified in advance by a terminal or a recommendation server, and may be set by recommending this service using its service attribute information in the same manner as the recommended content.

  As described above, the recommendation server 100 selects a service from the service setting information acquired from the user attribute information database 101 (S931) (S932). As shown in FIG. 10, the selection method may be arranged in the order of five as arranged by the user 150 in consideration of the priority, and the probability of being selected according to the priority, for example, five services set. Then, select one of these by generating 0.3, 0.25, 0.2, 0.15, 0.1 in order of priority and generating random numbers, or select five in the same way May be arranged in order. One or more recommended services selected in this way are transmitted to the display terminal 120 as recommended service information (S933).

  FIG. 12 is an explanatory diagram of data transmitted as recommended service information. The recommended service information includes, for example, “recommended service” 1201 as the event, the preference level 1202 of the content currently being viewed, the number 1203 of the currently distributable services to be recommended, a name indicating each service, 1204 and the URIs 1205 are shown for the number of recommended services. In the example of FIG. 12, since the number of recommended services is 4, parameters (1207 to 1211) of services 2 to 4 are described. In this example, the service 1 can receive the service by accessing the specific center with the URI “mail” of “svc: //center/user-hoge/S1.svc”. As the service 2, the most recently recorded program can be reproduced as the “recorded program” at the URI “svc: // rec? = Last-rec”. As the service 3, it is assumed that “game” is a URI “svc: /// internal / aux3 /”, and the video of the consumer game machine connected to the external input 3 of the terminal can be viewed here. The service 4 can be received as a “net video” by accessing the center at the URI “svc: //center/user-hoge/S3.svc”. With regard to this net video, the recommendation server may acquire the content stored in the external content server 160 or the content information thereof, and determine the recommendation order in advance according to the user's preference. The recommendation order is determined in this way, and when the terminal accesses the URI “svc: //center/user-hoge/S3.svc”, the display terminal 120 displays content information for receiving distribution from the external content server. May be received from the external content server from the terminal. Alternatively, the addresses of the respective external content servers and application servers may be set directly in the URI and transmitted to the terminal.

  Returning to FIG. 9B, when the recommended service information is sent from the recommendation server 100 (S912), the display terminal 120 receives it (S912), and presents the contents to the user 150 using the display unit 126 (S913). When the user confirms this (S902) and selects one service (S903), the terminal confirms this and obtains a URI for accessing the service from the recommended service information (S914).

  FIG. 13 is an explanatory diagram of a display screen that is displayed to the user from the display unit 126 when service recommendation is made from the recommendation server at the terminal and the service is selected. Here, along with the display of the content being viewed on the screen of the display unit, among the above-mentioned recommended service information, “Mail” of the recommended service 1, “Recorded program” of the recommended service 2, “Game” of the recommended service 3, and recommendation This indicates that the service 4 is requested to receive the “net video” service (1301), and the user uses, for example, a numeric keypad on the remote control connected to the input unit 127 of the terminal or an up / down cursor. The service is selected by pressing (S903). Note that only the recommended service 1 may be displayed and selected with the decision button, or may be displayed and selected separately one by one. Also, when there is no operation for a certain period of time or when there is an input indicating rejection, for example, when there is a remote control operation other than the designated enter button, it is determined that there is no intention to view the recommended content, and the recommended content display is terminated. For example, the content distribution currently being viewed on the terminal may be continued. Furthermore, as in the first embodiment, a process of recording the content being viewed may be performed.

  When the service selected by the user 150 is a recorded program (S915), the recorded program, for example, the last recorded program is displayed (S916). Here, the user may select the recorded program list. If the selected service is a service in the terminal, for example, “game” is selected and the external input is switched and displayed (S917), this in-terminal service, that is, the idle content information service is executed (S918). . In addition, when the selected service is a center service, for example, when an application such as “mail” or “web” is executed on the recommendation server 100 (S919), the service is requested (S920, S934), and the content is distributed from the server. You may receive the same video distribution and program distribution. In addition, when a mail or web function is already present in the terminal, it may be executed as an in-terminal service. In the case of a service connected to the external server 160, that is, a content server or an application server, such as “net video” (S922), the content information service during idle time is requested from the external server (S923, S941), and content distribution is performed. The same video distribution or program distribution may be received (S942, S924).

  First, when a service request is received as a service of a recommendation server, a specific service, for example, a service for receiving a recommendation of content according to user preference from a content group of an external content server in the description of FIG. The distribution may be received by changing the connection destination or receiving a transfer from the recommendation server, or by notifying the connection destination from the recommendation server to the terminal and connecting the terminal to the external content server that is the connection destination. As described above, the service selected by the user is received at the terminal, and can be viewed by the user by using the display unit 126, the input unit 127, and, if necessary, the decoding unit 128 (S904).

  As described above, if the recommendation server evaluates that the content currently being viewed by the user or other content that can be distributed on the terminal is lower than the set threshold, the service set in advance Can be received and viewed in a manner suitable for each.

  The content recommendation system of the third embodiment is the same system as that of the first embodiment described above, but separates content acquisition and scene analysis processing from user preference level evaluation processing. That is, resources can be reduced in the receiving unit 104, the scene analysis unit 109, and the like on the recommendation server by allowing a plurality of users to access the content distribution process and the scene search process in common. Specifically, it is possible to configure a content promotion server that does not have as many receiving units and scene analysis units as the number of products of each user and each content, but prepares them as many as the number of content and uses them in common in the user processing part To.

  FIG. 14 is a processing sequence diagram in which the content being viewed and other content in this embodiment are compared and evaluated with user preferences and recommended to the terminal. In the figure, the processing sequence in the recommendation server 100 and the external content server 160 is described, and the processing in the display terminal 120 and the user 150 is the same as that in FIG. 4, and in particular, the exchange by communication between the terminal and the recommendation server is the same. is there. Further, since the processing steps have the same contents, some of them use the same step numbers as those in FIG. Hereinafter, it demonstrates in order.

  In the recommendation server 100, processing in two systems of processing for each user (S1401) and processing for each content (S1421) may be performed in parallel and asynchronously. In the processing system for each user, user authentication may be performed by login (S221), and a processing sequence for the user may be generated. In the processing system for each content, a processing sequence may be generated for each external content distribution server, that is, for example, for each broadcast channel on a television. At this time, the reception units 104 and the scene analysis units 106 are made the same number, and the same number of processing sequences are generated. Further, when there is a separate content distribution by communication using the communication unit 108 such as the Internet, the number of scene analysis units is increased by the number of content streams from the corresponding external content distribution server.

  In the processing sequence of 100 in the recommendation server, first, processing preparation is performed in the processing system for each content (S1422). This processing preparation includes initialization of variables on the memory, connection and setting with the receiving unit 104 or the communication unit 108 for receiving content distribution from the external content server 180, and a scene analysis unit corresponding to each content 109 reservation is included. Next, each content information is acquired from an external content server (S451, S432), and distribution of content is started (S452, S433).

  On the other hand, in the processing system for each user, first, the recommendation server 100 authenticates the process in which the user 150 logs in using the display terminal 120 (S221). If authenticated, the user's information is acquired (S222), and the user's processing is actually started. Next, content information currently being viewed is acquired from the terminal (S431), and the preference for the scene is calculated by the number of contents acquired from each content analysis unit thereafter, and preparations are made for comparison across all contents (S1402). ). This preparation of processing of one content analysis system means, for example, initialization of in-memory variables, etc., acquisition of content information and securing of a communication method with all of the processing systems that analyze the corresponding content For example, as an event notification means, registration of a function as a destination of scene information or connection of a communication path is performed.

  In the processing system for each content, content scene analysis, for example, keyword acquisition is performed in the same manner as in the first embodiment while receiving delivery asynchronously for each content (S443). These keywords are transmitted to the processing system for each user by using, for example, event notification means for the function set above (S1423). The transmission of the scene information continues until the end of the content (S448). Even when the transmission ends, the process continues to the next content, for example, the next program in the television broadcast (S449) as necessary, and the program information is acquired and distributed. .

  In the processing system for each user, scene information is acquired by the above-described means corresponding to each content (S1403), and the degree of preference is evaluated for each content (S435). Next, each content preference level at that time is sequentially compared with the content preference level currently being viewed at regular intervals (S445). If the content currently being viewed is high, the next scene information is waited to be received. However, when there is a content with a high preference level for another content (S445), and when the content is higher than a predetermined threshold (YES at S446). The content information corresponding to the content with high preference is sent as recommended content information (FIG. 5) to the display terminal 120 as in the first embodiment, and the terminal performs the processing after S413 in FIG. Thus, another content is recommended to the user, and the program being viewed is recorded as necessary. If there is no content having a preference level exceeding the predetermined threshold (NO in S446), the recommended service information (FIG. 12) is obtained by performing the same processing (S931 to S933) as in the second embodiment. It may be sent to the display terminal and service recommendation of the processing of FIG. 9 may be performed.

  In this way, instead of having as many receivers and scene analyzers as the number of products of each user and each content, the processing system for each content that prepares them for the number of contents is common to the processing system parts for each user. It is possible to construct a content recommendation server that can be used for

  In the embodiment described above, the display terminal 120, the recommendation server 100, and another content server are configured as separate devices. However, a device including processes of some other devices or a part of processing is distributed among a plurality of servers. You may comprise so that a process may be performed. For example, the terminal performs all the processing of the recommendation server and makes the recommendation server unnecessary, or the recommendation server has a content distribution function and does not require another content server, or the processing system for each user in the third embodiment The processing system for each content and the processing system for each content may be further distributed.

  Further, the display terminal 120 and the recommendation server 100 may be configured by a software program using a general computer, and each unit constituting each may operate as a calculation unit.

  Further, the recommendation server 100 or another content server may have a recording function. In this case, when the user selects another content, a time difference occurs between the selection execution and the actual distribution, and the recommended scene is overlooked. It is possible to receive the distribution retroactively so as to prevent further oversight. Furthermore, recommended content information may be recorded and used so that the user views the recorded content as desired.

  In addition, when the recommended content information or recommended service information is received at the terminal, it may be set so that the user's selection confirmation is not performed. In this case, for example, the content with the highest preference or the first service is immediately selected. Deliver to and display. Thereby, instead of the user, a channel change, that is, a display change to a content or service having a high preference may be automatically performed as zapping.

DESCRIPTION OF SYMBOLS 100 Recommendation server 101 User attribute information database 102 Authentication part 103 Antenna 104 Reception part 105 User preference information database 106 Preference evaluation part 107 Calculation part 108 Scene analysis part 109 Communication part 110 Service setting information database 111 Service setting part 120 Display terminal 121 Communication part 122 Recording unit 123 Database of recorded data 124 Reception unit 125 Calculation unit 126 Display unit 127 Input unit 128 Decoding unit 130 Antenna 150 User 160 External content server 161 Antenna.

Claims (20)

A server configured to be connectable to a display terminal and a content providing source,
The server
Obtaining user preference information indicating the preference of the user using the display terminal;
Correlation between the first content currently being viewed on the display terminal and at least one other content other than the first content provided from the content providing source in parallel with the user preference information Are evaluated to determine the preferred degree of preference,
When the other content having the higher preference preference than the first content being viewed exists, recommended content information for recommending the other content having the higher preference preference level is generated and transmitted to the display terminal. A server characterized by that. In claim 1,
A server characterized in that scene information of each content is sequentially evaluated based on the preference suitability level, and the preference suitability level is determined for each content scene. In claim 1,
Obtaining in advance the threshold value of the preference level in the server and idle content information defining one or more contents to be displayed on the display terminal when there is no content to be recommended,
When the preference suitability level for evaluating the content being viewed and the other content does not exceed the preference threshold value, the idle content information is sent from the idle content information to the display terminal. A server characterized by In claim 1,
The server
A function of acquiring the user preference information;
A function of receiving viewing content information specifying the first content from the display terminal;
A function of receiving the first content from the content providing source synchronously or in advance with a display time on the display terminal;
A function of sequentially receiving at least one other content other than the first content from the content providing source in synchronization with the displayable time on the display terminal or in advance from the displayable time;
With respect to the first content and the other content, respectively, a function for evaluating a correlation with the user preference information and obtaining a preference suitable degree,
Regarding the display terminal, when there is the other content having the higher preference preference degree than the first content being viewed, recommended content information for specifying the other content item having the higher preference preference degree as recommended content Generating and transmitting the recommended content information to the display terminal;
And a function of changing the display on the display terminal from the first content being viewed to the recommended content when a request to change to the recommended content is received from the display terminal. Server. In claim 4,
The server
As many as the number of products of the display terminal and the content, the receiving unit and the scene analysis unit,
Two processing systems, a processing system for each user of the display terminal and a processing system for each content, are performed in parallel and asynchronously,
The processing system for each content performs processing for each external content distribution server that distributes the content. In claim 4,
The server
A server comprising: a reception unit and a scene analysis unit as many as the number of contents; and a user processing unit that allows a plurality of users to commonly access the content distribution process and the scene search process in the content. In claim 4,
The function for evaluating the correlation with the user preference information and obtaining the preference preference degree is given a threshold value for determining the preference preference degree in advance.
The function for generating the recommended content information includes recommended content information at idle time in which one or a plurality of contents are ordered as content to be displayed on the display terminal when there is no other content to be recommended. A server characterized by In claim 7,
The function of generating the recommended content information has the function of generating the recommended content information when the preference suitability degree for evaluating the viewing content and the other content does not exceed the preference threshold value. A server, wherein recommended content information is sent to the display terminal. In claim 7,
The function of generating the recommended content information includes adding a preference suitability level of the content being viewed and a preference suitability level of the recommended content to the recommended content information sent from the server, and sending it to the display terminal. Feature server. In claim 1,
When evaluating the correlation with the user preference information and obtaining the preference preference degree,
Acquire scene data consisting of video, audio, and character strings constituting the content, or scene labels indicating the contents of the scene data, as a scene unit that is sufficiently shorter than the duration of the content. And
The user preference information is composed of preference data composed of video, audio and character strings, or preference labels indicating the contents of the preference data, and respective weights,
The preference data and the scene data, or the preference label and the scene label are compared, and the calculation result using the weight is sequentially evaluated as the preference suitability for each scene unit while the content is continuously received. A server characterized by that. In claim 1,
The server
An arithmetic unit;
A receiving unit capable of receiving two or more of the contents;
A scene analysis unit capable of sequentially analyzing two or more of the contents;
A communication unit that performs communication using a network;
A user preference information section for storing preference information for each user;
A preference evaluation unit for comparing and evaluating the user preference information and scene data and scene labels from the scene analysis unit;
Means for acquiring user preference information indicating the user's preference from the user preference information unit by identifying the user with the display terminal that performs communication using the network;
Means for acquiring viewing content information for identifying the first content currently being viewed from the display terminal using the communication unit;
Means for receiving the first content specified from the viewing content information in synchronization with a display time on the display terminal or separately in advance;
Means for sequentially receiving at least one second content other than the first content being viewed from the content providing source in synchronization with the displayable time on the display terminal or in advance of the displayable time; ,
Means for obtaining a preference preference degree by evaluating a correlation with the user preference information according to an analysis criterion for predetermining the first and second contents;
When the preference suitability of the second content is higher than the first content being viewed on the display terminal, recommended content information for specifying the second content as recommended content from the communication unit A server comprising: means for sending to a display terminal. In claim 11,
A service setting information section that holds the preference level threshold value and idle content information that defines one or more content to be displayed on the display terminal when there is no content to be recommended;
And a service setting unit that identifies or orders idle content,
Obtaining in advance from the preference threshold, idle content information and the service setting information section,
When neither the preference preference degree that evaluates the viewing content or the different content exceeds the preference level threshold, the service setting information unit identifies one of the idle content information or orders a plurality of content items A server characterized in that recommended idle content information is sent from the communication unit to the display terminal. A content display terminal configured to be connectable to a content providing source and a server,
The display terminal is
A function of providing information of a user who uses the display terminal to the server;
A function of sending viewing content information identifying the first content being viewed displayed on the display unit to the server;
In the server, based on a result of determining that another content having a higher user preference than the first content being viewed is provided in parallel from the content providing source, from the server A function of receiving recommended content information for recommending other content with a high degree of preference,
A function of acquiring the other content specified from the received recommended content information from the content providing source and displaying the other content on the display unit in place of the first content being viewed; Content display terminal. In claim 13,
An arithmetic unit;
A first receiving unit capable of receiving a plurality of contents from the content providing source;
The display unit;
A communication unit that communicates with the server using a network;
A function of sending the viewing content information specifying the content being viewed displayed on the display unit to the server using the communication unit;
A function of acquiring the recommended content information specified as the recommended content sent from the server from the communication unit;
A content display terminal comprising: a function of acquiring the recommended content specified from the recommended content information from the first receiving unit, and displaying the recommended content on the display unit in place of the content being viewed. In claim 14,
A service setting information section that holds a preference level threshold and idle content information that defines one or more contents to be displayed on the display terminal when there is no content to be recommended;
And a service setting unit that identifies or orders idle content,
A function for specifying one of the idle content information in the service setting information section when the preference suitability for evaluating the viewing content and the other content does not exceed the preference threshold;
A content display terminal, comprising: a function of displaying idle content specified from the idle content information with the content being viewed on the display unit. In claim 15,
A second receiving unit and a recording unit for recording the content;
A function of recording a continuation of the content being viewed received by the first or second receiving unit using the recording unit;
A content display terminal comprising: a function of displaying the recommended content or the idle content received by the other receiving unit on the display unit in place of the viewing content. In claim 14,
A function of recording a continuation of the content being viewed in the display terminal when changing from the content being viewed to recommended content or recommended idle content;
A content display terminal, wherein the recommended content or recommended idle content is separately displayed. In claim 14,
The display terminal is
A content display terminal having a function of generating information about a user who uses the display terminal and user preference information indicating the user's preference and providing the user preference information to the server. A display terminal for displaying content and a server;
The display terminal has a function of sending viewing content information specifying content currently being viewed to the server,
The server
A function of acquiring user preference information indicating the user preference of the display terminal, and a function of acquiring a plurality of other contents from an external content providing source;
A function of receiving the content specified from the viewing content information in synchronization with a display time on the display terminal or separately receiving in advance;
One or more other contents other than the content being viewed on the display terminal are sequentially received from the other external content providing sources in synchronization with the displayable time on the display terminal or in advance of the displayable time. However, a function for determining the degree of preference preference by evaluating the correlation with the user preference information based on the analysis criteria that predetermine these contents,
A function of sending recommended content information for specifying the content as recommended content to the display terminal when the preference suitability of the other content is higher than the content being viewed on the display terminal;
The display terminal has a function of receiving the recommended content information and changing the content being viewed to the recommended content. In claim 19,
The server
A function of acquiring in advance a threshold value of preference and idle content information defining one or more contents to be displayed on the display terminal when there is no content to be recommended;
The content recommendation system, wherein the display terminal has a function of receiving recommended content information at the time of idle and changing display of the content being viewed to the content recommended at idle.

Images