Recommendation device and recommendation method

Abstract

This recommendation device comprises: a reception unit that receives, from a plurality of users, moving image information pertaining to a moving image including a plurality of scenes arranged in chronological order, the moving image information including information relating to a viewed user and information relating to excitement in each scene; a determination unit that determines a viewing moving image from among a plurality of candidate moving images on the basis of the moving image information relating to a plurality of viewing users viewing the moving image together; and a recommendation unit that recommends the viewing moving image determined by the determination unit.

Description

Recommendation device and recommendation method

[0001] One aspect of the present disclosure relates to a recommendation device and a recommendation method.

[0002] Patent Literature 1 discloses a technique for identifying the average level of excitement regarding a video among multiple users in order to know the reactions of other users who are simultaneously watching the video.

[0003] JP 2011-182109 A

[0004] In recent years, it has become common for multiple people to watch videos together. In such multi-person viewing, users can get excited when their favorite parts (scenes) of a video overlap, improving the viewing experience. However, it is not easy for users who meet for the first time to find a video that shares their favorite parts (a video that they can enjoy together).

[0005] The present disclosure has been made in consideration of the above-described situation, and provides a recommendation device and recommendation method that can appropriately recommend videos that can be enjoyed by multiple users who watch the videos together.

[0006] A recommendation device according to one aspect of the present disclosure includes a reception unit that receives video information from multiple users, the video including multiple scenes in chronological order, the video including information about the users who have viewed the video and information about the excitement in each scene; a determination unit that determines a video to be viewed from multiple candidate videos based on video information about the multiple users who have viewed the video together; and a recommendation unit that recommends the video to be viewed determined by the determination unit.

[0007] A recommendation device according to one aspect of the present disclosure receives video information from multiple users, the video information including information about the users and information about the excitement in each scene of the video. Then, a viewing video is determined based on the video information about the multiple viewing users who watch the video together, i.e., information about the excitement in each scene, and the viewing video is recommended. With this configuration, the recommended viewing video is determined taking into account information about the excitement in each scene for the multiple viewing users who watch the video together, so it is possible to appropriately recommend a viewing video that is likely to be popular among multiple viewing users who watch the video together. As described above, the recommendation device according to one aspect of the present disclosure can appropriately recommend a video that will be popular among multiple users who watch the video together. This can improve the user's viewing experience.

[0008] According to one aspect of the present disclosure, it is possible to appropriately recommend videos that can get multiple users excited about watching the videos together.

[0009] FIG. 1 is a diagram illustrating an overview of a recommendation device according to this embodiment. FIG. 2 is a diagram illustrating a functional configuration of a recommendation device according to this embodiment. FIG. 3 is a diagram illustrating an overview of creating a machine learning model. FIG. 4 is a diagram illustrating details of creating a machine learning model. FIG. 5 is a diagram illustrating an example of a method for determining a video to be viewed. FIG. 6 is a diagram illustrating an example of a method for determining a video to be viewed. FIG. 7 is a flowchart illustrating processing executed by the recommendation device. FIG. 8 is a diagram illustrating an example of a hardware configuration of a recommendation device.

[0010] Hereinafter, the embodiments will be described in detail with reference to the drawings. In the description, the same elements or elements having the same functions are denoted by the same reference numerals, and redundant description will be omitted.

[0011] FIG. 1 is a diagram illustrating an overview of a recommendation device according to this embodiment. The recommendation device is a device that recommends to multiple viewing users who watch videos together (e.g., simultaneously), videos that are expected to be enjoyed by the multiple viewing users together. The recommendation device is configured to determine such viewing videos and recommend the determined viewing videos. The multiple viewing users here may be multiple users who do not know each other, or multiple users who know each other. The multiple viewing users may be, for example, multiple users who belong to a common community, multiple users who receive a common service, or multiple randomly selected users. The community may be a community formed based on a common hobby, or may be a community in a metaverse space. For example, if the community is a robot anime community, videos that users who share the common hobby of robot anime can enjoy are recommended, thereby fostering a more united community. For example, if the community is a community in the metaverse space, videos that users in the metaverse space can enjoy (for example, exciting videos that make use of the metaverse space (such as singing and dancing together)) can be recommended, thereby fostering a more cohesive community.

[0012] In the following description, a video is assumed to be viewed simultaneously by multiple viewing users, but this is not limited thereto, and the video may be viewed at different times as long as it is viewed together. Note that "viewed simultaneously" may mean viewing in the same space at the same time, or viewing in different spaces at the same time.

[0013] The recommendation device may receive information about the excitement of each scene in a video that includes multiple scenes in chronological order in advance, and based on the received information, determine and recommend videos to be viewed simultaneously by multiple viewing users. In the graphs shown in FIGS. 1(a) and 1(b), the horizontal axis represents time and the vertical axis represents excitement level. Such excitement level is quantified, for example, by the number of times a scene is repeated (details will be described later). FIG. 1(a) shows the excitement levels of a woman in her 30s and a man in his 20s for a video X that includes multiple scenes in chronological order. The recommendation device may create a machine learning model that predicts information about the excitement level for a target user by associating and learning information about the viewing user (e.g., a woman in her 30s), information about each scene (such as information about the actors and lines in each scene), and information about the excitement level in each scene (such as the excitement level depending on the number of repetitions). Then, as shown in Figure 1(b), the recommendation device predicts information about the excitement of the target users for each scene of video Y, which is different from video X, by using the above-mentioned machine learning model. The recommendation device predicts information about the excitement of multiple target users (viewing users) for each of multiple candidate videos, and recommends viewing videos that multiple viewing users can enjoy together, taking into account the prediction results (details will be described later). Note that the method using the machine learning model is merely an example and is not limited to this.

[0014] Fig. 2 is a diagram showing the functional configuration of a recommendation device 20 according to this embodiment. Fig. 2 shows a recommendation system including the recommendation device 20 and a terminal 10. In the recommendation system, the recommendation device 20 and the multiple terminals 10 are configured to be able to communicate with each other via a network including a wireless communication network and a fixed communication network.

[0015] The terminal 10 is a terminal used by a user to watch a video. The terminal 10 may be, for example, a personal computer, a smartphone, a tablet terminal, a feature phone, a server device, a game console, or the like. Although FIG. 2 illustrates only one terminal 10, the recommendation system actually includes multiple terminals 10. When a user watches a video, the terminal 10 transmits video information (details will be described later) including various information related to the video viewing to the recommendation device 20. The timing for transmitting the video information from the terminal 10 to the recommendation device 20 may be when the user finishes watching the video, while the user is watching the video (in real time), or at predetermined time intervals (periodic timing). When video information is transmitted at predetermined time intervals, video information related to all videos viewed between the previous transmission and the current transmission may be transmitted together. In this case, the terminal 10 includes both the terminal of a user who sends video information to the recommendation device 20 in the learning process described below, and the terminal of a viewing user (a user who watches a video simultaneously with multiple people) who is recommended a video to watch.

[0016] The recommendation device 20 includes, as functional components, a reception unit 21, a model creation unit 22, a determination unit 23, a recommendation unit 24, and a storage unit 25. Note that the recommendation device 20 does not need to have the model creation unit 22 when recommending videos to be viewed without using a machine learning model. Below, the function of the "learning process" related to the creation of a machine learning model and the function of the "recommendation process" related to the recommendation of videos to be viewed will be described in order.

[0017] (Learning process) The reception unit 21 receives video information related to videos viewed by users from each of the terminals 10 (i.e., multiple users). In this way, the reception unit 21 receives multiple pieces of video information. The video information is information related to a video that includes multiple scenes in chronological order, including information about the user who viewed the video, information about each scene, and information about the excitement in each scene.

[0018] The information about the user who viewed the video is various information about the user who viewed the video via the terminal 10. The reception unit 21 may receive information about the user, such as the user's age (or generation), gender, place of residence, hobbies, and preferences, as the information about the user who viewed the video. Here, as an example, the description will be given assuming that the reception unit 21 receives the user's generation and gender as the information about the user who viewed the video.

[0019] The scene information is various information about the scene viewed by the user. The reception unit 21 may receive information about the performers in the scene as the scene information. The performer information is information about the performers who appear in the video if the video is live action, and information about the characters who appear in the video if the video is animation, CG, or the like. The performer information may be information indicating the performer (information such as the performer name that identifies the performer), information indicating the performer's attributes (e.g., young male, etc.), or information indicating the performer's role in the video (e.g., protagonist, partner, rival, parent and child, couple, boss, mentor, best friend, etc.). Note that the performer information can be made more versatile by vectorizing the performer information rather than the performer's name.

[0020] The reception unit 21 may receive a line sentiment analysis result as information about a scene. The line is a line spoken by a character in a video. The line sentiment analysis result is an analysis result of an emotion estimated from transcribed information of the line (or from the information and information about the music played at the same time), and is information indicating, for example, anger, sadness, joy, neutral, etc. The line sentiment analysis may be performed on the terminal 10, on the recommendation device 20, or by another server, etc.

[0021] The receiving unit 21 may receive information about music as information about a scene. The information about music may be information indicating whether music is playing, information indicating the type of music being played (e.g., bright music, dark music, etc.), or information indicating the role of the music being played in the video (e.g., main theme, insert song, etc.).

[0022] The receiving unit 21 may receive information about camera work or camera angle as information about a scene. The information about camera work or camera angle may be information indicating, for example, pan, tilt, zoom in, zoom out, high angle, low angle, horizontal angle, etc.

[0023] The receiving unit 21 may receive, as the information related to the excitement of a scene, at least one of the number of times the corresponding scene is repeated, information related to the volume of the playback of the corresponding scene, and information related to the reaction of the viewing user to the corresponding scene. The receiving unit 21 may also receive, as the information related to the excitement, information other than the above, as long as the information specifies the degree of excitement in the scene.

[0024] The number of times a corresponding scene has been repeated is information indicating how many times the viewing user has repeated the scene. The number of times a corresponding scene has been repeated may be the number of times it has been repeated within a predetermined period (for example, within a predetermined period of several minutes to several days). It is estimated that the greater the number of times a scene has been repeated, the more interested the viewing user is in the scene and the higher the degree of excitement.

[0025] The information about the volume during playback of the corresponding scene may be information indicating the volume itself or information indicating a change in volume over time. The louder the volume, the more interested the viewer is in the scene and the higher the level of excitement is estimated to be. Furthermore, if the volume is gradually increased during the scene, it is estimated that the viewer is interested in the scene and the higher the level of excitement is estimated to be.

[0026] The information regarding the reaction of the viewing user to the corresponding scene is information indicating how the viewing user evaluated the scene. The information regarding the reaction of the viewing user may be, for example, information indicating a qualitative evaluation such as "high rating" or "low rating," or information indicating a quantitative evaluation such as a score. It is estimated that the higher the rating, the more interested the viewing user is in the scene and the higher the degree of excitement.

[0027] The reception unit 21 stores a plurality of pieces of video information received from each terminal 10 in the storage unit 25. The storage unit 25 stores the plurality of pieces of video information input from the reception unit 21. The storage unit 25 also stores a machine learning model created by the model creation unit 22 (described later). The storage unit 25 also stores information determined by the determination unit 23 (described later).

[0028] The model creation unit 22 creates a machine learning model that predicts information related to excitement by performing machine learning using multiple pieces of video information stored in the storage unit 25. The model creation unit 22 stores the created machine learning model in the storage unit 25.

[0029] FIG. 3 is a diagram illustrating an overview of creating a machine learning model. As shown in FIG. 3, assume that video information for multiple videos (here, as an example, a woman in her 30s and a man in his 20s) is stored in the storage unit 25. Each piece of video information is information related to a video containing multiple scenes, and assumes that the degree of excitement in, for example, six consecutive scenes (scenes from 0 to 10 seconds, 11 to 20 seconds, 21 to 30 seconds, 31 to 40 seconds, 41 to 50 seconds, and 51 to 60 seconds) is indicated by a score. Using this video information, the model creation unit 22 machine-learns information indicating which users (here, a woman in her 30s or a man in his 20s) were excited by which scenes, and to what extent, thereby creating a machine learning model that predicts information about the excitement level for a target user.

[0030] FIG. 4 is a diagram illustrating the details of creating a machine learning model. FIG. 4 shows an example of detailed video information for a certain viewing user (a man in his twenties). Here, the video information shows scores representing the degree of excitement in six consecutive scenes (scenes from 0 to 10 seconds, 11 to 20 seconds, 21 to 30 seconds, 31 to 40 seconds, 41 to 50 seconds, and 51 to 60 seconds). These scores may be normalized to a value between 0 and 100, for example. These scores may be determined from statistical values ​​for the same attribute (a man in his twenties) or may be values ​​for a specific individual. Information about each scene also includes information about the cast, sentiment analysis results for the dialogue, and information about the music. For example, for the "0-10 second scene," the cast's "protagonist" and "partner" are "True," "rival" is "False" (the protagonist and partner appear, the rival does not appear), the sentiment analysis results for the lines show that "sadness" is "True" and "joy" is "False" (the lines are related to sadness, not joy), the music's "theme song" is "False," and "insert song" is "False" (the music is not the theme song, not an insert song).

[0031] The model creation unit 22 performs machine learning on multiple pieces of video information, using information about the user who viewed the video and information about each scene as explanatory variables and information about the excitement level of each scene as a target variable, to create a machine learning model that predicts information about the excitement level. For example, in the example of the "0-10 second scene" of video information shown in FIG. 4, information about the user who viewed the video (a man in his 20s) and information about the scene (the cast members are the protagonist and his partner, the sentiment analysis result of the dialogue is sad, and no theme song or insert song is played) are used as explanatory variables, and a score of "80" indicating the excitement level is used as the target variable. This process is performed for each scene of each piece of video information, and machine learning is performed to create a machine learning model that predicts information about the excitement level. Machine learning may be performed using an algorithm such as a random forest or a recurrent neural network. The machine learning model predicts information about the excitement level by inputting information about the user and information about the scene. Information about each scene may be characterized, and a machine learning model that semantically characterizes image features such as CLIP may be used for the characterization.

[0032] The model creation unit 22 may create a machine learning model by performing machine learning using only scenes of a plurality of video information whose excitement level (score) indicated by the excitement information is equal to or greater than a predetermined value. For example, when using the video information shown in Figure 4, if it is predetermined to use only scenes whose excitement level (score) is equal to or greater than 50, the "scene from 0 to 10 seconds" (score 80), the "scene from 31 to 40 seconds" (score 50), and the "scene from 51 to 60 seconds" (score 100) may be used as learning data, and other scenes may be excluded from the learning data.

[0033] The model creation unit 22 may create a machine learning model by learning not only information about the excitement of a single scene, but also information about changes in the information about the excitement over time. For example, the model creation unit 22 may perform machine learning using information about the excitement of each scene and information about the excitement of the scene following each scene as objective variables, to create a machine learning model that performs a first inference that predicts information about the excitement of each scene and a second inference that predicts information about the excitement of the scene following each scene. For example, in the example shown in FIG. 4 , machine learning may be performed using information about the excitement of successive scenes as objective variables, such that "40," which is the excitement level (score) of the "scene from 21 to 30 seconds," is set as "target variable 1," and "50," which is the excitement level (score) of the subsequent "scene from 31 to 40 seconds," is set as "target variable 2."

[0034] (Recommendation processing) The reception unit 21 receives information about multiple viewing users who watch a video together (simultaneously) from the terminal 10 of each viewing user. The information about the viewing user may be any information that can identify the viewing user, such as a user ID that uniquely identifies the viewing user. The reception unit 21 may receive the above-mentioned video information from the viewing user. For example, the reception unit 21 may receive information about each scene of a candidate video, which is a video that will be a candidate for a viewing video that will ultimately be recommended, from the terminal 10 of each viewing user. Note that the information about each scene of the candidate video does not necessarily have to be received from the terminal 10, but may be received from another server or the like, or may be stored in advance in the storage unit 25. The reception unit 21 stores the received various information in the storage unit 25.

[0035] The determination unit 23 determines a video to be viewed from among multiple candidate videos based on video information about multiple viewing users who are watching videos together (simultaneously). The determination unit 23 stores the determined information (information indicating the viewed video) in the storage unit 25. By referring to the storage unit 25, the determination unit 23 identifies the video information of the multiple viewing users (information about the viewing user, information about the excitement in each viewed scene, and information about each scene) and information about each scene of the multiple candidate videos.

[0036] The determination unit 23 then inputs information about the viewing user and information about each scene of the candidate video into the machine learning model stored in the storage unit 25, and predicts information about the excitement in each scene. The determination unit 23 executes this process for each combination of the viewing user and the candidate video, and determines the video to be viewed based on the prediction result.

[0037] The determination unit 23 may prioritize candidate videos that have a higher degree of agreement between scenes that increase the level of excitement among multiple viewing users as the viewed video. For example, suppose there are three candidate videos X, Y, and Z, each of which contains six consecutive scenes (scenes from 0 to 10 seconds, 11 to 20 seconds, 21 to 30 seconds, 31 to 40 seconds, 41 to 50 seconds, and 51 to 60 seconds). For example, if an excitement level of "50" or higher is determined to be "high excitement," then candidate video X has two scenes that increase the level of excitement among multiple viewing users, candidate video Y has one scene that increase the level of excitement among multiple viewing users, and candidate video Z has three scenes that increase the level of excitement among multiple viewing users. In this case, the determination unit 23 determines candidate video Z, which has the highest degree of agreement between scenes that increase the level of excitement among multiple viewing users, as the viewed video.

[0038] The determination unit 23 may prioritize a candidate video to be viewed that has a smaller variation in excitement among multiple viewing users, i.e., that has a higher likelihood of matching exciting points. The determination unit 23 may prioritize a candidate video to be viewed that has a smaller average value of variance of the excitement levels of each scene among multiple viewing users.

[0039] FIG. 5 is a diagram illustrating an example of a viewing video determination method. FIG. 5(a) shows scores indicating the degree of excitement of each viewing user (here, as an example, a woman in her 30s and a man in his 20s) for six consecutive scenes (0-10 seconds, 11-20 seconds, 21-30 seconds, 31-40 seconds, 41-50 seconds, and 51-60 seconds) of video X, a candidate video. The determination unit 23 derives the variance of the excitement levels among multiple viewing users for each scene of video X, and finally derives the average value (here, "100") of the variances of the excitement levels for each scene. Similarly, FIG. 5(b) shows scores indicating the degree of excitement of each viewing user (here, as an example, a woman in her 30s and a man in his 20s) for six consecutive scenes (0-10 seconds, 11-20 seconds, 21-30 seconds, 31-40 seconds, 41-50 seconds, and 51-60 seconds) of video Y, a candidate video. The determination unit 23 derives the variance of the excitement levels among multiple viewing users for each scene of video Y, and finally derives the average value (here, "25") of the variance of the excitement levels for each scene. Now, assuming that the candidate videos are video X and video Y, the determination unit 23 determines video Y, which has the smallest average value of the variance of the excitement levels for each scene, as the video to be viewed.

[0040] The determination unit 23 may derive the average value of the variance of the excitement level of each of the above-mentioned scenes for a specific user group identified from information about the viewing users, and may prioritize candidate videos with smaller average variance values ​​for the specific user group as the viewing video.

[0041] FIG. 6 is a diagram illustrating an example of a viewing video determination method. FIG. 6 shows scores representing the degree of excitement of each viewing user (here, a woman in her 30s and a man in his 20s, as an example) for six consecutive scenes (scenes from 0 to 10 seconds, 11 to 20 seconds, 21 to 30 seconds, 31 to 40 seconds, 41 to 50 seconds, and 51 to 60 seconds) of video X, a candidate video. The determination unit 23 derives the variance of the excitement levels among multiple viewing users for each scene of video X, and finally derives the average value (here, "100") of the variances of the excitement levels for each scene. Furthermore, the determination unit 23 derives the average value (here, "15") of the variances of the excitement levels for each of the above-mentioned scenes for each user group that shares information about the viewing users, for example, their hometowns (here, Saitama Prefecture). Video X has a large discrepancy between the average value of the variance of the excitement level of each scene for all viewing users (here, "100") and the average value of the variance of the excitement level of each scene for a group of users whose hometown is Saitama Prefecture (here, "15"), and can be said to be a video that is likely to be exciting for users from Saitama Prefecture (users who can get excited together). By recommending such video X, it is possible to recommend videos that are extremely likely to be exciting for a group of users with certain attributes (here, a group of users from Saitama Prefecture). Note that using a user group unit based on a specific hometown is just one example, and user groups based on, for example, age, gender, hobbies, preferences (soccer fans, idol fans, food lovers, robot anime fans), etc. may also be used.

[0042] The determination unit 23 may determine a representative viewing user based on the similarity in viewing tendencies among multiple viewing users according to information regarding the excitement in each scene, and may determine the video to be viewed based on the viewing tendencies of the representative viewing user.

[0043] The determination unit 23 derives the absolute value of the difference in excitement level between each viewing user for each scene of the same video, and determines that viewing users with a small absolute value of the difference in excitement level have similar viewing tendencies and assigns them to the same group. When a group has been created, the determination unit 23 assigns viewing users whose viewing tendency similarity with any viewing user in the group is within a predetermined value to the same group, and assigns viewing users whose similarity is not within a predetermined threshold to a different group. With the threshold set in this way, the determination unit 23 determines the similarity of viewing tendencies with existing groups for all viewing users, and assigns any viewing users who are not similar to any of the existing groups to a new group. The determination unit 23 repeats the above process until the number of groups does not change.

[0044] When a plurality of viewing users are divided into a plurality of groups whose viewing tendencies are similar to each other (when there are not one but multiple groups), the determination unit 23 may determine a representative viewing user from among the viewing users in the group that includes the largest number of viewing users. For example, the determination unit 23 may determine the viewing user whose viewing tendencies are most similar to those of other viewing users in the group as the representative viewing user.

[0045] When a plurality of viewing users are divided into a plurality of groups, the determination unit 23 may determine a representative viewing user from among viewing users in a small group (a group that is not the group with the largest number of viewing users) who are highly enthusiastic about watching videos. The group who are highly enthusiastic about watching videos may be, for example, a group that pays a high average amount for video services.

[0046] Furthermore, when multiple viewing users are divided into multiple groups, the determination unit 23 may determine a representative viewing user from each group. In this case, the determination unit 23 may determine the viewing video for each group based on the viewing tendencies of the representative viewing user. For example, in the case of a festival or the like in the metaverse space, multiple viewing videos can be played back spatially, so that the viewing videos for each of the above-mentioned groups can be played back.

[0047] The recommendation unit 24 identifies the viewing video determined by the determination unit 23 by referring to the storage unit 25, and recommends the viewing video to the terminals 10 of the multiple viewing users. The multiple viewing users watch the recommended viewing video via the terminals 10.

[0048] Next, the process executed by the recommendation device 20 will be described with reference to Fig. 7. Fig. 7 is a flowchart showing the process executed by the recommendation device 20.

[0049] 7 , first, the recommendation device 20 receives video information (plural pieces of video information) from each terminal 10 (step S1). The video information includes information about the user who viewed the video, information about each scene included in the video, and information about the excitement of each scene included in the video.

[0050] Next, the recommendation device 20 performs machine learning on multiple video information using information about the viewing user and information about each scene as explanatory variables and information about the excitement in each scene as the objective variable, and creates a machine learning model that predicts information about the excitement (step S2).

[0051] Next, in the recommendation device 20, information about the viewing user and information about each scene of the candidate video are input into the machine learning model, and information about the excitement in each scene is predicted for each combination of viewing user and candidate video, and the video to be viewed is determined based on the prediction results (step S3).

[0052] Finally, the video to be viewed is recommended to the terminals 10 of a plurality of viewing users (step S4).

[0053] Next, the effects of the recommendation device 20 according to this embodiment will be described.

[0054] The recommendation device 20 of this embodiment includes a reception unit 21 that receives video information from multiple users, the video including multiple scenes in chronological order, including information about the users who watched the video and information about the excitement in each scene; a determination unit 23 that determines a video to watch from multiple candidate videos based on video information about multiple viewing users who watch the video together; and a recommendation unit 24 that recommends the video to watch determined by the determination unit 23.

[0055] The recommendation device 20 according to this embodiment receives video information from multiple users, including information about the users and information about the excitement in each scene of the video. Then, a viewing video is determined based on the video information about the multiple viewing users who watch the video together, i.e., information about the excitement in each scene, and the viewing video is recommended. According to this configuration, the recommended viewing video is determined taking into account information about the excitement in each scene for the multiple viewing users who watch the video together, so it is possible to appropriately recommend a viewing video that is likely to be popular among multiple viewing users who watch the video together. As described above, the recommendation device 20 according to this embodiment can appropriately recommend a video that can be enjoyed by multiple users who watch the video together. This can improve the user's viewing experience.

[0056] The determination unit 23 may determine the video to be viewed as a video that is viewed simultaneously by a plurality of viewing users, thereby making it possible to appropriately recommend a video that can be enjoyed by multiple viewing users simultaneously.

[0057] The recommendation device 20 further includes a model creation unit 22 that creates a machine learning model that predicts information about the excitement of each scene included in a video, wherein the reception unit 21 receives video information that further includes information about each scene, and the model creation unit 22 performs machine learning on the multiple video information using information about the user and information about each scene as explanatory variables and information about the excitement in each scene as a target variable to create a machine learning model, and the determination unit 23 inputs information about the viewing user and information about each scene of the candidate video into the machine learning model to predict information about the excitement in each scene for each combination of the viewing user and the candidate video, and determines the video to be viewed based on the prediction result. In this way, by using a machine learning model that has machine-learned information about the excitement of multiple users, it is possible to appropriately predict the excitement level of each viewing user in each scene of the candidate video and appropriately recommend videos that each viewing user can enjoy together.

[0058] The determining unit 23 may preferentially determine a candidate video to be viewed if the candidate video has a higher degree of coincidence of scenes that are exciting to a plurality of viewing users. With this configuration, it is possible to appropriately recommend videos that allow each viewing user to get excited about the same scenes together.

[0059] The determination unit 23 may prioritize a candidate video that has a smaller variation in the degree of excitement among a plurality of viewing users as the video to be viewed. With this configuration, it is possible to appropriately recommend a video that can be equally exciting for each viewing user.

[0060] The determination unit 23 may prioritize a candidate video for viewing, the smaller the average value of the variance of the excitement level of each scene among multiple viewing users. With this configuration, it is possible to appropriately recommend videos that can equally excite each viewing user.

[0061] The determination unit 23 may derive the average value of the variance for each specific user group identified from the information about the viewing users, and prioritize candidate videos with smaller average values ​​of the variance for the specific user group as the viewing video. With this configuration, it is possible to appropriately recommend videos that can be enjoyed together, specifically for a desired user group.

[0062] The determination unit 23 may determine a representative viewing user based on the similarity of viewing tendencies among multiple viewing users according to information related to the excitement in each scene, and may determine a viewing video based on the viewing tendencies of the representative viewing user. With this configuration, for example, a viewing user whose viewing tendencies are highly similar to those of many viewing users is selected as the representative viewing user, and as a result, videos that are in line with the viewing tendencies of many viewing users (i.e., videos that many viewing users can enjoy together) can be recommended.

[0063] When a plurality of viewing users are divided into a plurality of groups based on the similarity of their viewing tendencies, the determination unit 23 may determine a representative viewing user from among the viewing users in the group that includes the largest number of viewing users. With this configuration, it is possible to recommend videos that many viewing users can enjoy together.

[0064] The recommendation device and recommendation method of the present disclosure have the following configuration.

[0065] [1] A recommendation device comprising: a reception unit that receives video information from multiple users, the video including multiple scenes in chronological order, the video including information about the users who watched the video and information about the excitement in each scene; a determination unit that determines a video to watch from multiple candidate videos based on the video information about multiple viewing users who watch the video together; and a recommendation unit that recommends the video to watch determined by the determination unit.

[0066] [2] The recommendation device according to [1], wherein the determination unit determines the viewed video as a video that is viewed simultaneously by the multiple viewing users.

[0067] [3] The recommendation device according to [1] or [2], further comprising a model creation unit that creates a machine learning model that predicts information related to the excitement of each scene included in a video, wherein the reception unit receives the video information that further includes information related to each scene, the model creation unit performs machine learning on a plurality of pieces of video information using information related to the user and information related to each scene as explanatory variables and information related to the excitement of each scene as a target variable, and creates the machine learning model, and the determination unit inputs information related to the viewing user and information related to each scene of the candidate video into the machine learning model to predict information related to the excitement of each scene for each combination of the viewing user and the candidate video, and determines the video to be viewed based on the prediction result.

[0068] [4] The recommendation device according to [3], wherein the determination unit prioritizes determining a candidate video as the video to be viewed if the candidate video has a higher degree of coincidence between scenes that are highly exciting among the multiple viewing users.

[0069] [5] The recommendation device according to [3], wherein the determination unit prioritizes the candidate video with smaller variation in excitement level among the plurality of viewing users as the viewed video.

[0070] [6] The recommendation device according to [5], wherein the determination unit prioritizes determining a candidate video as the video to be viewed if the candidate video has a smaller average value of variance of the excitement level of each scene among the plurality of viewing users.

[0071] [7] The recommendation device described in [6], wherein the determination unit derives the average value of the variance for a specific user group identified from information about the viewing users, and prioritizes the candidate video with a smaller average value of the variance for the specific user group as the viewing video.

[0072] [8] The recommendation device according to [1] or [2], wherein the determination unit determines a representative viewing user based on similarities in viewing tendencies among the plurality of viewing users according to information related to the excitement in each scene, and determines the viewing video based on the viewing tendencies of the representative viewing user.

[0073] [9] The recommendation device according to claim 8, wherein, when the plurality of viewing users are divided into a plurality of groups based on the similarity of the viewing tendencies, the determination unit determines the representative viewing user from among the viewing users in a group including the largest number of viewing users.

[0074]

[10] A recommendation method performed by a recommendation device, comprising: receiving, from a plurality of users, video information relating to a video including a plurality of scenes in chronological order, the video information including information about the users who have viewed the video and information about the excitement of each scene; determining a video to be viewed from a plurality of candidate videos based on the video information about a plurality of viewing users who view the video together; and recommending the determined video to be viewed.

[0075] The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and / or software. Furthermore, the method for realizing each functional block is not particularly limited. That is, each functional block may be realized using a single device that is physically or logically coupled, or may be realized using two or more physically or logically separated devices that are connected directly or indirectly (e.g., via wire, wirelessly, etc.) and these multiple devices. The functional block may also be realized by combining the single device or multiple devices with software.

[0076] Functions include, but are not limited to, judgment, determination, assessment, calculation, computation, processing, derivation, investigation, search, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, consideration, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating, mapping, and assignment. For example, a functional block (component) that performs transmission is called a transmitting unit or transmitter. As mentioned above, there are no particular limitations on how these functions are implemented.

[0077] For example, a recommendation device 20 constituting a recommendation system according to an embodiment of the present disclosure may function as a computer that performs processing of the control method of the present disclosure. FIG. 8 is a diagram illustrating an example of the hardware configuration of the recommendation device 20 according to this embodiment. The recommendation device 20 described above may be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like. Note that the recommendation device 20 may be configured as a computer device including at least one processor such as a CPU or GPU, or may be configured as a computer device including multiple processors, or may be configured to include multiple computer devices. The terminal 10 may also have a similar hardware configuration.

[0078] In the following description, the term "device" can be interpreted as a circuit, a device, a unit, etc. The hardware configuration of the recommendation device 20 may be configured to include one or more of the devices shown in the figure, or may be configured to exclude some of the devices.

[0079] Each function of the recommendation device 20 is realized by loading specific software (programs) onto hardware such as the processor 1001 and memory 1002, causing the processor 1001 to perform calculations, control communication via the communication device 1004, and control at least one of reading and writing data in the memory 1002 and storage 1003.

[0080] The processor 1001 controls the entire computer by running, for example, an operating system. The processor 1001 may be configured by a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. For example, the above-mentioned reception unit 21, model creation unit 22, determination unit 23, recommendation unit 24, etc. may be realized by the processor 1001.

[0081] The processor 1001 also reads programs (program code), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002 and executes various processes in accordance with these. The program used may be a program that causes a computer to execute at least some of the operations described in the above-described embodiments. For example, the reception unit 21, the model creation unit 22, the determination unit 23, and the recommendation unit 24 may be implemented by a control program stored in the memory 1002 and running on the processor 1001, and similar implementations may be used for other functional blocks. While the above-described various processes have been described as being executed by one processor 1001, they may also be executed simultaneously or sequentially by two or more processors 1001. The processor 1001 may be implemented by one or more chips. The programs may also be transmitted from a network via a telecommunications line.

[0082] The memory 1002 is a computer-readable recording medium and may be configured, for example, by at least one of a read-only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a random access memory (RAM), etc. The memory 1002 may also be called a register, a cache, a main memory (primary storage device), etc. The memory 1002 can store executable programs (program codes), software modules, etc. for implementing a control method according to an embodiment of the present disclosure.

[0083] Storage 1003 is a computer-readable recording medium, and may be composed of at least one of, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disk, a digital versatile disk, a Blu-ray (registered trademark) disk), a smart card, a flash memory (e.g., a card, a stick, a key drive), a floppy (registered trademark) disk, a magnetic strip, etc. Storage 1003 may also be referred to as an auxiliary storage device. The above-mentioned storage medium may be, for example, a database, a server, or other appropriate medium including at least one of memory 1002 and storage 1003.

[0084] The communication device 1004 is hardware (transmission / reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also referred to as, for example, a network device, a network controller, a network card, a communication module, etc. The communication device 1004 may be configured to include a high-frequency switch, a duplexer, a filter, a frequency synthesizer, etc. to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). For example, the above-mentioned reception unit 21, recommendation unit 24, etc. may be realized by the communication device 1004.

[0085] The input device 1005 is an input device (e.g., a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that accepts input from the outside. The output device 1006 is an output device (e.g., a display, a speaker, an LED lamp, etc.) that outputs to the outside. Note that the input device 1005 and the output device 1006 may be integrated into one device (e.g., a touch panel).

[0086] Furthermore, each device, such as the processor 1001 and the memory 1002, is connected by a bus 1007 for communicating information. The bus 1007 may be configured using a single bus, or may be configured using different buses between each device.

[0087] The recommendation device 20 may also be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, the processor 1001 may be implemented using at least one of these pieces of hardware.

[0088] The notification of information is not limited to the aspects / embodiments described in the present disclosure and may be performed using other methods. For example, the notification of information may be performed by physical layer signaling (e.g., Downlink Control Information (DCI) and Uplink Control Information (UCI)), higher layer signaling (e.g., Radio Resource Control (RRC) signaling, Medium Access Control (MAC) signaling, broadcast information (Master Information Block (MIB) and System Information Block (SIB))), other signals, or a combination thereof. Furthermore, the RRC signaling may be referred to as an RRC message, and may be, for example, an RRC Connection Setup message, an RRC Connection Reconfiguration message, or the like.

[0089] The order of the procedures, sequences, flowcharts, etc. of each aspect / embodiment described in this disclosure may be changed unless it is consistent. For example, the methods described in this disclosure present elements of various steps using an example order, and are not limited to the particular order presented.

[0090] Input and output information may be stored in a specific location (for example, memory) or may be managed using a management table. Input and output information may be overwritten, updated, or added to. Output information may be deleted. Input information may be sent to another device.

[0091] The determination may be made based on a value represented by one bit (0 or 1), a Boolean value (true or false), or a numerical comparison (e.g., comparison with a predetermined value).

[0092] The aspects / embodiments described in this disclosure may be used alone, in combination, or switched depending on the implementation. Notification of predetermined information (e.g., notification that "X is true") is not limited to explicit notification, but may be implicit (e.g., not notifying the predetermined information).

[0093] Although the present disclosure has been described in detail above, it is clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the spirit and scope of the present disclosure as defined by the claims. Therefore, the description of the present disclosure is intended to be illustrative and does not have any limiting meaning on the present disclosure.

[0094] Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

[0095] Software, instructions, information, etc. may also be transmitted or received over a transmission medium. For example, if software is transmitted from a website, server, or other remote source using wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)), and / or wireless technologies (such as infrared, microwave), then these wired and / or wireless technologies are included within the definition of transmission media.

[0096] The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0097] Note that terms described in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings. For example, at least one of a channel and a symbol may be a signal (signaling). Furthermore, a signal may be a message. Furthermore, a component carrier (CC) may be called a carrier frequency, a cell, a frequency carrier, etc.

[0098] Furthermore, the information, parameters, etc. described in the present disclosure may be expressed using absolute values, relative values ​​from a predetermined value, or other corresponding information. For example, a radio resource may be indicated by an index.

[0099] The names used for the above-described parameters are not intended to be limiting in any way. Furthermore, the mathematical expressions using these parameters may differ from those explicitly disclosed in this disclosure. The various channels (e.g., PUCCH, PDCCH, etc.) and information elements may be identified by any suitable names, and therefore the various names assigned to these various channels and information elements are not intended to be limiting in any way.

[0100] In this disclosure, the terms "Mobile Station (MS)," "user terminal," "User Equipment (UE)," "terminal," and the like may be used interchangeably.

[0101] A mobile station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or some other suitable terminology.

[0102] As used in this disclosure, the terms "determining" and "determining" may encompass a wide variety of actions. "Determining" and "determining" may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, inquiring (e.g., searching in a table, database, or other data structure), ascertaining, and the like. "Determining" and "determining" may also include receiving (e.g., receiving information), transmitting (e.g., sending information), input, output, accessing (e.g., accessing data in memory), and the like. Furthermore, "judgment" and "decision" can include regarding resolving, selecting, choosing, establishing, comparing, etc. as having been "judged" or "decided." In other words, "judgment" and "decision" can include regarding some action as having been "judged" or "decided." Furthermore, "judgment (decision)" can be interpreted as "assuming," "expecting," "considering," etc.

[0103] The terms "connected," "coupled," or any variation thereof, refer to any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are "connected" or "coupled" to each other. The coupling or connection between elements may be physical, logical, or a combination thereof. For example, "connected" may be read as "access." As used in this disclosure, two elements may be considered to be "connected" or "coupled" to each other using one or more wires, cables, and / or printed electrical connections, as well as electromagnetic energy having wavelengths in the radio frequency range, microwave range, and optical (both visible and invisible) range, as some non-limiting and non-exhaustive examples.

[0104] As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

[0105] As used in this disclosure, any reference to an element using a designation such as "first," "second," etc. does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient method of distinguishing between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must in some way precede the second element.

[0106] When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Furthermore, when the term "or" is used in this disclosure, it is not intended to be an exclusive or.

[0107] In this disclosure, where articles are added by translation, such as a, an, and the in English, the disclosure may include that the nouns following these articles are in the plural form.

[0108] In the present disclosure, the term "A and B are different" may mean "A and B are different from each other." The term may also mean "A and B are each different from C." Terms such as "separate" and "coupled" may also be interpreted in the same way as "different."

[0109] 20... Recommendation device, 21... Reception unit, 22... Model creation unit, 23... Determination unit, 24... Recommendation unit.

Claims

1. A recommendation device comprising: a reception unit that receives video information from multiple users, the video including multiple scenes in chronological order, the video including information about the users who watched the video and information about the excitement in each scene; a determination unit that determines a video to watch from multiple candidate videos based on the video information about multiple viewing users who watch the video together; and a recommendation unit that recommends the video to watch determined by the determination unit.

2. The recommendation device according to claim 1, wherein the determination unit determines the viewed video as a video that is viewed simultaneously by the plurality of viewing users.

3. A recommendation device as described in claim 1, further comprising a model creation unit that creates a machine learning model that predicts information regarding the excitement of each scene included in a video, wherein the reception unit receives the video information further including information regarding each scene, the model creation unit performs machine learning on a plurality of pieces of video information using information regarding the user and information regarding each scene as explanatory variables and information regarding the excitement of each scene as a target variable, to create the machine learning model, and the determination unit inputs information regarding the viewing user and information regarding each scene of the candidate video into the machine learning model to predict information regarding the excitement of each scene for each combination of the viewing user and the candidate video, and determines the video to be viewed based on the prediction result.

4. The recommendation device according to claim 3, wherein the determination unit prioritizes the candidate video to be viewed if the candidate video has a higher degree of coincidence of scenes that are highly exciting among the plurality of viewing users.

5. The recommendation device according to claim 3, wherein the determination unit gives priority to determining a candidate video as the viewed video for which the degree of excitement varies less among the plurality of viewing users.

6. The recommendation device according to claim 5, wherein the determination unit gives priority to determining a candidate video to be viewed when the candidate video has a smaller average value of the variance of the excitement level of each scene among the plurality of viewing users.

7. A recommendation device as described in claim 6, wherein the determination unit derives the average value of the variance for a specific user group identified from information about the viewing users, and prioritizes the candidate video with the smaller average value of the variance for the specific user group as the viewed video.

8. A recommendation device as described in claim 1, wherein the determination unit determines a representative viewing user based on the similarity in viewing tendencies among the plurality of viewing users according to information regarding the excitement in each scene, and determines the viewing video based on the viewing tendencies of the representative viewing user.

9. A recommendation device as described in claim 8, wherein the determination unit determines the representative viewing user from among the viewing users in the group containing the largest number of viewing users when the plurality of viewing users are divided into a plurality of groups based on the similarity of the viewing tendencies.

10. A recommendation method performed by a recommendation device, comprising: receiving video information from multiple users, the video including multiple scenes in chronological order, the video including information about the users who watched the video and information about the excitement in each scene; determining a video to watch from multiple candidate videos based on the video information about multiple viewing users who watch the video together; and recommending the determined video to watch.