JP2021068090A - Content recommendation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To recommend a content suitable for a user's taste from all the contents which can be viewed by the user regardless of whether the user's preference for the content is unknown or known. A content recommendation system according to the present disclosure calculates a preference degree for a plurality of contents presented by a content presenting unit 103 capable of reproducing at least sound or video content, and a user. Content to be reproduced by the content presentation unit 103 from the information stored in the content information storage unit 101, the content information storage unit 101 that stores the content in association with the content, and the content information storage unit 101 that determines the preference. It has a content extraction unit 102 for extracting the content, and a content presentation unit 103 for presenting the content extracted by the content extraction unit 102 to the user, and recommends the content extracted by the content extraction unit 102. [Selection diagram] Fig. 1

Description

The present disclosure relates to a content recommender system used for content playback devices such as music and video.

Conventionally, in order to recommend content according to the user's taste, there is a technique of recommending content based on an estimated user's emotion. For example, in Patent Document 1, emotion information is given to the content in advance, the user's emotion is estimated from the user's content viewing time, and the content having the same emotion information as the estimated user's emotion is recommended next.

In addition, the user's past search history and search queries (for example, words and phrases) are used to calculate the degree of suitability with the user's taste for each content category, and the search results are presented based on the results. There is technology. For example, in Patent Document 2, the user feature vector is calculated for each category based on the goodness of fit of the search query and the category, the click operation history of the user, and the past user feature vector, and the user feature vector is calculated according to the obtained user feature vector. To change the priority of search results.

JP-A-2015-228142 Japanese Unexamined Patent Publication No. 2016-11260

However, in the technology of Patent Document 1, since the viewing time of the content and the emotion of the user are linked one-to-one, it is possible to recommend the content close to the emotion of the user, but the content according to the taste of the user can be recommended. There are cases where recommendation is difficult. For example, when the user's emotion is estimated to be "sad" from the viewing time, content having the same "sad" emotion information as the user's emotion is recommended. However, even if the user's emotions are "sad", there is a problem that the optimum content cannot be recommended to such users because they may desire "fun" content to change their mood.

Further, the technique of Patent Document 2 is a technique of obtaining a user feature vector of a category to be recommended and recommending the category having the highest vector evaluation value, but in the case of a user who does not have a viewing history or a click operation history, a search is performed. Since the user feature vector is obtained only from the degree of correspondence between the query and the category, it becomes difficult to recommend the category according to the user's preference. For example, if a user wants to listen to a "calm" song sung by an "idol", a user simply searching for "idol" will recommend "fun" content that is generally assumed to be popular. , "Idol calm song ballad" etc. need to be refined in the search query. However, this means eventually requires the user to think about his / her own taste and input it, which imposes a heavy burden on the user.

This disclosure solves the conventional problems, and includes content whose degree of conformity with the user's taste is unknown, and content recommendation that can recommend content that suits the user's taste from among the content that can be viewed by the user. The purpose is to provide a system.

In order to solve the above-mentioned conventional problems, the content recommendation system according to the present disclosure calculates at least a content presentation unit capable of reproducing sound or video content and a degree of preference for a plurality of contents presented by the content presentation unit. Then, from the preference judgment unit that determines the user's preference, the content information storage unit that stores the main body data of the content and the emotional information of the content in association with each other, and the information stored in the content information storage unit, the content presentation unit It has a content extraction unit that extracts the content to be played, and recommends the content.

According to the content recommendation system according to the present disclosure, by obtaining the degree of preference for a plurality of contents, it is possible to determine the direction in which the user is interested, and to improve the accuracy of content recommendation that suits the user's preference. Can be done.

It is a figure which shows the structure in Embodiment 1 of this disclosure. It is a flow chart which shows the process in a preference determination part. It is explanatory drawing which shows the information stored in the content information storage part. It is a flow chart which shows the process in a content extraction part. It is a figure which shows the state of the preference determination in the preference determination unit when the mathematical expression (1) is used. It is a figure which shows the structure in Embodiment 1 including the operation information acquisition part. It is a flow chart which shows the calculation procedure of the goodness-of-fit degree using operation information acquired by the operation information acquisition part. It is a figure which shows the structure in Embodiment 1 including a sensor information acquisition part. It is a flow chart which shows the calculation procedure of the goodness-of-fit degree using the sensor information acquired by the sensor information acquisition part. It is a figure which shows the structure in Embodiment 1 including a scene determination part. It is a flow chart which shows the operation procedure of content recommendation using a scene judgment part. It is a figure which shows the structure in Embodiment 1 including the user information storage part. It is a flow figure which shows the operation procedure of content recommendation using a user information storage part. It is a flow chart which shows the operation procedure of the content recommendation using the preference judgment result change part. It is a figure which shows the structure in Embodiment 2 of this disclosure. It is a figure which shows the example of the analysis result in the vehicle interior sound environment analysis department. It is a figure which shows the structure in Embodiment 2 including this driver monitor camera, a speaker, and a control unit. It is a figure which shows the mechanism of operation in Embodiment 2 including this driver monitor camera, a speaker, and a control unit. It is a figure which shows the structure in Embodiment 2 which can display a tire performance. It is a figure which shows the example of the analysis result in the vehicle interior sound environment analysis department. It is a figure which shows the display example of a tire performance.

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

It should be noted that all of the embodiments described below show comprehensive or specific examples. The numerical values, shapes, materials, components, arrangement positions and connection forms of the components, steps, the order of steps, etc. shown in the following embodiments are examples, and are not intended to limit the present disclosure. Further, among the components in the following embodiments, the components not described in the independent claims will be described as arbitrary components.

Further, each figure is a schematic view and is not necessarily exactly illustrated. Therefore, for example, the scales and the like do not always match in each figure. Further, in each figure, substantially the same configuration is designated by the same reference numerals, and duplicate description may be omitted or simplified.

(Embodiment 1)
FIG. 1 is a block diagram showing a functional configuration of a content recommendation system according to the first embodiment of the present disclosure. The content recommendation system according to the present embodiment has a preference determination unit 100 that calculates the degree of suitability (degree of preference) with the user's preference for the content and determines the information of the content that matches the user's preference, and the content information and the content. The content information storage unit 101 that stores the emotional information in association with each other, the content extraction unit 102 that extracts the content from the content information storage unit 101 based on the determination result of the preference determination unit 100, and the content extraction unit 102. It includes a content presenting unit 103 that presents the content to the user.

For example, the preference determination unit 100 and the content extraction unit 102 use a CPU (Central Processing Unit), the content information storage unit 101 uses a storage medium such as an HDD (Hard Disk Drive), and the content presentation unit 103 uses an output device such as a display or a speaker. Can be configured.

Here, the emotional information is information about emotions that a person feels toward the content when viewing the content. The emotional information may be text data such as "fun" or "sad", or may be numerical data. Here, the format of emotional information is not limited.

The content recommended by the present embodiment refers to all media contents such as music, video, and electronic books, and the present embodiment is applied to a system or device for a user to view these media contents.

For example, when the present embodiment is applied to a music player, the preference determination unit 100 calculates the preference degree for a plurality of played songs, and the emotion of the next recommended song based on the plurality of preference degree. Determine the information. The content information storage unit 101 stores in advance the audio file of the music and the emotional information felt by a person when listening to the music in association with each other. The content extraction unit 102 selects a song from the content information storage unit 101 based on the determination result of the preference determination unit 100. The content presentation unit 103 transmits the selected music to a speaker, headphones, or the like, and reproduces the music.

Further, when the present embodiment is applied to a video player, the preference determination unit 100 calculates the preference degree for the reproduced plurality of video works, and the next recommended video work based on the plurality of preference suitability. Determine emotional information. The content information storage unit 101 stores the video data of the video work and the emotional information held by the person when the video is viewed in association with each other. The content extraction unit 102 selects a video work from the content information storage unit 101 based on the determination result of the preference determination unit 100. The content presentation unit 103 transmits a video work to a display, a projector, or the like, and reproduces the video work.

Further, when the present embodiment is applied to an electronic book reader, the preference determination unit 100 calculates the preference degree for a plurality of selected books, and the emotion of the book recommended next based on the plurality of preference degree. Determine the information. The content information storage unit 101 stores the text data of the book and the emotional information held by the person when the book is read in association with each other. The content extraction unit 102 selects a book from the content information storage unit 101 based on the determination result of the preference determination unit 100. The content presentation unit 103 transmits book data to a display, a speaker, or the like, and displays the book and reproduces the sound.

A music player, a video player, and an electronic book reader have been given as application examples of this embodiment, but the scope of application of this embodiment is not limited here, and may be applied to other media players. Absent.

Further, the content recommended by the present embodiment is not limited to one category (for example, music, video, etc.), and may include a plurality of categories. For example, a combination of video and an electronic book may be recommended.

FIG. 2 is a flow chart showing an example of processing in the preference determination unit 100. The preference determination unit 100 may include, for example, a preference degree calculation unit, a preference determination result storage unit, and a search request transmission unit.

As shown in FIG. 2, the preference degree calculation unit calculates the degree of suitability of the content with the user's preference when the user is viewing the content (S21). The calculation method of the goodness of fit will be described below.

The preference determination result storage unit stores the preference degree calculated by the preference degree calculation unit in association with the content (S22). When calculating the degree of suitability with the user's preference for a plurality of contents, the preference degree calculation unit may calculate the preference degree as numerical data such as 1 to 100, and the first place to the first place. It may be calculated as ranking data such as 3rd place. In the present embodiment, the preference degree is handled in the range of 0 to 1, and it is determined that the higher the value, the more suitable the user's preference, but the range of the preference degree is not limited. The information stored in the preference determination result storage unit may be initialized or kept every time the system to which the present embodiment is applied is started.

If the number of preference degrees required for content recommendation is stored in the preference judgment result storage unit (S23; YES), the search request transmission unit is based on the plurality of preference degrees stored in the preference judgment result storage unit. Then, a search request for the content to be recommended next is created (S24).

If the preference determination result storage unit does not store the number of preference degrees required for content recommendation (S23; NO), a search request is created so that the content to be recommended next is arbitrary content (S25).

The search request transmission unit creates a search request for the content to be recommended next based on the preference degree stored in the preference determination result storage unit, and transmits the search request to the content extraction unit 102. For example, when considering the emotional information of the content in terms of the coordinate values in the two-dimensional space, the search request may be the coordinate value of the content to be recommended next, or may be a range surrounded by a plurality of points. Here, the format of the data to be transmitted as a search request is not limited.

The configuration of the preference determination unit 100 is not limited to the above. For example, the preference degree is not calculated by the preference degree calculation unit, but instead has a preference degree input unit to allow the user to input the preference degree. You may use the display of. Of course, the preference degree input unit may use an output device such as a display used in the content presentation unit 103, or may be provided separately from the content presentation unit 103.

The content information storage unit 101 stores the original data of the content and the emotional information felt by a person by viewing the content in a database in association with each other. For example, as shown in FIG. 3, emotion information may be stored as coordinate values in a two-dimensional space with the horizontal axis as the comfort axis and the vertical axis as the arousal degree axis, and "clean feeling" and "power". , And may be stored as coordinate values of "brightness" in the three-dimensional space. This makes it possible to quantify and evaluate emotions such as "fun" and "relaxation", and also consider the intensity of each emotion. For example, when this embodiment is applied to a music player, it is possible to recommend content that gradually makes the song more enjoyable, or content that changes from anxious content to relaxing content. It becomes.

The content information storage unit 101 may store the content information in a storage medium such as an HDD or SSD (Solid State Drive) inside the system to which the present embodiment is applied in advance, or is outside the system to which the present embodiment is applied. It may be stored in the attached HDD or USB (Universal System Bus) memory, or may be downloaded from an external server via the network. The means for storing content information is not limited here.

When the content information storage unit 101 handles emotional information of contents as numerical data, it is preferable to limit the range of numerical values. For example, when handling emotional information in a two-dimensional space, the lower limit of the coordinate values on the horizontal axis and the coordinate values on the vertical axis may be set in a rectangular range with -1 and the upper limit being 1, or a radius centered on the origin. It may be set in a circular range of 1. As a result, it is possible to prevent the content from having coordinate data extremely far from the center in the coordinate space, so that the content can be evenly recommended to the user.

As the method of acquiring the emotional information of the content to be stored in the content information storage unit 101, the incidental information of the content may be used. In this case, the content information storage unit 101 may extract emotional information from the incidental information of the content. For example, regarding music content, information that means the genre of the content such as "rock", "ballad", and "classic", the name of the author of the music, the name of the performer of the music, the time when it was created, the country where it was announced, etc. You may acquire emotional information based on the information of. For video content and e-books, emotional information is based on keywords that mean the content such as "SF", "love story", and "spocon", as well as information such as the director, author, time of creation, and actors. You may get. Here, the incidental information required to acquire the emotional information of the content is not limited.

Further, the emotional information of the content may be acquired based on the feature amount obtained by analyzing the content instead of the incidental information of the content. For example, for music content, "tempo", which is an index of the speed felt by the listener, "melody", which is the main melody of the music, and "chord progression", which is information on changes in the chords that make up the music, etc. Emotional information may be acquired based on the acoustic feature amount. In addition, a model for predicting the emotions that a person has when viewing the content may be constructed using the acoustic features and the subjective evaluation data by the person. For example, a regression model may be created using acoustic features as learning data and subjective evaluation data by humans as teacher data. The means for acquiring emotional information of the content to be stored in the content information storage unit 101 is not limited here.

The content extraction unit 102 extracts the content from the content information storage unit 101 based on the determination result received from the preference determination unit 100. The content extraction unit 102 may be composed of, for example, a search request receiving unit, a content search unit, and a content transmitting unit. FIG. 4 is a flow chart showing an example of processing in the content extraction unit 102.

The search request receiving unit receives the emotional information (request) of the recommended content transmitted from the search request transmitting unit (S41).

The content search unit accesses the content information storage unit 101 based on the request received by the search request receiving unit, and searches for the content (S42). For example, when emotion information is stored as coordinates in a two-dimensional space, the flow of content search when coordinate data (X, Y) is received as a request will be described. The content search unit may refer to the coordinate data of the content stored in the content information storage unit 101 and extract the content closest to (X, Y). Further, when a plurality of coordinate data are received as a request, arbitrary contents within the range composed of those coordinates may be extracted. The configuration of the content extraction unit 102 is not limited to the above.

The content transmission unit transmits the content searched by the content search unit to the content presentation unit 103 (S43).

In the present embodiment, since the user's preference is determined by the preference degree for a plurality of contents, it is possible to recommend the content according to the user's preference as compared with the method of recommending the content from the preference preference degree for one content. Become.

In a conventional system that recommends content based on the degree of preference of one content, for example, when the degree of preference is calculated to be high for content having "sad" emotion information, the user is similarly provided with "sad" emotion information. You can recommend the content you have. However, when the degree of preference is calculated low, content with emotional information different from "sad" should be recommended, but whether "fun" content should be recommended or "calm" content should be recommended here is correct. It is difficult to judge.

On the other hand, in the present embodiment, for example, "sad" content has a low degree of preference, "calm" content has a low degree of preference, and "fun" content has a slightly high degree of preference, and should be recommended next. As for the content, it is possible to recommend "intense" content with stronger "fun" emotional information. In this way, by calculating the goodness of fit for a plurality of contents, it is possible to determine which emotional information the user's preference vector is suitable for, and it is possible to recommend the content according to the user's preference.

In addition, by determining which emotional information the user's preference vector is suitable for, it is possible to recommend content that the user has not listened to and whose preference degree is unknown. The method of calculating the preference of a plurality of contents and recommending the one with the highest preference can be recommended only from the contents for which the preference is known, so the recommended contents are limited. Will be done.

On the other hand, in the present embodiment, by determining which emotional information the user's preference includes from the content whose preference degree is known, it is possible to recommend even the content whose preference preference degree is unknown.

Since the basic configuration in this embodiment has been described in general, the flow of content recommendation will be described next. Here, a case where the next content is recommended based on the degree of preference for the three contents is considered. The goodness of fit is treated as a coordinate value in a two-dimensional space. When treating the user's preference as a vector, only the content on the straight line connecting the coordinates of the two contents can be recommended only by the two preference degrees in the two-dimensional space, but if the three preference degrees are known, the two-dimensional space Since all of the above contents can be recommended, it is desirable to first calculate the degree of preference for at least three contents at the time of initial content recommendation to determine the user's preference. Here, three contents are arbitrarily extracted and presented. The three presented contents are referred to as content A, content B, and content C. As a result of calculating these three taste preference degrees by the preference determination unit 100, it is assumed that the content A has the highest preference degree, the content B has the next highest preference degree, and the content C has the lowest preference degree. In this case, the content to be recommended next is determined to be the content D having emotion information different from the content C and having the emotion information similar to the contents A and B.

Here, an example of recommending the next content based on the degree of preference for the three contents is shown, but the number of contents used for determining the preference may be any number as long as it is two or more. Further, the number of contents used for preference determination is not constant and may be changed. For example, when the present embodiment is applied to a music player, the number of songs used for determining the user's preference may be increased by one each time the user listens to one song, and the content of the three songs viewed immediately before is always used. The user's preference may be judged from the preference degree and the content may be recommended.

とし、コンテンツＢの位置ベクトル（座標）を

とし、コンテンツＣの位置ベクトル（座標）を

として考えることができる。例えば、嗜好適合度がコンテンツＣ＜コンテンツＢ＜コンテンツＡの順で高くなっていると算出された場合、次に推薦するコンテンツＤの位置ベクトルである

は次の数式（１）のように算出しても良い。また、図５は数式（１）を用いた場合の嗜好判断の過程を示したものである。 When the content information storage unit 101 stores the emotional information of the content as coordinate values in the two-dimensional space, the position vector (coordinates) of the content A is stored.

And the position vector (coordinates) of the content B

And the position vector (coordinates) of the content C

Can be thought of as. For example, when it is calculated that the degree of preference is higher in the order of content C <content B <content A, it is the position vector of the content D to be recommended next.

May be calculated as in the following mathematical formula (1). Further, FIG. 5 shows the process of preference determination when the mathematical formula (1) is used.

ｍ、ｎ、α：任意の係数

m, n, α: Arbitrary coefficient

The value of α in the formula (1) may be changed at any time according to the calculated value of the goodness of fit. For example, the formula (2) may be used such that if the maximum value of the calculated goodness of fit is low, the value of α is increased, and if the maximum value is high, the value of α is decreased.

For example, when β in the formula (2) is 1, and the maximum preference degree is 0.8, α is 0.2, and the content having coordinates close to the content having the highest preference degree can be recommended next. .. On the other hand, when β is set to 1 and the maximum preference degree is set to 0.3, α is 0.7, and the content whose coordinates are different from the reproduced content can be recommended next. That is, it is possible to next recommend content having different emotional information from the content having a low degree of preference, and next to recommend content having similar emotion information to the content having a high degree of preference.

As a result, it is possible to not recommend the content that does not match the user's taste, and to recommend the content that matches the user's taste next. Further, even if all of the recommended plurality of contents do not match the user's taste, it is possible to recommend the content according to the user's taste without continuing to recommend the same content. Similarly, if all of the recommended contents match the user's taste, the next recommended content can also recommend the same content.

The content recommendation method is not limited to the mathematical formula (1) and the mathematical formula (2), and the content of the mathematical formula may be changed according to the result of the degree of preference and the situation of the user. You may use multiple formulas properly.

The preference determination unit 100 may weight the degree of suitability of the determined content with the user's preference according to an arbitrary criterion. For example, when weighting is performed according to the order in which the contents are played, it is possible to recommend the contents corresponding to the change in the user's tastes over time by highly evaluating the preference degree of the contents played immediately before. It becomes. At this time, the degree of weighting may always be constant, or may be changed each time the content is recommended. The weighting criteria are not limited to the above.

When the user selects the content by himself / herself, weighting may be performed so as to highly evaluate the preference degree of the selected content. This makes it possible to extract content having characteristics similar to the content directly selected by the user and present it to the user at the time of content recommendation from the next time onward. For example, when this is applied to a music player, the user is first presented with the recommended music based on the determination result of the preference determination unit 100, but when the user himself searches for the music and selects the music. By making the preference degree of the music selected by the user higher than the preference preference degree of the music that has been played up to the previous time, it becomes possible to recommend the music that reflects the user's preference next.

The content recommendation system according to the present disclosure has an operation information acquisition unit 204 for acquiring operation information related to content playback operation control by the user in the preference determination unit 200, determines the preference based on the acquired operation information, and determines the content. The extraction unit 202 may extract the content from the content information storage unit 201 and present the extracted content to the content presentation unit 203. Playback operation control refers to, for example, fast-forwarding, rewinding, and volume adjustment in the case of a music player. Also, if it is an e-book reader, it means turning pages, sandwiching bookmarks, and so on.

The operation information acquisition unit 204 may cooperate with an application such as a music player or an electronic book reader to which the present embodiment is applied to receive operation information from the cooperation destination application, or may be an application target of the present embodiment. If it is difficult to obtain the operation information from the application, the operation information estimated by a camera or microphone not connected to the application may be obtained. Here, the means for acquiring operation information is not limited.

Hereinafter, a content recommendation method using the data acquired by the operation information acquisition unit 204 will be described. FIG. 6 shows the configuration of the embodiment having the operation information acquisition unit 204, and FIG. 7 is a flow chart showing an example of the calculation procedure of the preference degree using the operation information.

As shown in FIG. 7, first, the operation information acquisition unit 204 confirms whether the operation information can be acquired from the application target of the embodiment, and acquires the acquireable operation information (S61). In addition, the number of acquired operation information is also stored (S62).

Next, the preference determination unit 200 calculates the preference degree based on the operation information acquired by the operation information acquisition unit 204 (S63). When there are a plurality of acquired operation information (S64; NO), the preference degree is calculated based on all the operation information (S65).

After completing the calculation of the preference degree based on the operation information (S64; YES), the overall preference degree is determined from the plurality of preference degree (S66). The overall preference degree may be, for example, the average of a plurality of calculated preference degrees, or may be the highest preference degree. The calculation method is not limited here.

The operation information acquisition unit 204 may measure the time from the playback of the content to the fast-forwarding, and use the obtained value for the calculation of the goodness of fit. For example, if the time until the fast-forward is performed is short, the preference degree may be evaluated low, and if the time until the fast-forward is performed is long, the preference degree may be evaluated high.

The formula for measuring the time from the playback of the content to the fast-forwarding and calculating the preference degree using the obtained value may be the following formula (3).

When calculating the preference degree using the mathematical formula (3), for example, assuming that the total playback time of the content for which the preference degree is calculated is 240 seconds and the playback time when the user fast-forwards is 30 seconds, the preference degree is calculated. Is 0.125. The number of significant digits of the calculated goodness of fit may be arbitrarily determined in the present embodiment.

As a result, when the present embodiment is applied to a music player, it can be determined that the played music does not meet the user's taste when the played music is fast-forwarded in the middle, and the played music is not fast-forwarded to the end. It is possible to determine that the user's taste is suitable.

The operation information acquisition unit 204 may measure the number of times the content is rewound and use the obtained value for calculating the goodness of fit. For example, if the number of times of rewinding is large, the degree of preference may be evaluated high, and if the number of times of rewinding is small, the degree of preference may be evaluated low.

The formula for measuring the number of times the content is rewound and calculating the degree of preference using the obtained value may be the following formula (4). In formula (4), by setting the upper limit of the number of times the content is rewound to 5, the number of times the content is rewound is counted up to 5 times, and 5 is substituted after the 5th time, but the content is rewound. The upper limit of the number of times may be arbitrarily set as long as it is a value of 2 or more.

ｃｎｔＭａｘ：コンテンツ巻き戻し回数の上限
ｃｎｔ：上限を考慮したコンテンツ巻き戻し回数

ctMax: Maximum number of content rewinds ct: Number of content rewinds considering the upper limit

When calculating the preference degree using the mathematical formula (4), for example, when the number of times of rewinding the content for which the preference degree is calculated is 2 and the upper limit of the number of rewinds is 5, the preference degree is It becomes 0.4. If the product is never rewound, the goodness of fit is 0.

As a result, when the present embodiment is applied to a music player, it is determined that the music is suitable for the user's taste when it is rewound to the beginning and viewed again after the music is played to the end, and the degree of preference is calculated to be high. can do. Further, when the user tries to watch the next song without rewinding, it is determined that the song does not suit the user's taste, and the degree of preference suitability can be calculated low.

The operation information acquisition unit 204 may measure the value of the volume changed by the volume adjustment and use the obtained value for the calculation of the goodness of fit. For example, if the volume control is very small, the preference degree may be evaluated high, and if the volume is greatly lowered, the preference degree may be evaluated low.

The formula for measuring the value of the volume control and calculating the degree of preference using the obtained value may be the following formula (5). In formula (5), the large volume adjustment threshold is set to 20% of the original volume, the preference dynamics is calculated high if the volume is adjusted within the threshold, and the adjusted volume is adjusted if the volume is adjusted above the threshold. The degree of preference is lowered accordingly. The threshold value T in the mathematical formula (5) may be arbitrarily set as long as it is a value of 0 or more.

When calculating the goodness of fit using the mathematical formula (5), for example, when the original volume of the content for which the goodness of fit is calculated is 30, the adjusted volume is 27, and the threshold value T is 0.8. Since a is 0.9, which is equal to or higher than the threshold value T, the goodness of fit is 1. Further, when the original volume is 30, the adjusted volume is 12, and the threshold T is 0.8, a is 0.4, which is equal to or less than the threshold T. Therefore, the goodness of fit is divided by a and the threshold T. The value is 0.5.

As a result, when the present embodiment is applied to a music player, it is determined that the user's preference is met when the volume of the music being played is adjusted to a volume that is easy for the user to hear, and the preference degree can be calculated high. it can. Further, when the user lowers the volume of the music being played to half or less, it is determined that the music does not suit the user's taste, and the degree of preference suitability can be calculated low.

The goodness of fit with the user's preference for the content may be calculated based on the plurality of information obtained from the operation information acquisition unit 204 described so far. In this case, for example, the calculation formula may be the following formula (6) to calculate the goodness of fit.

ｘ：早送りされた時の再生時間から求めた嗜好適合度
ｙ：巻き戻しの回数から求めた嗜好適合度
ｚ：音量調節の値から求めた嗜好適合度

x: Goodness of fit obtained from the playback time when fast-forwarding y: Goodness of fit obtained from the number of rewinds z: Goodness of fit obtained from the value of volume control

By using a plurality of operation information obtained from the operation information acquisition unit 204, it is possible to recommend contents in consideration of individual differences in the operation. For example, some users do not fast-forward but adjust the volume diligently, or all operations are performed, and it is considered that the operations differ depending on the user. When calculating the preference degree from one operation information, for example, for a user who does not fast-forward, if the preference is judged only by the fast-forward operation information, there is a problem that the preference degree is highly evaluated for most contents. appear. However, the above problem can be solved by calculating the goodness of fit by combining a plurality of operation information.

In addition, weighting may be performed for each goodness of fit used in the mathematical formula (6). In this case, for example, the calculation formula may be the following formula (7) to calculate the goodness of fit. α, β, and γ are coefficients for weighting and may be set to arbitrary values.

By weighting the goodness of fit obtained from the operation information, individual differences in the above operations can be further reduced. For example, when calculating the preference degree of a user who rarely fast-forwards, it is preferable to reduce the weight of the preference degree obtained from the fast-forward operation information and increase the weight of the preference degree obtained from other operation information. .. In this way, by reducing the weight of the goodness of fit calculated from the operations that are rarely performed, it is possible to determine the preference in consideration of individual differences.

The formula for obtaining the goodness of fit based on the information obtained from the operation information acquisition unit 204 is not limited to the above formula. Further, in calculating the goodness of fit, a plurality of mathematical formulas may be used properly, or the values of variables in the mathematical formulas may be changed.

By acquiring the operation information by the operation information acquisition unit 204, it is possible to recommend the content that suits the user's preference without having the user input the preference. For example, conventionally, there is a means of requesting an answer to a questionnaire form for confirming a user's preference and determining the preference based on the answer result, but such an answering work is troublesome for the user. On the other hand, in the present embodiment, since it is possible to recommend the content that suits the user's taste without forcing the user to perform tasks such as answering a questionnaire, the user can concentrate on the content.

The content recommendation system according to the present disclosure has a sensor information acquisition unit 304 for acquiring information from a sensor attached to the application destination of the present embodiment or an external sensor, and the sensor information acquired by the sensor information acquisition unit 304. The content extraction unit 302 may extract the content from the content information storage unit 301 and present the extracted content to the content presentation unit 303.

Hereinafter, the procedure of content recommendation using the data acquired by the sensor information acquisition unit 304 will be described. FIG. 8 shows a configuration when the sensor information acquisition unit 304 is provided, and FIG. 9 is a flow chart showing an example of a calculation procedure of the goodness of fit using the sensor information.

As shown in FIG. 9, first, the sensor information acquisition unit 304 confirms whether the sensor information can be acquired from the sensor attached to the application target of the embodiment or the sensor outside the embodiment, and acquires the acquireable sensor information. (S71). In addition, the number of acquired sensor information is also stored (S73).

From the acquired sensor information, the physiological index determination unit determines the user state (S72).

Next, the preference determination unit 300 calculates the preference degree based on the sensor information acquired by the sensor information acquisition unit 304 (S74). When there are a plurality of acquired sensor information (S75; NO), the preference degree is calculated based on all the sensor information (S76).

When the calculation of the preference degree based on the sensor information is completed (S75; YES), the overall preference degree is used based on the calculated sensor information and the judgment result of the physiological index judgment unit. Is determined (S77). The procedure for calculating the goodness of fit using sensor information is not limited to the above.

The sensor information acquisition unit 304 acquires the movement of the user's body detected by using the vibration sensor, transmits the obtained data to the preference determination unit 300, and the preference determination unit 300 determines the preference based on the received data. You may calculate the goodness of fit. For example, the criteria for determining preference may be changed depending on whether or not the body is shaking, or the feature amount may be extracted from the degree of shaking of the body such as the timing and tempo of the body shaking, and the criteria for determining preference may be changed based on the feature amount. It doesn't matter. As a result, when the present embodiment is applied to a music player, for example, the degree of synchronization between the shaking of the user's body and the tempo of the music being played is measured, and if they are synchronized, the degree of preference is highly evaluated and synchronized. If not, the degree of preference may be evaluated low. Further, when the present embodiment is applied to a video player, for example, if the user's body is moving in a scene that is exciting in a video work, the preference degree is highly evaluated, and if the user's body is not moving, the preference degree is evaluated low. It doesn't matter.

When the present embodiment is applied to a video player, it is desirable that the content information storage unit 301 stores the scene information of the video content in association with each other in advance. For the scene information, the video content may be divided into arbitrary time divisions, and emotional information about what kind of scene the scene is may be used for each division. For example, keywords such as "intense" and "calm" may be used, and emotional information may be quantified.

Further, the preference determination unit 300 may change the calculation method of the goodness of fit using the vibration sensor based on the emotional information of the content. For example, in the case of content that makes people feel happy, the degree of preference may be highly evaluated by shaking the body, and conversely, in the case of content that makes them relaxed, the degree of preference may be evaluated low by shaking the body. This makes it possible to determine whether or not the content suits the user's taste by confirming whether or not the emotional information originally possessed by the content acts on the user's behavior.

The sensor information acquisition unit 304 may acquire the user's voice detected by using the microphone and use the obtained value for the calculation of the goodness of fit. For example, you can get the loudness of the user's voice and estimate that it is excited if the voice is loud and relaxed if the voice is quiet. You may presume that there is. When this embodiment is applied to a video player, if the user who sees the exciting scene in the video work raises his voice from surprise or excitement, it is judged that it suits the user's taste, and if the user's reaction is weak. It is possible to determine that the user's taste is not met. The preference criteria may be changed based on the emotional information of the content.

The formula for acquiring the voice of the user who is viewing the video content and obtaining the degree of preference may be the following formula (8).

When calculating the preference degree using the mathematical formula (8), for example, the number of scene divisions of the video content to be calculated for the preference degree is 10, the emotion information estimated from the user's voice and the matching division of the emotion information in each scene. When the number is 8, the degree of preference is 0.8.

As a result, when the user utters a voice as a reaction when viewing each scene of the video, the degree of preference is calculated by calculating whether the emotional information estimated from the voice matches the emotional information of the content. It becomes possible to do.

Further, when the present embodiment is applied to a music player, the degree of matching between the user's singing voice and the melody line of the music being played is measured, and if they match, the degree of preference is highly evaluated, and if they do not match, the degree of preference is highly evaluated. The degree of preference may be evaluated low. As a result, for example, if the user is listening to music and singing a song in accordance with the music, it can be determined that the preference is met, and if the user's reaction is weak, it can be determined that the preference is not met.

In this case, it is desirable that the content information storage unit 301 stores the melody line of the music content in advance in association with the content. The melody line may be acquired by analyzing the contents, or may be acquired as metadata such as MIDI (Musical Instrument Digital Interface: a standard for exchanging performance data between electronic musical instruments).

The formula for acquiring the voice of the user who is watching the music content and obtaining the degree of preference may be the following formula (9).

When calculating the goodness of fit using the mathematical formula (9), for example, when the number of notes on the melody line of the music content for which the goodness of fit is calculated is 300, and the number of notes where the melody line and the singing voice match is 150. The goodness of fit is 0.5.

Thereby, when the user sings a song according to the music, the goodness of fit can be calculated.

The sensor information acquisition unit 304 may acquire the face of the user photographed by the camera and use the obtained feature amount for the calculation of the goodness of fit. For the feature amount, for example, the degree of opening of the eyes, the degree of opening of the mouth, the number of blinks, and the like may be used. As a result, emotions may be estimated from the user's facial expression, and content having emotional information close to the estimated emotions may be preferentially recommended. For example, if the user's facial expression is a smile, the preference degree may be evaluated high, and if the user's facial expression is steep, the preference degree may be evaluated low.

For example, when the face of the user who is viewing the video content is acquired and the goodness of fit is obtained, the mathematical formula (8) may be used in the same manner as when the goodness of fit is obtained from the voice of the user.

This makes it possible to calculate the goodness of fit from the facial expression when the user is viewing the content.

The sensor information acquisition unit 304 may acquire the user's respiration acquired by using the respiration sensor and use the obtained value for the calculation of the goodness of fit. For example, when the present embodiment is applied to a music player, the degree of synchronization between the user's breathing rhythm and the rhythm of the music is measured, and if they are synchronized, the degree of preference is highly evaluated, and if they are not synchronized, the degree of preference is highly evaluated. May be underestimated.

In this case, it is desirable that the content information storage unit 301 stores the rhythm information of the music content in advance in association with the content. The rhythm information may be acquired by analyzing the contents, or may be acquired as metadata such as MIDI.

This makes it possible to calculate the goodness of fit from the user's breathing when viewing the content.

The sensor information acquisition unit 304 has a physiological index determination unit for determining a user's physiological index from the acquired data, and the preference determination standard of the preference determination unit 300 may be changed based on the determination result. The physiological index refers to the drowsiness of the user, the concentration ratio of the user, the attention and interest of the user, and the like. The method for determining the physiological index may be, for example, estimating drowsiness from the facial expression of the user, or estimating concentration, interest, etc. from the direction of the user's line of sight.

As a result, for example, when it is determined that the user is sleepy, the criteria for preference determination may be changed so that content that awakens is recommended, and it is determined that the user's concentration is lost. If so, the criteria for preference judgment may be changed so that content that can be concentrated is recommended. In this way, it is possible to recommend content based not only on the user's preference but also on the user's physiological index.

By acquiring the sensor information in the sensor information acquisition unit, it is possible to determine the user's preference and recommend the content that suits the preference even in a situation where the operation information cannot be acquired.

Further, unlike the operation information, the information obtained by the sensor information acquisition unit includes actions that the user naturally takes with respect to the content, so that the accuracy of the user's preference determination can be further improved. For example, when watching a song, fast-forwarding the song is done by the user's intention, but it is not the user's intention that the body shakes or the facial expression is softened according to the song. It is considered that this is a natural behavior or change. In this way, it is possible to determine the preference for the content that appears in the unconscious behavior by acquiring the sensor information.

The content recommendation system according to the present disclosure has a scene determination unit 406, and may determine the user's situation based on the data obtained from the operation information acquisition unit 404 and the sensor information acquisition unit 405. The preference determination unit 400 weights the preference degree according to the scene determined by the scene determination unit 406, and the content extraction unit 402 limits the range of the content extracted from the content information storage unit 401, and the content presentation unit 403. It may be presented to.

Hereinafter, the procedure of content recommendation using the judgment result of the scene judgment unit 406 will be described. FIG. 10 is a diagram showing a configuration of an embodiment when the scene determination unit 406 is provided, and FIG. 11 is a flow diagram showing an example of an operation procedure of content recommendation.

As shown in FIG. 11, the scene determination unit 406 confirms whether the operation information can be acquired from the operation information acquisition unit 404, and confirms whether the sensor information can be acquired from the sensor information acquisition unit 405 (S81). ). If either information can be acquired (S81; YES), the scene is determined based on the acquired information (S82). The scene determination result is transmitted to the preference determination unit 400 (S84). If neither information can be obtained, the scene determination result is determined as "undecidable" (S83) and transmitted to the preference determination unit 400 (S84).

When the scene judgment unit 406 determines that the user is driving a car, for example, on a highway, the preference judgment unit 400 changes the preference judgment standard so that awakening content is preferentially recommended. However, on general roads, the preference determination unit 400 may change the criteria for preference determination so that intense content is not recommended so that the driver does not speed up too much.

In addition, when there are passengers, the criteria for determining preference may be changed according to the passengers. For example, if a mistress is on board, content that makes you feel romantic may be recommended, and if a child is on board, content that makes you feel happy may be recommended. It doesn't matter.

In addition, the criteria for determining the preference of the content may be changed according to the traveling noise generated during traveling. For example, if the running noise is noisy, a song with a violent tune may be recommended, and if the running noise is quiet, the standard may not be changed.

When the scene determination unit 406 determines that the user is taking a bath, for example, relaxing content may be prioritized and recommended. Further, the bathing time may be measured, and if the bathing time is long, the preference degree may be evaluated high, and conversely, if the bathing time is short, the preference degree may be evaluated low. The formula for measuring the bathing time and obtaining the goodness of fit may be the following formula (10). The maximum bathing time in the formula (10) may be set to an arbitrary time according to the user.

Furthermore, if the scene judgment unit 406 determines that the user is training in the gym or playing sports such as jogging, the content that motivates the user may be prioritized and recommended. .. For example, when applied to a music player, music content containing motivational lyrics such as "Do your best" or "You can do it" may be recommended. In this case, it is desirable that the content information storage unit 401 stores the music content and the lyrics in association with each other in advance.

It should be noted that although the estimated scenes include driving a car, bathing, and sports, the scenes judged by the content recommendation system according to the present disclosure are not limited here, and the preference is given according to other scenes. The determination unit 400 may change the criteria for determining preference.

In this way, by determining the scene by the scene determination unit 406 and changing the criterion of preference determination according to the determination result, it is possible to recommend not only the user's preference but also the content according to the user's scene.

The preference determination unit 400 combines the content in combination with other arousal control technologies according to either or both of the user's physiological index determined by the sensor information acquisition unit 405 and the scene determined by the scene determination unit 406. May be presented. For example, if it is determined that the user is getting sleepy in the residential space, the brightness of the indoor lighting may be reduced. In addition, if it is determined that the user is getting sleepy during work, the chair may be vibrated. It should be noted that the awakening control technology combined with the content presentation is not limited here.

This makes it possible to control the user's emotions, drowsiness, etc. by enhancing the effect of awakening or, conversely, relaxing, rather than viewing a single content.

In the present embodiment, it is possible to deal with the case where the content is recommended by one user or a plurality of users. When there are a plurality of users, the preference degree may be calculated for each user and the next content may be recommended based on the result, or a representative user may be determined and only the preference degree of the user is focused on. The following content may be recommended.

When the content is recommended based on the preference degree of a plurality of users, the preference preference degree may be obtained by the following mathematical formula (11).

As a result, it is possible to judge the preference of multiple users at the same time and recommend the content. Therefore, for example, the music to be played in a restaurant or office, the video content, etc. are recommended according to the user's preference. It becomes possible to do.

The content recommendation system according to the present embodiment may have a user information storage unit 506 for recording information about the user. In that case, the user information storage unit 506 may be composed of a user attribute information storage unit for storing the user's attribute information and a user viewing history storage unit for recording the user's content viewing history.

Hereinafter, the procedure of content recommendation using the data of the user information storage unit 506 will be described. FIG. 12 is a diagram showing a configuration of an embodiment when the user information storage unit 506 is provided, and FIG. 13 is a flow diagram showing an example of an operation procedure of content recommendation.

As shown in FIG. 13, first, the user information storage unit 506 confirms whether or not the user information exists in the user attribute information storage unit (S91). Here, if the user information cannot be confirmed (S91; NO), the user attribute information input unit instructs the user to input the user information, or either the operation information acquisition unit 504 or the sensor information acquisition unit 505. User information is estimated based on the information obtained from (S92), and the input or estimated user information is stored in the user attribute information storage unit (S93). If the user information can be confirmed (S91; YES), the content is played back as it is (S94).

After the content is reproduced, the preference determination unit 500 calculates the preference degree (S95), and the user viewing history storage unit stores the viewing history (S96). The search request unit recommends the content based on the information of the user information storage unit 506 (S97). The configuration of the user information storage unit 506 is not limited to the above, and the processing performed inside the user information storage unit 506 is not limited to the above.

The user attribute information storage unit may have a user attribute information input unit and store information for which the user has input the attribute information. As the user attribute information, a unique ID for identifying the user, age, nationality, gender, and the like may be stored. The user attribute information input unit may present a GUI (Graphical User Interface) such as a questionnaire form to the user by the content presentation unit 503, or may receive data from an external device such as a smartphone.

The user attribute information storage unit may estimate and store the user attribute information from the data acquired by the sensor information acquisition unit 505. For example, the age, nationality, gender, etc. may be estimated and stored from the face image acquired by the camera. In addition, the gender, personality, etc. may be estimated from the voice acquired by the microphone and memorized.

Each time the content is played, the user viewing history storage unit may store the ID of the user who played the content, the date and time when the content was played, and the degree of preference in association with the content. The user ID is preferably associated with the user ID of the user attribute information storage unit.

The content extraction unit 502 may refer to the user viewing history storage unit and preferentially extract unlistened content or content that has been viewed less frequently. This makes it possible to prevent similar contents from being continuously extracted, to have emotional information similar to the user's taste, and to recommend fresh contents.

The content information storage unit 501 may refer to the user attribute information storage unit and use a database of content information used for content search according to the user's attributes. For example, if the user is a man, the database for men may be referred to, and if the user is a woman, the database for women may be referred to. This makes it possible to recommend content in consideration of differences in tastes caused by differences in user attributes.

The preference determination unit 500 may have a preference determination status presentation unit and present the preference determination result of the preference determination unit 500 to the user. For example, when the content information storage unit 501 stores the emotional information of the content as the coordinate values in the two-dimensional space, the location of each content in the two-dimensional space on the display, the degree of preference for each content, and the following. You may display the location near the coordinates of the content that is likely to be recommended by. As a result, the user can confirm the judgment status of the system according to the present disclosure.

Further, after the user confirms the result of the preference determination unit 500, a preference determination result changing unit may be provided in the content recommendation system according to the present disclosure so that the previous preference determination result can be changed. The preference determination result changing unit acquires information regarding the preference determination result change (preference determination result change information), and changes the preference determination result of the preference determination unit 500. As the preference judgment result change information, a display or the like for inputting the preference judgment result change information may be provided in the present embodiment so that the user can input the preference judgment result change information, or the preference judgment result change information transmitted from an external terminal such as a smartphone may be input. You may get it.

The procedure for content recommendation using the preference judgment result changing unit will be described below. FIG. 14 is a flow chart showing a procedure for recommending content.

As shown in FIG. 14, when the content is reproduced (S101), the preference degree for the content is calculated (S102), and then the change information is input to the preference determination result changing unit (S103; YES), the information is included. Based on this, the result calculated by the preference determination unit 500 is changed (S104).

The search request is determined using the changed goodness of fit (S105), and the content is recommended (S106). After that, this is repeated.

As a result, when the user is dissatisfied or has doubts about the content recommendation, the accuracy of the content recommendation can be improved by changing the preference determination result by the user.

The present embodiment has a content processing unit and may process the content. For example, in the case of a musical piece, the tempo and pitch may be changed, and acoustic effects such as reverb (strengthening the reverberation) and equalizer (changing the strength of a specific frequency component) may be added. Further, if it is a video, the playback speed, color tone, etc. may be changed.

As a result, even if the content is the same, the emotional information of the content can be changed by processing the content, and even if the content that can be viewed is small, it is possible to recommend the content that suits the user's taste. Become.

Further, for music, the sound quality of the sound environment such as running noise may be changed by active noise control (ANC: active noise control) to recommend the content. As a result, the noise can be canceled, the influence of the noise on the emotional information of the content can be eliminated, and the accuracy of content recommendation according to the user's taste can be expected to be improved.

(Embodiment 2)
The content recommendation system according to the second embodiment will be described with reference to the schematic diagram of FIG. In the present embodiment, an example in which the content recommendation system according to the present disclosure is used for an automobile is shown. In FIG. 15, the automobile 1000 is equipped with a microphone 1001, a vehicle interior sound environment analysis unit 1002, and a display 1003.

When the automobile 1000 is traveling, the microphone 1001 collects the sound environment in the vehicle interior, and inputs the collected sound information to the vehicle interior sound environment analysis unit 1002. Here, the sound information is all kinds of sound information in the vehicle interior, and includes sounds flowing from the car stereo, running noise such as road noise and wind noise emitted from the tires during running, and the like. The sound information input to the vehicle interior sound environment analysis unit 1002 is analyzed by the vehicle interior sound environment analysis unit, for example, how easily the sound is likely to induce drowsiness. For example, if it is only monotonous running noise, it is analyzed that the arousal level is low, and it is found that the sound environment is relatively sleepy because it is arranged at a low position on the Russell ring as shown in FIG. In addition, when classical music such as a loose piano solo is played, the sound environment of the vehicle interior is analyzed by the sound environment analysis unit 1002 with a low degree of arousal as shown in FIG. 16, and the sound environment is relatively sleepy. I understand. On the other hand, when intense rock'n'roll music is playing, the vehicle interior sound environment analysis unit 1002 analyzes the sound environment with a high degree of arousal as shown in FIG. By showing this result on the display 1003, the driver 1004 can grasp how sleepy the current sound environment in the vehicle interior is likely to be.

By doing this, in a sound environment that tends to make you sleepy, the driver 1004 likes caffeine such as coffee and actively talks to passengers to make silence (only driving noise), classical music, etc. You can be careful not to get sleepy while driving even in a quiet sound environment, and you can comfortably reduce the risk of traffic accidents.

As the content to be displayed on the display 1003, the Russell ring of FIG. 16 may be displayed as it is, or it may be displayed as other text information. For example, when the environment is analyzed in the lower half of the Russell ring diagram, "It is a sound environment that tends to make you sleepy" may be displayed. In addition, for example, in the corresponding sound environment, the level of drowsiness that is predicted to reach 15 minutes ahead is measured in advance, and the drowsiness 15 minutes ahead is the vehicle interior sound from the sound information input based on the data. It may be calculated by the environmental analysis unit 1002 and displayed as "sleepiness level after 15 minutes: 4" or the like. The scale of drowsiness level is not limited, but for example, it is displayed in 5 stages, and it is set as level 1: not sleepy at all, level 2: slightly sleepy, level 3: sleepy, level 4: quite sleepy, level 5: very sleepy. You may display it. Of course, it does not matter whether it is 10 minutes or 20 minutes ahead instead of 15 minutes ahead, and here it is not limited as long as it expresses the state of the driver. Further, although drowsiness has been described here as an index analyzed by the vehicle interior sound environment analysis unit 1002, for example, comfort may be used. In that case, the result analyzed by the vehicle interior sound environment analysis unit 1002 is reflected in the left-right axis direction of the Russell ring. In that case, for example, if the driver recognizes that the vehicle interior sound environment is low in comfort, the driver recognizes that the environment is likely to be frustrating while driving, so that the driver is careful to have a pleasant talk with the passenger, for example. Of course, by selecting different songs to create a sound environment that is less frustrating, you can comfortably reduce the risk of traffic accidents. Of course, emotions other than drowsiness and comfort may be used, and for example, the sound environment may be easy to concentrate on, and the present is not limited thereto. Further, although the display 1003 is shown as a display device, the driver 1004 may be notified by voice, for example, by a speaker or the like, or may be notified by vibration, and the method is not limited here.

Next, another embodiment of the present embodiment will be described with reference to FIG. FIG. 17 is the same as FIG. 15, except that the driver monitor camera 1015, the speaker 1016, and the control unit 1017 are added.

The driver monitor camera 1015 is a camera that estimates the state of the driver 1014 from the captured image of the driver 1014, and calculates and outputs, for example, the drowsiness level of the driver 1014. The output drowsiness level is input to the control unit 1017 and connected to the speaker 1016 which is a content presentation unit. Further, the vehicle interior sound environment analysis unit 1002 becomes the content information storage unit 1020, and the output of the content information storage unit 1020 is input to the control unit 1017. The microphone 1011 is also connected to the content information storage unit 1020. Further, in this example, the display 1013 may not be provided.

The operation mechanism of this embodiment will be described with reference to FIG. First, the content information storage unit 1020 includes a vehicle interior sound environment analysis unit 1012, and the vehicle interior sound environment analysis unit 1012 is connected to the microphone 1011 and further connected to music content available in the vehicle 1010. .. By doing so, the content information storage unit 1020 can determine the ease of getting sleepy with respect to the sound environment including the music A collected by the microphone 1011 and the running noise when the music A is flowing in the music content, for example. By analyzing with the vehicle interior sound environment analysis unit 1012 in the information storage unit 1020, it is possible to analyze how sleepy the sound environment is. Furthermore, by analyzing in advance how sleepy the sound environment is for songs other than Music A, the content information storage unit 1020 determines how sleepy each music content is in the car 1010. Is memorized in advance. On the other hand, the drowsiness of the driver estimated by the driver monitor camera 1015 is input to the preference determination unit 1018, and for example, when the drowsiness level is 3 or more, it is determined that the drowsiness of the driver is deepened. Of course, the drowsiness level to be judged may be 2 or 4.5, and is not limited here. If it is determined that the drowsiness level is 3 or higher (Yes in the figure), the content extraction unit 1019 determines that the arousal level is relatively high from the content information storage unit 1020. The music (Music A) is selected and played by the driver 1014 on the speaker 1016, which is a content presentation unit.

By doing this, it is possible to automatically provide the driver with a sound environment that makes it difficult to get sleepy by automatically selecting a song that can increase alertness even when drowsiness decreases, resulting in a traffic accident. It becomes possible to reduce the risk of.

Further, although drowsiness has been described here as an index analyzed by the vehicle interior sound environment analysis unit 1012, for example, comfort may be used. In that case, the result analyzed by the vehicle interior sound environment analysis unit 1012 is reflected in the left-right axis direction of the Russell ring. In that case, for example, when the driver monitor camera 1015 determines that the driver is uncomfortable, the sound environment is less frustrating by selecting a song analyzed by the vehicle interior sound environment analysis unit 1012 as having a high degree of comfort. Can be provided, and the risk of traffic accidents can be comfortably reduced.

Moreover, although the embodiment has been described above using the automobile 1000 and the automobile 1010, this is not limited to the automobile, and may be similarly used for an airplane, a ship, a bus, a train, etc., and here, the present invention is used. It does not limit the use.

Further, an index showing the road noise performance during traveling will be described with reference to FIG. A microphone 1051 and a vehicle interior sound environment analysis unit 1052 are installed in the automobile 1050. At this time, when the vehicle is traveling at a predetermined speed in the automobile 1050, the coordinates collected by the microphone 1051 and analyzed by the vehicle interior sound environment analysis unit 1052 are displayed as the tire performance. For example, when the road noise at a predetermined speed when the tire A is mounted is analyzed by the vehicle interior sound environment analysis unit 1052 as shown in FIG. 20, the performance of the tire A is arousal +0.8 and comfort +0 as shown in FIG. By displaying .2, a person can easily understand how hard it is to get sleepy or uncomfortable by wearing this tire A. At this time, by determining the automobile 1050 as a predetermined automobile, it becomes possible to compare the performance with other tires B and C, and a person can use it as a reference index when purchasing a tire. ..

In this embodiment, the tires were replaced with the automobile as a predetermined vehicle, but of course the opposite is also possible. The predetermined tires are determined, and the vehicle interior sound environment analysis when the automobile 1050 is used as various vehicles at a predetermined speed is analyzed. By extracting and displaying the coordinates analyzed by the unit 1052, it is possible to easily understand how much the running noise such as the wind noise of the automobile 1050 is less likely to be sleepy or uncomfortable.

Needless to say, the configurations shown in the first and second embodiments described above are examples, and various changes can be made without departing from the spirit of the invention.

The content recommendation system in the present disclosure makes it possible to provide content that matches the taste of the user, and is useful as a music or video playback application or the like. It can also be applied to applications such as driving a car or living space such as a house in combination with technologies for maintaining arousal and improving comfort.

100, 200, 300, 400, 500, 1018 Preference judgment unit 101, 201, 301, 401, 501, 1020 Content information storage unit 102, 202, 302, 402, 502, 1019 Content extraction unit 103, 203, 303, 403 , 503 Content presentation unit 204, 404, 504 Operation information acquisition unit 304, 405, 505 Sensor information acquisition unit 406 Scene judgment unit 506 User information storage unit 1000, 1010, 1050 Automobile 1001, 1011, 1051 Mike 1002, 1012, 1052 Car Room Sound Environment Analysis Unit 1003, 1013 Display 1004, 1014 Driver 1015 Driver Monitor Camera 1016 Speaker 1017 Control Unit

Claims (10)

At least a content presentation unit that can play sound or video content,
A preference determination unit that calculates the degree of preference for a plurality of contents presented by the content presentation unit and determines a user's preference, and a preference determination unit.
A content information storage unit that stores the content in association with emotional information of the content.
It has a content extraction unit that extracts content to be reproduced by the content presentation unit from the information stored in the content information storage unit.
The content presenting unit presents the content extracted by the content extracting unit to the user.
A content recommendation system that recommends the content extracted by the content extraction unit. The content recommendation system according to claim 1, wherein the preference determination unit weights the calculated preference degree for the plurality of contents. The content recommendation system has an operation information acquisition unit that acquires operation information related to the playback operation control of the content performed by the user.
The content recommendation system according to claim 1, wherein the preference determination unit determines the user's preference based on the operation information acquired by the operation information acquisition unit. The operation information that can be acquired by the operation information acquisition unit includes a content playback operation, a content stop operation, a content fast forward operation, a content rewind operation, and a list in which the user can refer to the content later. The content recommendation system according to claim 3, which includes at least one of a saving operation and a content volume changing operation. The preference determination unit acquires the operation information from the operation information acquisition unit, and determines the user's preference based on at least one of the time when the operation was performed, the number of times the operation was performed, and at least one of the operation information. The content recommendation system according to claim 4. The content recommendation system has a sensor information acquisition unit that acquires a value measured by a sensor attached to the content recommendation system or a value measured by a remote sensor as sensor information.
The content recommendation system according to claim 1, wherein the preference determination unit determines a user's preference for the content based on the sensor information acquired by the sensor information acquisition unit. The sensors include a vibration sensor that detects the movement of the user's body, a microphone that acquires the voice emitted by the user or the environmental sound around the user, a camera that captures the user's face, and the user's breathing. The content recommendation system according to claim 6, further comprising at least one of a breathing sensor that detects the above and a GPS (Global Positioning System) that detects the position information of the user. The content recommendation system according to claim 6, wherein the preference determination unit determines a user's preference according to the user's physiological index including at least one of the user's drowsiness and the user's emotion by the sensor. The content recommendation system according to claim 2 or 6, wherein the preference determination unit estimates the user's current scene by a sensor and weights the preference degree for the plurality of contents according to the estimated scene. The content recommendation system has a user information storage unit that stores attribute information of the user including at least one of the user information registered in advance and the user information extracted by a sensor, and stores the user information. The content recommendation system according to claim 1, wherein a plurality of the content information storage units are used according to the attribute information of the user stored in the unit.

Images