JP2003173350A - Music and video content recommendation system - Google Patents

Abstract

(57) [Summary] [Problem] When a user's past viewing history is sent over the Internet, the contents are analyzed to create a preference model suitable for the user, and content identification information such as a new song selected by the preference model. In the recommendation service that recommends to users, users could expand the physical space of accessible content, but could only receive the recommendation service corresponding to the narrow mental space captured by their preferences . SOLUTION: With the permission of a charisma in the field, such as a talent, an artist, a disc jockey, a music critic, etc., their selection model can be selected and used by the user, and the content identification information can be provided to the user. Transmission over the Internet allows users a fresh experience of expanding their mental space.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention receives a user's viewing history or the like transmitted on the Internet, analyzes the content of the user, and statistically estimates the user's preference, and the user's interest through the Internet. It relates to the function and configuration of a recommendation engine that provides a recommendation service that recommends and presents music and video contents that you have.

The present invention also provides the above-mentioned recommendations.
Regarding the function and configuration of the recommendation engine and related systems when the engine is installed in the user's music / video playback device.

[0003]

2. Description of the Related Art Originally, a user's preference for audiovisually appealing music / video content is extremely personal, and listening, not listening, seeing, or not seeing a certain content is entirely based on individual free will. It is an act performed. Conventionally, each user knows the existence of these contents from various media that he / she touches in daily life, and listens to or sees the contents he / she prefers. In other words, it filters out personal preferences from the myriad of content and selects only those that pass it.

However, if the user listens to or watches the content that has passed his or her filter only from the information obtained from the medium within the range in which he or she touches in his or her daily life space, it may happen that the user is outside the living space and passes the filter. There is an opportunity loss that you miss the valuable content you get. As a countermeasure, by analyzing the choices made in the user's past daily life and creating a preference model that is similar to the user's filter and recommending the content selected by the preference model to the user, The concept of a recommender system was born, which allows the accessible content space to be dramatically expanded.

It is said that the days of mass production and mass consumption have passed. Today, one-to-one marketing that responds to diversifying user needs, accumulates personal information of each user, and provides appropriate products and services based on it is becoming established mainly in e-commerce. The recommender system is rapidly advancing as a core technology for achieving this, and various methods have been put to practical use.
Collaborative filtering methods and content-based filtering methods are typical examples.

The present invention does not go into the analysis method itself of the recommender system. When the target recommendation engine inputs the user's viewing history in the form of recognition information of music / video content, it creates a preference model suitable for the user, and the music / video selected by the preference model is selected for the user. It has an operation mode of outputting the identification information of the content. Content is C
When D is used as a medium, at present, identification information called CDDB operated by Gracenote Co. of the United States of America is used as a de facto standard. In the case of media other than CD, an identification information system corresponding to CDDB will be established at a time when the diffusion rate of the media becomes high in the future.

[0007]

Now, with the recent increase in storage capacity, it is possible to store music contents corresponding to thousands of CDs in an HDD (hard disk drive) and listen to music the user likes. The size of a stationary or in-vehicle type music playback device called a digital jukebox that can be used has become smaller and more familiar. This device can record the CD reproduction of the user in the form of the number of reproduction times for each identification information of the CDDB. If this record is sent to a site that provides a recommendation service via the Internet, the site provides the user with identification information such as a new song selected by the user's preferred model through the Internet using the recommendation engine described above.
The user can add the recommended content to his digital jukebox by means of purchasing the CD using the identification information or downloading the content from the content providing site via the Internet.

Certainly, the recommended content obtained by this method must be close to the user's preference. But is this enough for the user to get emotional satisfaction? There is often charisma that users want in this area. Specific talents, artists, disc jockeys, music critics, etc.
What would he / she recommend to the user? At present, there is no recommendation service that enables such recommendation.

Let's expand the subject further. If the user is a friend / acquaintance, in such a case, the user may be interested in what content to recommend. At present, there is no recommendation service that enables such recommendation.

[0010]

According to the present invention, among these charisma, the charisma which permits the disclosure of one's own preference model can receive the recommendation service by his / her selection model if the user specifies. Make the recommendation engine work. And his / her selection model is managed and stored in the public space of the storage device.

On the other hand, the user's preference model is to protect and manage privacy by protecting it in a private space of the storage device so that no one other than the user can access it.
However, only if the user has access with the special permission of a friend, acquaintance, etc. who is another user of the recommendation service, if the user specifies, the recommendation service based on the preference model of the other user is provided. Make the recommendation engine work for you.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION First Embodiment of the Invention The embodiment of the present invention as described above will be described in detail with reference to FIG. The processing process for the general recommendation service is as follows. The device 2 of the user accesses the Internet 1 and transmits to the RME (recommendation engine) 3 of the recommendation service site the identification information of the content, which is the viewing history of the user, and the record of the number of times of reproduction thereof. The RME 3 analyzes the data to create a user's preference model, and sends identification information such as a new song selected by the preference model to the user's device 2 via the Internet 1 by the user's designation. The user's preferred model is housed in private space 4 of a storage device protected by confidentiality.

The processing process for receiving the charismatic recommendation service is as follows. The user's device 2 accesses the Internet 1, and the recommendation service site R
Specify charisma in ME3. The RME 3 takes out a pre-registered preference model of the charisma from the public space 5 of the storage device, selects all the contents selected by the preference model or only new songs according to the user's specification, and identifies the identification information on the Internet. 1 to the user's device 2.

When a user receives a recommendation service based on a preference model of another user such as a friend or acquaintance, the following is performed. The user presents to the RME 3 an encrypted signal or the like indicating that the user has been given special access rights. RME3 collates and confirms this, extracts the other user's preference model from the private space of the storage device,
All the contents selected by this preference model or only new songs are selected according to the user's specification, and the identification information is sent to the user's device 2 via the Internet 1.

The user's preference model is obtained by mechanically analyzing the content identification information, which is their viewing history, and the number of times the content is played, but the charisma's selection model requires more information and more elaboration. It can be created by analysis means. This is because most of them / they are celebrities, their position in music etc. has been established, and a high degree of analysis is possible by using many elements related to the preference model.

Second Embodiment of the Invention In the embodiment of the invention described above, the user's device and the recommendation engine exchange information by interposing the Internet. Next, as another mode of application of the present invention, the user's device has a recommendation engine built in, analyzes the database of the user's viewing history, operates in a closed system state, and recommends songs and the like to the user. The case will be described with reference to FIG.

In FIG. 2, a home music server or the like 12 is an electronic device for controlling the reproduction of music / video, and its control center is a PLM (playlist manager).
It is a program called 14. The PLM 14 is an HDD (hard disk device) or the like for storing contents 1
5, UI (user interface module program) 13, which is an interface with the audio / video playback hardware 11, RME (recommendation engine module program) 16, which is built in above, IOU (input / output device) for inputting / outputting to / from a storage medium ) 17, NCE with wired or wireless access to Internet 1
(Network communication engine / module program and related hardware) 18 and the like are controlled.

Now, it is assumed that the selection models of charisma and friends / acquaintances are recorded in a storage medium such as a memory stick or a CD-R. The IOU 17 reads selection models of charisma and friends / acquaintances from these storage media, and stores them in the HDD 15 through the PLM 14. When the user selects charisma or friends / acquaintances and inputs the fact from the music / video reproduction hardware 11, the home music server 12 etc. sends the PLM 14 via the UI 13.
Then, the selected model is retrieved from the HDD 15 or the like and passed to the RME 16. The RME 16 activates a necessary recommendation function to specify the recommended content stored in the HDD 15 or the like. The PLM 14 takes out the content from the HDD 15 or the like based on the designation and reproduces the content by the music / video reproduction hardware 11 via the UI 13. The NCE (network communication engine module program and related hardware) 18 accesses the Internet 1 each time it is needed, selects from the new contents that appeared after the previous access with the RME 16, etc., and downloads it to the HDD 1
5 is for storing.

[0019]

Since the present invention is constructed as described above, it has the following effects.

According to the present invention, the user can receive the recommendation service according to his / her preference model from the recommendation service site or the system built in the device owned by the user, and also according to the charisma preference model specified by the user. You can receive the selected recommendation service. This is not just a physical expansion of the content / access space, but you can experience the wonderful world of expansion of the spiritual space that goes into the individuality of charisma that you respect and worship.

Further, according to the present invention, the user can think of his / her friend / acquaintance, etc., and if he / she is presented with what kind of content should be selected in such a situation, the user can have an empathy. You can enjoy the blissful experience of enjoying the contents.

[Brief description of drawings]

FIG. 1 shows a system diagram corresponding to claim 1 according to the present invention.

FIG. 2 shows a system diagram corresponding to claim 2 according to the present invention.

[Explanation of symbols]

1 Internet 2 User Equipment 3 Recommendation Engine 4 Private Space of Storage Device 5 Public Space of Storage Device 11 Music / Video Playback Hardware 12 Home Music Server etc. 13 UI (User Interface Module Program) 14 PLM (Play List manager) 15 HDD (storage device such as hard disk) 16 RME (Recommendation engine module program) 17 IOU (I / O device) 18 NCE (Network communication engine module program and related hardware) 19 Content providing site

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) G06F 17/60 502 G06F 17/60 502 G10K 15/02 G10K 15/02 H04N 7/173 610 H04N 7/173 610Z 640 640Z 17/00 17/00 Z

Claims (2)

[Claims] 1. A user's viewing history is received via the Internet, the content is analyzed to create a preference model suitable for the user, and the identification information of the new music / video content selected by the preference model is transmitted via the Internet. In the recommendation service that is sent to the user via the user, the selection model is stored in the public space of the storage device with the permission of the charisma of the relevant field such as talent, artist, disc jockey, music critic, and the preference model of each user. Is stored in a private space of a storage device provided with a protection mechanism that cannot be taken out without user's permission, and when the user specifies a particular charisma via the Internet, the selection model of that charisma is described above in the storage device. It can be taken out of public space and activated to allow users to gain access from specific users. If it is systematically approved, when the user specifies the specific user, the preference model of the specific user is taken out from the private space of the storage device and activated, and the selected music Recommendation engine system structure that provides identification information of video contents to users via the Internet. 2. A user's viewing history is stored as a database, the content is analyzed to create a preference model suitable for the user, and the user is provided with the identification information of the new music / video content selected by the preference model. In the recommendation function, the CD of the selection model of charisma in the field such as talent, artist, disc jockey, music critic, or other user's preference model provided by another user is provided.
-When a user specifies a particular charisma or other user by obtaining and inputting in the form of a storage medium such as a ROM or a memory stick, the charisma or the selection model of the other user is activated, and the selected content A system structure of a device-equipped recommendation engine that controls music and video content based on identification information so that it can be provided to users.