WO2018001223A1

WO2018001223A1 - Playlist recommending method and device

Info

Publication number: WO2018001223A1
Application number: PCT/CN2017/090234
Authority: WO
Inventors: 贺冯良
Original assignee: 中兴通讯股份有限公司
Priority date: 2016-06-27
Filing date: 2017-06-27
Publication date: 2018-01-04
Also published as: CN107547916A

Abstract

The present disclosure provides a playlist recommending method and a device, the method comprising: acquiring program watching behavior data of a user by means of a set top box; generating a recommended program playlist according to the behavior data; and displaying the recommended program playlist. The method solves the problem in the related art that programs to be played cannot be intelligently recommended to the programs of the set top box according to personal preferences of the user, thus improving the user experience.

Description

Playlist recommendation method and device

Technical field

The present disclosure relates to the field of communications, and in particular, to a playlist recommendation method and apparatus.

Background technique

With the development of communication technology, the application of multimedia technology has become more and more extensive. Cloud computing, cloud storage and big data analytics technology are also gradually matured and commercialized, and users' needs are understood based on user behavior data. At the same time, the development of targeted services and product solutions, and the targeted sale of these services and even more new products to customers will become one of the trends in the future commercialization of big data. At present, big data is also gradually being promoted and used in social networking (such as Facebook) and e-commerce (such as Amazon).

As one of the important consumer terminals of the digital home, set-top boxes have been widely used and applied on a large scale. The set-top box plays an important role in the audio-visual entertainment engine in the digital home. Program-on-demand, live broadcast and other interactive services will become a highlight of the commercial potential of the set-top box. How to dig out the user character, hobbies and needs of the set-top box side to provide accurate, personalized and efficient products and services for the set-top box users has become an important requirement in the field of set-top boxes.

In view of the problem in the related art that the program cannot be recommended for the set-top box program intelligently according to the user's personal preference, an effective solution has not been proposed.

Summary of the invention

The embodiments of the present disclosure provide a playlist recommendation method and apparatus to solve at least the problem that the program cannot be recommended for the set top box program smart according to the user's personal preference.

According to an embodiment of the present disclosure, a playlist recommendation method is provided, including: acquiring behavior data of a user watching a program through a set top box; generating a recommended program playlist according to the behavior data; and displaying the recommended program playlist.

Optionally, after obtaining the behavior data of the user watching the program through the set top box, the method further includes: storing the behavior data on the cloud server.

Optionally, generating the recommended program playlist according to the behavior data includes: generating different recommended program playlists according to the behavior data according to different time periods.

Optionally, generating the recommended program playlist according to the behavior data comprises: searching for an object related to the attribute of the behavior data from a library containing all data according to an attribute of the behavior data; filtering according to a predetermined rule, remaining For the recommended program playlist.

Optionally, displaying the recommended program playlist includes: authenticating the user's authority; and in the case of passing the authentication, displaying the recommended program playlist.

Optionally, the behavior data includes at least one of the following: a page browsing record, a play, a fast forward, a pause, a volume adjustment, a loop play, a search count, and a rating of the program.

Another aspect of the embodiments of the present disclosure further provides a playlist recommendation apparatus, including: an acquisition module, configured to acquire behavior data of a user watching a program through a set top box; and a generating module configured to generate a recommended program play according to the behavior data a display module configured to display the recommended program playlist.

Optionally, the device further includes: a storage module configured to store the behavior data on the cloud server.

Optionally, the generating module includes: a generating unit, configured to generate different recommended program playlists according to the behavior data according to different time periods.

Optionally, the generating module includes: a searching unit, configured to search for an object related to the attribute of the behavior data from a library containing all data according to an attribute of the behavior data; and the filtering unit is set to follow a predetermined rule Filtered, and the rest is the recommended program playlist.

Through the disclosure, the behavior data of the user watching the program is acquired by the set top box; the recommended program play list is generated according to the behavior data; and the recommended program play list is displayed, and the related technology cannot be recommended for the smart set top box program according to the user's personal preference. The problem with the program has improved the user experience.

DRAWINGS

The drawings described herein are provided to provide a further understanding of the present disclosure, which is a part of the present disclosure, and the description of the present disclosure and the description thereof are not intended to limit the disclosure. In the drawing:

FIG. 1 is a flowchart of a playlist recommendation method according to an embodiment of the present disclosure;

2 is a flowchart of a playlist recommendation device in accordance with an embodiment of the present disclosure;

3 is a flowchart 1 of a playlist recommendation device in accordance with a preferred embodiment of the present disclosure;

4 is a second flowchart of a playlist recommendation device in accordance with a preferred embodiment of the present disclosure;

5 is a flowchart 3 of a playlist recommendation device in accordance with a preferred embodiment of the present disclosure;

6 is a flow diagram of recommending a playlist for a user in accordance with an embodiment of the present disclosure.

detailed description

The present disclosure will be described in detail below with reference to the drawings in conjunction with the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments may be combined with each other without conflict.

It is to be understood that the terms "first", "second", and the like in the specification and claims of the present disclosure are used to distinguish similar objects, and are not necessarily used to describe a particular order or order.

In this embodiment, a playlist recommendation method is provided. FIG. 1 is a flowchart of a playlist recommendation method according to an embodiment of the present disclosure. As shown in FIG. 1, the process includes the following steps:

Step S102, obtaining, by the set top box, behavior data of the user watching the program;

Step S104, generating a recommended program playlist according to the behavior data;

Step S106, displaying the recommended program playlist.

Through the above steps, the behavior data of the user watching the program is obtained by the set top box; the recommended program playlist is generated according to the behavior data; and the recommended program playlist is displayed, which solves the problem that the relevant technology cannot recommend the program for the set top box program according to the user's personal preference. The problem is to improve the user experience.

Optionally, after the behavior data of the user watching the program is obtained by the set top box, the method further includes: storing the behavior data on the cloud server.

In order to further improve the user experience, different recommended program playlists may be generated according to the behavior data according to different time periods.

In an optional embodiment, generating the recommended program playlist according to the behavior data may include: searching, according to the attribute of the behavior data, an object related to the attribute of the behavior data from a library containing all the data; filtering according to a predetermined rule The rest is the recommended show playlist.

In order to achieve personalized display for different customers, the user's authority can be authenticated; in the case of authentication, the recommended program playlist is displayed.

Optionally, the behavior data may include at least one of the following: a page browsing record, a play, a fast forward, a pause, a volume adjustment, a loop play, a search count, and a rating of the program.

The embodiment of the present disclosure further provides a playlist recommendation apparatus. FIG. 2 is a flowchart of a playlist recommendation apparatus according to an embodiment of the present disclosure. As shown in FIG. 2, the method includes: an acquisition module 22 configured to acquire a user view through a set top box. The behavior data of the program; the generating module 24 is configured to generate a recommended program playlist according to the behavior data; and the display module 26 is configured to display the recommended program playlist.

3 is a flowchart 1 of a playlist recommendation apparatus according to a preferred embodiment of the present disclosure. As shown in FIG. 3, the apparatus further includes a storage module 32 configured to store the behavior data on a cloud server.

4 is a flowchart 2 of a playlist recommending apparatus according to a preferred embodiment of the present disclosure. As shown in FIG. 4, the generating module 24 includes: a generating unit 42 configured to generate different recommended programs according to the behavior data according to different time periods. playlist.

5 is a flowchart 3 of a playlist recommendation apparatus according to a preferred embodiment of the present disclosure. As shown in FIG. 5, the generation module 24 includes: a search unit 52 configured to search from a library containing all data according to attributes of the behavior data. An object related to the attribute of the behavior data; the filtering unit 54 is configured to filter according to a predetermined rule, and the rest is the recommended program playlist.

Embodiments of the present disclosure also provide a storage medium. Optionally, in this embodiment, the foregoing storage medium may be configured to store program code for performing the following steps: Step S1, acquiring behavior data of a user watching a program through a set top box; and step S2, generating a recommended program according to the behavior data. a playlist; in step S3, the recommended program playlist is displayed.

Optionally, in this embodiment, the foregoing storage medium may include, but not limited to, a USB flash drive, a Read-Only Memory (ROM), a Random Access Memory (RAM), a mobile hard disk, and a magnetic memory. A variety of media that can store program code, such as a disc or a disc.

Optionally, in the embodiment, the processor performs the above steps S1, S2, and S3 according to the stored program code in the storage medium.

For example, the specific examples in this embodiment may refer to the examples described in the foregoing embodiments and the optional embodiments, and details are not described herein again.

The embodiment of the present disclosure exploits the user behavior habits of the user-side set-top box by means of the big data technology of the Internet at that time. Interesting, potential needs, to achieve accurate, personalized, efficient programs and product recommendations for users on the set-top box.

In order to perform accurate, personalized, and efficient program and product recommendation for the user on the set top box, FIG. 6 is a flowchart for recommending a program list for the user according to an embodiment of the present disclosure. As shown in FIG. 6, the following is mainly completed.

In steps S602-S604, big data is acquired, and big data can be obtained in two ways.

Pathway 1. Capture the behavior data of the set-top box user (such as page browsing record, play, fast forward, pause, adjust volume, loop play, search times, comment score, etc.) and store it on the cloud server in the background to form a A huge consumer database.

Path 2, sharing consumer behavior data through a third-party search company or video site.

The big data system they have accumulated, in addition to providing traditional program heat rankings, program sorting, and more can provide users with the habit of watching programs.

At step S606, processing of big data is performed. The characteristics of big data are: large scale, multiple types of data, fast data processing, and low data density, while current cloud storage, cloud computing, and data mining calculations can be completed. The reason why traditional recommendation is to choose popular recommendation as the benchmark algorithm is also obvious. Users always prefer programs that everyone likes. However, popular recommendations are antonyms of personalized recommendations, and the results generated for each user are the same. Therefore, our goal is to find a personalized sorting algorithm that is better than popular recommendations to meet the different tastes of different users. That is, a personalized recommendation algorithm, recommending the right program or service to the right user at the right time. There are many algorithms for machine learning and data mining in this category, such as the cinematch algorithm of Netflix.

At steps S608-S610, a recommendation of a program or other service is performed.

The basic implementation idea of the recommendation is: according to the attribute of the operation collection object, according to the collected attribute condition, the object related to it is searched from the library containing all the data, filtered according to a certain rule, and the rest is recommended to the user.

1) Recommend a program of interest to the user based on a specific user.

If the current user A is a fan of American drama and enjoys action movies, the user's page may recommend similar scripts such as "Jailbreak 4", "Shenzhen Agent", "American Spy Dream", "Arrow" and the like.

The background server classifies the resources and creates different collection classes. Regional collection: American drama class Aa_Class, Korean drama Ba_Class, Japanese drama Ja_Class, Taiwanese drama Ta_class, Hong Kong drama Ha_Class, mainland drama Ca_Class; type collection: history class Ht_Class, country Ct_Class, idol drama Lt_Class; adventure class: Adt_Class, action class: Action_Class. Inspirational class: inspir_Class.

The same reason can be established by the collection of chronological points, according to the collection of actors.

Establish a user's behavioral correlation coefficient by capturing the user's clicks and ratings.

The system establishes the coefficients of the user and resource mapping. Regional coefficient: American drama coefficient Accoff, Korean drama coefficient Bcoff, Japanese drama coefficient: Jcoff, Taiwan drama coefficient Tcoff, mainland drama Ccoff: Hong Kong drama coefficient Hcoffee; similarly by type, according to actors, by age, etc. also establish corresponding coefficients. When the user clicks on the program, the corresponding coefficient will be incremented once. For example, if you click "Da Qin Empire", the mainland drama coefficient Ccoff, the historical drama coefficient Hcoff, the age coefficient 2013coff will increase by 1, the more clicks, the greater the correlation coefficient.

According to the correlation coefficient size, the relevant resource classes are matched, and then a related program is taken out from the related resources.

A correlation coefficient (coff) of 0 is not recommended.

Programs with a large correlation coefficient (coff) are preferred, and are sorted by the correlation coefficient size.

The programs recommended by the correlation coefficient are relatively extensive, and may not accurately meet the user's preferences.

If you want to recommend the program more accurately, use recursive screening:

N recommendation coefficients

For example, first select the mainland dramas by region, and then recommend the selection of mainland historical dramas on the basis of mainland dramas, and then choose war on the basis of mainland historical dramas. In this way, it is recommended to recommend to the user such as "Da Qin Empire" and "Chu Han Fengyun".

2) In special programs, according to the user's interests, special program segments can also be recommended while providing complete programs.

The user's interest type is analyzed by the user viewing the historical data of the program.

Analysis from the behavior correlation coefficient.

From the idol drama coefficient Lcoff, the action drama coefficient ActionCoff, the expedition coefficient Cooff, the motivation coefficient: the spira coefficient, etc., whether the user is an idol drama preference or an action class preference, or an adventure class preference type, or an inspirational preference type, so that a certain The pieces whose source matches the feature are intercepted and recommended to the user.

Because these clip users may be more interested, and even repeat the recollection.

For example, in "Prison Break", you can cut the most exciting dialogues (classic lines), exciting movements, touching pictures (such as parting pictures) into small pieces, and recommend to users to select on-demand to attract users. Go to the on-demand.

3) Recommend different programs for users according to different time periods.

The user has 7 days in a week, 24 hours a day, and the time is different, and the program category he decides is different.

Programs are classified according to duration.

Continuous play class: such as serial S_Class; 90 minutes program class: movie Film_Class, variety film Z_Class; program within 30 minutes: such as microfilm Minfilm_Class, news video News_Class.

The user's online time of one week is recorded according to the time of the user's online time of the week: T1, T2, T3, T4, T5, T6, T7.

The corresponding program class is recommended according to the Tn size. If Tn is greater than 2 hours, serials and movies are recommended. If Tn is less than 2 hours and more than half an hour, recommend movies, variety films, micro movies, news videos. If Tn is less than an hour, Only recommended variety films, micro-movies, news videos, such as Tn less than 30 minutes, only recommend micro-movies and news videos.

Such as: Tuesday, Tuesday, Thursday night, go home late at night, just want to see micro-movies, news videos and so on. On Wednesday night, on Friday night, when I got home from work, I wanted to see programs like "Decoding Financials", "Financial Lang Lang" and "Not You Are". I want to see the full movie, TV series on weekend night. The system accurately recommends programs based on time periods and personal interests.

4) User identification.

Personalization is for every family, and different members of each family are likely to have inconsistent interests. We have to make recommendations for "dad", "mother", "child" or the whole family. The recommendation system can set a sub-account, complete the user's identification for the sub-account and enter the page system of different roles.

Through the embodiment of the present disclosure, the data is analyzed and processed based on the data of the big data platform on the streaming media platform, and the deep mining further discovers the behavior habits, hobbies and potential needs of the user of the set top box, and then the user on the set top box is accurately performed. Personalized, efficient programs or recommendations for value-added products.

It will be apparent to those skilled in the art that the various modules or steps of the present disclosure described above can be implemented by a general-purpose computing device that can be centralized on a single computing device or distributed across a network of multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device such that they may be stored in the storage device by the computing device and, in some cases, may be different from the order herein. The steps shown or described are performed, or they are separately fabricated into individual integrated circuit modules, or a plurality of modules or steps thereof are fabricated as a single integrated circuit module. As such, the disclosure is not limited to any specific combination of hardware and software.

The above description is only a preferred embodiment of the present disclosure, and is not intended to limit the disclosure, and various changes and modifications may be made to the present disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and scope of the present disclosure are intended to be included within the scope of the present disclosure.

Claims

A playlist recommendation method includes:

Obtaining behavior data of a user watching a program through a set top box;

Generating a recommended program playlist according to the behavior data;

The recommended program playlist is displayed.
The method of claim 1, wherein after obtaining the behavior data of the user watching the program through the set top box, the method further comprises:

The behavior data is stored on the cloud server.
The method of claim 1, wherein generating the recommended program playlist based on the behavior data comprises:

Different recommended program playlists are generated according to the behavior data according to different time periods.
The method of claim 1, wherein generating the recommended program playlist based on the behavior data comprises:

Searching for an object related to an attribute of the behavior data from a library containing all data according to an attribute of the behavior data;

Filtered according to predetermined rules, and the rest is the recommended program playlist.
The method of claim 1 wherein displaying the recommended program playlist comprises:

Authenticate the user's authority;

In the case of passing authentication, the recommended program playlist is displayed.
The method according to any one of claims 1 to 5, wherein

The behavior data includes at least one of the following: page browsing record, play, fast forward, pause, adjust volume, loop play, number of searches, and score the comment of the program.
A playlist recommendation device includes:

Obtaining a module, configured to obtain, by the set top box, behavior data of the user watching the program;

Generating a module, configured to generate a recommended program playlist according to the behavior data;

a display module configured to display the recommended program playlist.
The apparatus of claim 7 wherein said apparatus further comprises:

A storage module configured to store the behavior data on a cloud server.
The apparatus of claim 7, wherein the generating module comprises:

The generating unit is configured to generate different recommended program playlists according to the behavior data according to different time periods.
The apparatus of claim 7, wherein the generating module comprises:

a search unit configured to search for an object related to the attribute of the behavior data from a library containing all data according to an attribute of the behavior data;

The filtering unit is set to filter according to a predetermined rule, and the rest is the recommended program playlist.
A computer storage medium storing execution instructions for performing the method of any one of claims 1 to 6.