KR20140052387A - An intelligent recommendati on method of smart tv contents - Google Patents

Abstract

The present invention relates to an intelligent content recommendation method for a smart TV and the method comprises the following steps: S10) contacting a content sensor; S11) automatic operation; S12) preference registration/conversion; S13) matching content and a program; S14) analyzing targeted recommendable content; S15) content recommendation; S16) content selection and operation; and S17) content update. The present invention can recommend proper content for each user in a social service system considering the preference of a smart TV user, favorite TV programs and use history, so that users′ satisfaction can be improved and it is highly possible for users to accept the recommended content reflecting TV program watching patterns and preference of each user. In addition, smart TV manufacturers can improve its brand image, user satisfaction and service qualities by providing content which can be utilized through each manufacturer′s smart TV. Moreover, content providers can create higher profits since the usage possibility of content provided based on the conditions of each user in various ways can be improved.

Description

TECHNICAL FIELD The present invention relates to an intelligent recommendation method for smart contents on a smart TV,

The present invention relates to a real-time content recommendation method, and more particularly, to a real-time content recommendation method for use in a content service system in consideration of a preference of a smart TV viewer, a frequently viewed TV program, The present invention relates to an intelligent content recommendation method on a smart TV.

As the era of digital broadcasting becomes popular, users who are looking for a smart TV (TV) that provides various functions and contents are increasing day by day. Since the functions provided in the existing smart device are defective with the TV, the smart TV can generate a synergy effect through a wider screen and an easier user interface than other smart devices.

In the end of 2012, various smart TVs are pouring out in the era of digital broadcasting and many contents are newly added as users increase.

Unlike analog TV, digital TV provides detail information of a program, and it can derive user 's program viewing pattern by using it. That is, the user can select one of various contents provided on the smart TV, but the technology for recommending the contents suitable for the user is not provided, and the list is displayed at the category level.

The conventional technology developed so far is disclosed as a " real-time content recommendation apparatus and method "as Patent Publication No. 2012-0091457, in a published patent publication.

The "real-time content recommendation apparatus and method" of the above-mentioned Japanese Patent Application Laid-Open Publication No. 2012-0091457 is a system including a server for providing contents as shown in FIG. 1, The content recommendation apparatus comprising: an input unit (11) for inputting a dialog based on a social network activity among a plurality of users; A first storage unit (12) for storing information on contents collected by the input unit in units of keywords in real time; A second storage unit (13) receiving from the server and storing metadata of contents broadcasted by the server in real time; Analyzing the correlation between the keyword and the content in real time with reference to the keyword stored in the first storage unit (12) and the metadata of the content stored in the second storage unit (13) An analysis unit (14) for deriving a social issuing ranking of the currently broadcasted content based on the ranking information; And a recommendation unit (15) for recommending the content to the user terminal in real time with reference to the issue issuing rank analyzed by the analyzing unit (14). The recommendation unit (15) , It is possible to analyze the social relations of image contents and to recommend images that become issues in real time.

However, the above-mentioned " real-time content recommendation apparatus and method " of the above-mentioned Patent Publication No. 2012-0091457 is limited to a list of categories.

Also, as a flowchart showing a method of executing a conventional content recommendation method, a content recommendation method as shown in Fig. 2 is used.

However, the content recommendation shown in FIG. 2 is automatically activated when the smart TV is turned on and connected to the content center (step S1), and the operator recommendation method such as the latest popularity ranking is adopted (step S2). However, There is a problem in that it is not possible to recommend the smart TV contents most suitable for the user's use of the contents (S4 step) by selecting the TV programs and the contents usage statistics which have been used in the past.

The present invention has been made to solve various drawbacks and problems caused by the conventional content recommendation method described above, and its object is to provide a smart TV viewer's preference, a frequently used TV program, a past history of use of contents It is possible to improve the acceptability and satisfaction of the recommendation result by recommending the content reflecting the preference of the user and the watching pattern of the TV program on the users side by recommending the most suitable content for use in the content service system, And to provide an intelligent content recommendation method on the basis of the content.

It is another object of the present invention to provide a smart TV manufacturer that provides convenience of contents that can be utilized through its own smart TV in order to improve service quality, user satisfaction and product image, And to provide a content recommendation method.

It is still another object of the present invention to provide a content providing system and a content providing method capable of recommending contents in consideration of a user's situation in various aspects, The content recommendation method.

According to an aspect of the present invention, there is provided an intelligent content recommendation method on a smart TV, comprising: a content center accessing step (S10) for a user to input an ID and a password to a content center to access the content center; An automatic driving step (S11) of driving the content recommendation apparatus; A preference registration / changing step (S12) of registering or changing the preference of the user's content with the preference registration module 10; A content / program matching step (S13) of connecting a category of content provided in the smart TV with a TV program category by the content / program matching module 20; A recommendation target content analysis step (S14) for analyzing the multi-dimensional recommendation target content for deriving the recommendation target content analysis module 30; The content recommendation module 40 calculates the final score by summing the recommendation score R1 of the user preference, the recommendation score R2 of the program side, and the recommendation score R3 of the content side, A recommendation step (step S15); The content selection and execution module 50 retrieves and excludes the content already being used by the recommendation target from the content_Stat database DB and calculates a recommended content list according to the final score R calculated by the content recommendation module 40 Selecting and executing a content to be presented on a user's smart TV (step S16); The content update information is updated to the content_stat database DB recommended by the content update module 60 and the content usage statistical information is updated in a content update step S17 used in the next content recommendation.

The present invention proposes content most suitable for use in a content service system in consideration of a preference of a smart TV viewer, a frequently viewed TV program, past use history, etc., It will increase the acceptability and satisfaction of recommendation results by recommending contents reflecting preference and watching patterns of TV programs, and providing convenience of contents that can be utilized through its smart TV in terms of Smart TV Manufacturers The service quality, the user satisfaction, and the product image can be improved. In addition, since the content provider can recommend the content considering the user's situation in various aspects, the availability of the provided content is increased, Capable of creating There are special advantages.

1 is a block diagram showing an embodiment of a conventional real-time content recommendation apparatus,
FIG. 2 is a flow chart showing an execution procedure of an intelligent content recommendation method on a conventional smart TV,
3 is a flow chart showing the execution of the intelligent content recommendation method on the smart TV of the present invention,
4 is a conceptual diagram of an intelligent content recommendation method on the smart TV of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an intelligent content recommendation method on the smart TV of the present invention will be described in detail with reference to the accompanying drawings.

4 is a conceptual diagram of an intelligent content recommendation method on the smart TV of the present invention. The apparatus for implementing the intelligent content recommendation method on the smart TV of the present invention is also an apparatus A preference registration module 10 for registering or changing a preference for a user's content as shown in FIG. 4; a content / program matching module 20 for linking a content category and a TV program category provided in the smart TV; A recommendation target content analysis module 30 for analyzing multi-dimensional recommendation target content for deriving a target content to be recommended; A content recommendation module 40 for calculating a final score by summing up a recommendation score R1 on the user preference side, a recommendation score R2 on the program side, and a recommendation score R3 on the content side to determine recommendation target contents; A recommended content list is calculated based on the final score R calculated by the content recommendation module 40 by searching for and excluding content already being used by the recommendation target content DB from the Content_Stat database DB, A selection and execution module 50; The recommended content is updated in the Content_Stat database (DB), and the content usage statistical information is composed of a content update module 60 used in the next content recommendation.

Meanwhile, an intelligent content recommendation method on the smart TV includes a content center accessing step (S10) in which a user inputs an ID and a password to a content center and accesses the content center; An automatic driving step (S11) of driving the content recommendation apparatus; A preference registration / modifying step (S12) of registering or changing the preference of the user's content with respect to the user's content with the preference registration module 10; A content / program matching step (S13) of connecting a category of content provided in the smart TV with a TV program category by the content / program matching module 20; A recommendation target content analysis step (S14) for analyzing the multi-dimensional recommendation target content for deriving the recommendation target content analysis module 30; The content recommendation module 40 calculates the final score by summing the recommendation score R1 of the user preference, the recommendation score R2 of the program side, and the recommendation score R3 of the content side, A recommendation step (step S15); The content selection and execution module 50 retrieves and excludes the content already being used by the recommendation target from the content_Stat database DB and calculates a recommended content list according to the final score R calculated by the content recommendation module 40 Selecting and executing a content to be presented on a user's smart TV (step S16); The content recommended and used by the content update module 60 is updated to the Content_Stat database DB, and the content usage statistical information is composed of a content update step S17 used in the next content recommendation.

The analysis in the recommendation target content analysis step (S14) is user preference analysis, TV program analysis, and content analysis.

Hereinafter, the operation of the intelligent content recommendation method on the smart TV of the present invention as described above will be described in detail as an embodiment.

Example

If the user inputs an ID and a password to the content center in the content center access step (step S10), the content recommendation apparatus is automatically activated in the automatic driving step (step S11), and then the preference registration / , The preference registration module 10 registers or changes the preference for the user's content. At this time, the content category is based on the category provided on the smart TV and can be expanded as needed. Since TV is generally used by individuals as well as families, and there may be many tastes of individuals, preferences can be assigned to a plurality of preferences, and preferences set are stored in a preference DB. This is used in analyzing the content to be recommended together with other elements.

Then, in the content / program matching step (step S13), the recommendation target content analysis module 30 links the category of the content provided on the smart TV with the TV program category. Since a certain content category can correspond to several TV program categories and one TV program can correspond to several content categories, a many-to-many (M: N) relationship is established between them, : N), we introduced Contents / Program Matching DB which resolve. This is to reflect the category of the TV program that I usually enjoy when I derive the content to be recommended at a later stage.

Subsequently, in the recommendation target content analysis step (step S14), the recommendation target content analysis module 30 has performed a multi-dimensional analysis task for deriving the target content. It is divided into user preference analysis function, TV program analysis function, and content analysis function.

1) The user preference analysis extracts the information about the content preference set by the user in the registration module from the Preference DB, finds matches with the categories provided on the smart TV, and sorts them in descending order of ratings. This is expressed in the database language Structured Query Language (SQL) as follows.

SELECT Category_Name, Category_Rating

FROM Contents DB, Preference DB

WHERE User_Preference_Category = Contents_Category

ORDER BY Category_Rating DESC

And calculates a recommendation score (R1) in terms of the user's preference using the extracted result.

2) Analysis of the TV program was carried out from the program database by analyzing the list of the program which was watched on the smart TV. The program information may be modeled by the category provided by the supplier (drama / current affairs / sports / entertainment / children, etc.) and its subclasses. In other words, sub-categories of current affairs education include news, documentary, accusation program, and quiz program. Extract the program that watched the most programs viewed during a specific period. Since the broadcasting period (daily / weekly / once a week / twice a week, etc.) differs for each program, the audience rating of the program P1 on the specific smart TV can be calculated using the following equation (1).

In Equation (1), the denominator is a value predefined by the broadcaster, and the numerator is calculated using the statistics of the user. Equation (1) only calculates for a program viewed at least once during a predetermined period. All program ratings for each program are grouped by program category and sorted in reverse order from highest score. Find content categories linked to the top ranked categories and sort them in order of highest viewer rating.

The recommendation score (R2) of the program side is calculated by using the result performed in the same manner as the SQL statement.

3) Content analysis calculates recommendation score for each content category based on content usage statistics. It is possible to calculate the usage score of the content based on the usage frequency and the usage time within a given period from the Content_Stat DB as shown in Equation (2).

In the formula (2), the use score calculated for each content is aggregated for each category to calculate a recommendation score (R3) on the content side.

라 하면 최종 점수는 이를 가중합하여 계산되고, 초기에는 세 가중치

을 미리 정의된 값으로 설정하고, 이후 필요에 따라 수학식 3으로 조절할 수 있다.Then, in the content recommendation step (S15), the final score is calculated by combining the recommendation score (R1) of the user preference, the recommendation score (R2) of the program side, and the recommendation score (R3) of the content side. The weights of R1, R2, and R3 are

The final score is calculated by weighting it, and initially, the three weights

May be set to a predefined value, and then adjusted to Equation (3) as needed.

Next, in the content selection and execution step (step S16), the recommended target content is searched for and removed from the content_stat DB, and the recommended content list is calculated according to the final score R calculated by the content recommendation module 40, On the smart TV.

Finally, the content recommended and used in the content update step S17 is updated in the content_stat DB, and the content usage statistical information is used in the next content recommendation.

While the present invention has been described with reference to the preferred embodiments, it is to be understood that the invention is not limited thereto and that various changes and modifications may be made therein without departing from the scope of the invention.

10: preference registration module 20: content / program matching module
30: Recommended content analysis module 40: Content recommendation module
50: Content selection and execution module 60: Content update module

Claims (5)

The intelligent content recommendation method on the smart TV includes a content center connection step (S10) in which a user inputs an ID and a password to a content center and accesses the content center; An automatic driving step (S11) of driving the content recommendation apparatus; A preference registration / changing step (S12) of registering or changing the preference of the user's content with the preference registration module 10; A content / program matching step (S13) of connecting a category of content provided in the smart TV with a TV program category by the content / program matching module 20; A recommendation target content analysis step (S14) for analyzing the multi-dimensional recommendation target content for deriving the recommendation target content analysis module 30; The content recommendation module 40 calculates the final score by summing the recommendation score R1 of the user preference, the recommendation score R2 of the program side, and the recommendation score R3 of the content side, A recommendation step (step S15); The content selection and execution module 50 retrieves and excludes the content already being used by the recommendation target from the content_Stat database DB and calculates a recommended content list according to the final score R calculated by the content recommendation module 40 Selecting and executing a content to be presented on a user's smart TV (step S16); The contents used and recommended by the content update module 60 are updated in the Content_Stat database DB. The content usage statistical information is updated in the content recommendation step S17, Way. The intelligent content recommendation method according to claim 1, wherein the analysis in the recommendation target content analysis step (S14) is user preference analysis, TV program analysis, and content analysis. The method of claim 2, wherein the user preference analysis is performed by extracting, from the preference DB, information about the content preference set by the user in the preference registration module (10) And expressed as a database language, Structured Query Language (SQL), as follows.
SELECT Category_Name, Category_Rating
FROM Contents DB, Preference DB
WHERE User_Preference_Category = Contents_Category
ORDER BY Category_Rating DESC
And calculating a recommended score (R1) in terms of user preference using the extracted result. The intelligent content recommendation method according to claim 2, wherein the TV program analysis is performed according to Equation (1) by analyzing a program list previously viewed on a smart TV from a program DB.
[Equation 1]

3. The content analysis method according to claim 2, wherein the content analysis is to calculate a recommendation score for each content category based on content usage statistics, and calculates a usage score of the content based on the usage frequency and the usage time within a given period from the Content_Stat DB, 2 < / RTI >
&Quot; (2) "

Images