JP2007049322A - Program recommendation device - Google Patents

Abstract

In a program recommendation device for recommending a program suitable for a user's preference, an appropriate program is recommended for a plurality of users who are watching at the same time.
A program recommendation device for recommending a program based on user preferences, a user identification unit for identifying a set of users viewing at the same time, a user profile management unit for managing profile information of a plurality of users, A recommended program determination unit that determines a recommended program using user information managed by the profile management unit, and the recommended program determination unit uses the profile information of each user managed by the user profile management unit. Recommendations that match the user set identified by the user identification unit are made.
[Selection] Figure 1

Description

  The present invention relates to an apparatus for recommending a user with a television program that matches user preferences.

  With the increase in the number of TV channels, it has become difficult for the user to select a favorite program from a large amount of program information posted in TV program guides such as newspapers and EPG (electronic program guide).

  For this reason, devices have emerged that profile user preferences from the user's viewing history, detect programs that match the preferences, present them to the user, and automatically record.

  Further, in order to deal with a plurality of users, an invention relating to an apparatus that includes means for identifying users, profiles each user's preference, and presents a recommended program that matches the preference to each user is disclosed in Patent Document 1 or 2 is disclosed.

JP 2000-287189 A

JP 2003-219287 A

  However, even if a program is recommended for each individual user, if multiple users view the program at the same time, the user must select the program by comparing the recommended programs presented to the individual users. It is not easy to select a program suitable for a plurality of users.

  The present invention identifies a set of users who are watching at the same time, manages a profile of a plurality of users, and collectively uses a plurality of user profiles belonging to the identified user set to collect a set of users. The recommended program is determined for the entire program.

  Further, the program selection priority between users is profiled, and the program selection priority profile is used for recommended program determination.

  Since an appropriate program recommendation is performed for the entire set of users who are watching at the same time, it is not necessary to select a program by comparing program recommendations for individual users, and the program can be selected easily.

  In addition, by using the program selection priority profile between users, it is possible to make a recommendation reflecting the daily program selection habits when used by a plurality of users.

  Embodiments of the present invention will be described below with reference to FIGS.

  FIG. 1 is a diagram showing the configuration of this embodiment. The program recommendation device 101 includes a control unit 102, a user identification unit 103, a user profiling unit 104, a user profile management unit 105, a recommended program determination unit 106, and a user interface unit 107, which are provided in a computer such as a personal computer. It can consist of external memory such as memory and HDD, real-time clock, microprocessor, console, and keyboard. The operation of each part of the present invention can be realized by software such as a program module installed and operated on a computer such as the personal computer described above. Further, the present invention can be implemented not only in a personal computer but also in a computer built-in device such as a television, a hard disk recorder, and a portable information terminal having a similar configuration.

  Hereinafter, each part which comprises a present Example is demonstrated.

  The control unit 102 controls the entire program recommendation device 101.

  The user identification unit 103 identifies a plurality of users who are viewing at the same time by identifying a face image captured by the camera, a fingerprint authentication system, or the like. The user may input personal identification information via input means such as a remote controller.

The user profiling unit 104 receives electronic data of the operation history 102 such as a user's viewing program and the viewing time and recorded program. As the viewing program information, electronic data of program information such as a program title, a genre, a performer name, and a program outline description included in an EPG received by a television receiver or an EPG for the Internet obtained via a network is used. The operation history records and saves the operation history of the user with respect to the television receiver and the recording device. Furthermore, language processing such as morphological analysis, which is a known means, for text information such as program title, genre, performer name, and program outline description included in the program information, which constitutes the program information that has been viewed or recorded. Is used to extract words included in the program information, and the user's preference strength (preference level) is obtained based on the operation history for the extracted words. A method of calculating a user's preference for a word based on a frequency of a target word included in program information of a program viewed or recorded by the user, a ratio of viewing or recording time to a broadcast time of a program including the word in the program information, and the like There is. The user may input a word and its preference degree evaluation.
The operation history used when obtaining the preference degree may be handled as the history of all the users who have viewed or viewed as the history of the representative when viewed by a plurality of users. The representative inputs the user or refers to the existing history, and selects the user whose selected program is most suitable for the preference.

  The user profile management unit 105 manages and manages the profiles of a plurality of users generated by the user profiling unit 104.

  The program recommendation determining unit 106 manages, for the user set identified by the user identifying unit 103, the words included in the user profiles of the users constituting the user set and the degree of preference for the words, which is managed by the user profile managing unit 105. Whether or not to recommend a program to the user set is calculated by comparing the extracted program information of the program to be determined in the same manner as the user profiling unit 104 with the extracted word, and calculating the recommendation degree for the target program. judge.

  The user interface unit 107 notifies the user of the program determined as the recommended program by the program recommendation determination unit 106.

  FIG. 2 is a diagram illustrating an example of a user profile managed by the user profile management unit 105. The user profile 201 includes user information 202 and profile information 203 corresponding to the user for each user. The profile information 203 is configured as a set of words 204 and word preference levels 205.

  FIG. 3 is a diagram illustrating a processing method of the program recommendation determination unit. In step 301, the user set U identified by the user identifying unit 102 is acquired.

  A word that belongs to the user set U obtained in step 301 and refers to the profile of the user X with respect to the user X whose user profile is held by the user profile management unit 105, and whose preference degree 205 is a value that exceeds the threshold value Sth 204 is selected, and a high preference word set Kx of the user X is created (step 302). Next, in order to obtain a word having a high preference level for all users belonging to the user set U, a common high preference word set Ka, which is a common set of the high preference word sets Kx for all users X, created in step 302 is obtained. Extract (step 303). When Ka is an empty set, there is no high preference word common to all users, and therefore, a recommendation level based determination process 304 for determining a recommended program from the program recommendation level calculated based on each user preference is performed. If Ka is not an empty set, a common preference base determination process 305 is performed because there is a high preference word common to all users.

  FIG. 4 is a diagram showing a processing method of the recommendation degree base determination process 304. This process determines whether or not to recommend a program to the entire user set based on the recommendation level to individual users belonging to the user set (the degree of adaptation to the user preference of the program). The recommendation degree Tpx of the target user X is obtained for the program P that is the target of whether or not to recommend (step 401). The recommendation level Tpx refers to the profile of the user X, and is a value obtained by adding the preference levels of all users to the words included in the word set extracted from the program information of the program P, or all users included in the word set. This is calculated as the number of words having a high preference level. If the recommendation level Tpx of all target users X for the program P obtained in step 401 is larger than the threshold Tth, the program P is determined as a recommended program for all (step 402). The recommendation level determination process 304 is performed when there is no word having a high preference level common to all users, that is, when each user has various preferences. Therefore, the recommendation result of the common preference base determination process 305 that performs the recommendation determination based on the preference common to all users has a higher degree of match with the user preference. However, since a program having a recommendation level exceeding the threshold is recommended, there is no user who recommends a program that does not meet the preference at all, and program selection among a plurality of users can be performed with ease. If the recommendation level of all programs targeted for recommendation does not exceed the threshold value and the recommended program cannot be found, the threshold value is lowered or the recommendation level for a certain number of users in the user set exceeds the threshold value. A recommended program can be selected by setting a program exceeding the recommended program. In this case, the degree of coincidence of the recommended program with the user preference level decreases, but unexpectedness and new discovery can be provided to the user.

FIG. 5 is a diagram illustrating a processing method of the common preference base determination processing 305. This process determines whether or not to recommend a program based on preferences common to users belonging to the user set. A recommendation degree Tp for the program P is calculated using the words included in the common high preference word set Ka obtained in step 303 of FIG. Step 501). The recommendation degree Tp is the average value of the preference degree Spw of each user X with respect to the word w, or the number of words of the common high preference word set Ka included in the program information of the program P or the preference degree Sw of the word w included in Ka. And the sum of the word preference degrees Sw included in the program information of the program P can be calculated. When the recommendation level Tp for the program P obtained in step 501 is larger than the threshold value Tth, the program P is determined as a recommended program for all users (step 502).
If it is not determined that all the programs subject to recommendation determination are recommended programs, the threshold value Sth for extracting a high-preference word set in step 302 in FIG. 3 is decreased, or the threshold value Tth is decreased, Alternatively, the recommended program can be determined by the recommendation level determination process 304.

  Next, an embodiment that considers the priority for viewing between users will be described. FIG. 6 is a diagram illustrating an example of a program selection priority profile. The program selection priority profile 601 includes user set information 602, viewing time zone information 603, and priority information 604. The user set information 602 is information on who is watching at the same time. The viewing time zone information 603 is a time zone that the user is viewing, and includes information such as day of the week and time zone, for example. The priority information 604 includes user information 605 and user program selection priority 606. In the example of FIG. 6, when user A and user B are watching from 19:00 to 21:00 on Wednesday, the program selection priority of user A is 3.0, and the program selection of user B is It consists of information that the priority is 1.0. The program selection priority represents a rate at which each viewer selects a program in a situation where a plurality of users are viewing the program. For example, the first line in FIG. 6 shows that the number of times that user A selects a program is three times the number of times user B selects a program in the situation where users A and B are watching the program. ing. Which user has selected the program can be determined by the user identification unit (103).

  Note that the priority time zone information 603 is not essential, and the program priority profile may be held by integrating all time zones without holding the program priority profile for each time zone.

  FIG. 7 is a diagram showing a processing method of program selection priority profile generation processing. First, information on the user set U currently being viewed identified by the user identifying unit 102 is acquired (step 701). For all users X belonging to the user set U acquired in step 701, the recommendation level Tpx for the viewed program P is obtained (step 702). When the user U included in the user set U and the set Ux of the user X whose recommendation degree Tpx for the program P obtained in step 702 is larger than the threshold Tth is not empty, the value of the recommendation degree Tpx of the user X belonging to the set Ux is Update the program selection priority profile held in the user profile manager. When there is a program selection priority profile that uses the user set U as user set information, the priority of each user of the corresponding priority information is updated with the value of the recommendation level Tpx, and the user set U is used as user set information. If no program selection priority profile exists, a new program selection priority profile for the user set U is generated. The priority for each user of the program selection priority profile can be calculated by the sum of recommendation degrees. When a program recommendation for a user set is performed using the program selection priority profile, it is possible to perform a program recommendation that emphasizes the preference level of a user with a high priority.

  FIG. 8 is a diagram showing a processing method for calculating a threshold used for program recommendation determination processing using program priority information.

The user set U identified by the user identification unit 102 is acquired (step 801), and the current time Tc is acquired (step 802). The current time can be obtained from a real-time clock provided in the computer. The user selection priority profile corresponding to the user set U and time Tc is referred to, and the priority of each user belonging to the user set U is acquired (step 803). The threshold value of each user is calculated using the priority acquired in step 803 (step 804).
For example, if the threshold Tb of the user B with low priority is raised, the threshold Ta of the user A with high priority is lowered, and Ta <Tb, the user B is recommended only for programs with a priority higher than Tb. The ratio of recommending a program that suits the preference is lower than that of the user A to be recommended for a program with a higher recommendation level. Because the user's daily program selection status is reflected, recommendations that match the viewing habits can be obtained. Conversely, by lowering the threshold for users with low priority and increasing the threshold for users with high priority, it is possible to reverse the daily program selection priority and create fairness in program selection.
FIG. 9 is an example of a screen that presents a recommended program for the user set. The identified user name and recommended program information are displayed. The display makes it easy for the user to select a program.

  The present invention can be applied to home video equipment such as a DVD recorder and a TV, a personal computer with a TV function, an information equipment such as a PDA and a mobile phone, a set-top box for reproducing a net distribution program, and the like.

It is a figure explaining the structure of one Example of this invention. It is a figure explaining the example of a structure of a user profile. It is a figure explaining the processing system of a program recommendation determination process. It is a figure explaining the processing system of recommendation degree base determination processing. It is a figure explaining the processing method of a common preference base determination process. It is a figure explaining the structure of a program selection priority profile. It is a figure explaining the processing system of a program selection priority profile production | generation process. It is a figure explaining the processing method of the threshold value calculation using program priority information. It is a figure explaining the example of the screen which presents a recommendation program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Program recommendation apparatus, 102 Control part, 103 User identification part, 104 User profiling part, 105 User profile management part, 106 Recommended program determination part, 107 User interface part, 201 User profile, 202 User, 203 profile, 204 Word, 205 Priority, 601 Program selection priority profile, 602 User set, 603 Viewing time zone, 604 Priority information, 605 User, 606 Priority.

Claims (4)

A program recommendation device for recommending programs based on user preferences,
A user identification unit for identifying a user set of a plurality of users who are simultaneously viewing;
A user profile management unit for managing profile information of each of the plurality of users;
A recommended program determination unit that determines a recommended program using user information managed by the profile management unit,
The recommended program determination unit determines a recommended program for a user set identified by the user identification unit using profile information of a plurality of users managed by the user profile management unit. .   The recommended program determining unit determines whether or not the recommended program is determined by determining whether the program information includes a word having a preference level equal to or higher than a predetermined value for each of all users constituting the user set. The program recommendation device according to claim 1.   2. The program recommendation device according to claim 1, further comprising: a user profiling unit that generates a user profile managed by the user profile management unit by profiling an operation history of each user. The user profiling unit determines priorities among a plurality of users who are watching at the same time,
The program recommendation device according to claim 3, wherein the recommended program determination unit determines a recommended program using the priority information.

Images