KR20050016895A - Method, system and program product for locally analyzing viewing behavior - Google Patents

Abstract

The present invention provides methods, systems, and program products for locally analyzing (television) viewing habits. In particular under the present invention, a single time interval of the watched program is fragmented into multiple time windows of the watched program. Then, for each program in each time window, conditional probabilities are calculated. The conditional probability is then compared to the noise threshold to determine the recommended program for each time window. The recommended program can be added to the user profile and / or output to the viewer.

Description

METHOOD, SYSTEM AND PROGRAM PRODUCT FOR LOCALLY ANALYZING VIEWING BEHAVIOR}

The present invention generally relates to methods, systems and program products for locally analyzing viewing habits. In particular, the present invention allows a single time interval of television viewing habits to be analyzed in a smaller time window so that accurate viewing recommendations can be made.

As the use of cable and satellite television has increased, television networks have increasingly provided viewers with an excess of programming. Such programs not only overwhelm television viewers, but also make it difficult for the network to analyze viewing habits (e.g., to determine what programs are likely to be watched). Also, with the development of consumer electronic devices such as set-top boxes and hard disk recorders, television viewers have more functionality. For example, many devices now allow viewers to set up user profiles, from which viewer recommendations can be made. In addition, many devices allow tracking of programs and / or program types being viewed. This type of information is commonly called viewing history / behavior and is particularly useful for television networks.

So far, viewing behavior has been analyzed on an overall basis. In particular, the programs and / or program types that have been watched over a single time interval (eg, 12 months) are identified. Once confirmed, the viewing frequency of each program is calculated. Based on this frequency, viewing preferences can be determined.

However, there are some problems in globally analyzing viewing habits. In particular, many viewers are not watching programs with the same consistency over different time periods. For example, suppose a global single time interval is 12 months. Also, consider that "baseball" related programs represent 97% of the programs watched during October, but only account for 3% of the programs watched over the entire 12 months (i.e., only for a particular viewer). Unfortunately, under global analysis, no attention will be paid to the high viewing frequency during October. Rather, an overall “global” percentage of 3% will be analyzed to determine the viewer's viewing preference. Thus, global analysis of viewing habits can be an inaccurate way to measure viewer preferences.

In view of the above, there is a need for a method, system, and program product for locally analyzing viewing habits. There is also a need for a single time interval to be chunked into multiple time windows of a program. There is also a need for conditional probabilities that each program in each time window will be calculated. In addition, there is a need for noise thresholds to be applied locally (eg, for each conditional probability) so that a particular program can be recommended with respect to each time window.

1 illustrates a recommendation system having an analysis system in accordance with the present invention.

2A shows a single time interval of a program watched according to the system.

2b illustrates a time window of a watched program in accordance with the present invention;

3 shows a method flow diagram in accordance with the present invention.

The present invention generally provides methods, systems and program products for locally analyzing viewing habits. In particular under the present invention, a single time interval of the watched program is fragmented into multiple time windows of the watched program. Then, for each program in each time window, conditional probabilities are calculated. The conditional probability is then compared with the noise threshold to determine the recommended program for each time window. Recommended programs may be added to the user profile and / or output to the viewer.

According to a first aspect of the present invention, a method of locally analyzing viewing habits is provided. The method includes (1) fragmenting a single time interval of the watched program into a plurality of time windows of the watched program; (2) calculating conditional probabilities for each of the watched programs in the plurality of time windows; And (3) comparing the noise threshold with conditional probabilities to determine the recommended program.

According to a second aspect of the present invention, a method of locally analyzing viewing habits is provided. The method includes the steps of (1) providing a single time interval of the watched program; (2) fragmenting a single time interval into a plurality of time windows of the watched program; (3) calculating conditional probabilities for each watched program in each of the plurality of time windows; And (4) locally applying a noise threshold to each of the watched programs to determine a recommended program for each of the plurality of time windows, wherein the calculated conditional probability for the particular watched program of the particular time window is specified. The time threshold must be at least equal to the noise threshold for the particular program that will be the recommended program.

According to a third aspect of the present invention, a system for locally analyzing viewing habits is provided. The system includes (1) a fragmentation system for fragmenting a single time interval of a watched program into a plurality of time windows of the watched program; (2) a probability system for calculating conditional probabilities for each watched program in the plurality of time windows; And (3) a threshold system that compares conditional probabilities and noise thresholds to determine a recommended program.

According to a fourth aspect of the present invention, there is provided a program product stored on a recordable medium for locally analyzing viewing habits. When executed, the program product includes (1) program code for fragmenting a single time interval of the watched program into a plurality of time windows of the watched program; (2) program code for calculating a conditional probability for each watched program of the plurality of time windows; And (3) program code for comparing the conditional probability and the noise threshold to determine a recommended program.

Therefore, the present invention provides a method, system and program product for locally analyzing viewing habits.

These and other features of the present invention will be more readily understood from the following detailed description of various aspects of the invention, taken in conjunction with the accompanying drawings.

The drawings are only schematic representations and are not intended to depict particular parameters of the invention. The drawings are only intended to depict exemplary embodiments of the invention and therefore should not be considered as limiting the scope of the invention. In the figures, like numerals represent like elements.

The present invention generally provides methods, systems and program products for locally analyzing viewing habits. In particular, a single time interval of a program watched under the present invention is fragmented into multiple time windows of the watched program. For each watched program in each time window, conditional probabilities are calculated. The conditional probabilities are then compared to the noise threshold to determine the recommended program for each time window. The recommended program can be added to the user profile and / or output to the viewer.

As used herein, it should be understood that the term "program" may refer to a particular program (eg, LAW AND ORDER) or the type / genre of the program (eg, crime drama). In this regard, the teachings described herein are not intended to be limited to one particular interpretation of the term "program."

Referring now to FIG. 1, an exemplary recommendation system 10 is shown. In general, the recommendation system 10 may be any computerized system that is capable of receiving the viewing habits of the user / viewer 36 and recommending the program 42 based on its local analysis. To this extent, the recommendation system 10 may be implemented as and / or within a set top box or other consumer electronic device (eg, a hard disk recorder, etc.). It is also to be understood that the term "viewing habit" as used herein is intended to refer to the program 40 (ie, the type of a particular show or program) watched by the viewer 36. As depicted, the recommendation system 10 generally includes a central processing unit (CPU) 12, memory 14, bus 16, input / output (I / O) interface 18, external device / resource 20 and database 22. The CPU 12 may include a single processing unit or may be distributed through one or more processing units at one or more locations on the client and server, for example. The memory 14 may be any known type of data storage device and / or transmission medium, including magnetic media, optical media, random access storage (RAM), read-only storage (ROM), data caches, data objects, and the like. It may include. Also similar to the CPU 12, the memory 14 may reside in one physical location including one or more types of data storage devices or may be distributed through multiple physical systems in various forms.

I / O interface 18 may include any system that exchanges information to / from an external source. The external device / resource 20 may include external devices including speakers, CRTs, LED screens, hand-held devices, keyboards, mice, voice recognition systems, voice output systems, printers, monitors, fax machines, pagers, and the like. It can include any known type. The bus 16 provides a communication link between each component in the recommendation system 10 and may likewise include any known type of transmission link including electrical, optical, wireless, and the like. Also, although not shown, additional components, such as cache memory, communication systems, system software, and the like, may be incorporated into the recommendation system 10.

The database 22 provides a storage device for the information needed to carry out the invention. Such information may include, in particular, watched programs, recommended programs, user profiles, noise thresholds, and the like. As such, database 22 may include one or more storage devices such as magnetic disk drives or optical disk drives. In another embodiment, the database 22 includes data distributed over a local area network (LAN), wide area network (WAN), or storage area network (SAN) (not shown), for example. Database 22 may also be configured in a manner that one of ordinary skill in the art can interpret as including one or more storage devices.

Stored in the memory 14 of the recommendation system 10 is the analysis system 24 (shown as a program product). As depicted, analysis system 24 includes a fragmentation system 26, a probability system 28, a threshold system 30, a profile system 32, and an output system 34. Under the present invention, the fragmentation system 26 will fragment a single time interval of viewing behavior (ie, watched program) into multiple time windows of the watched program. In particular, referring to FIG. 2A, a single time interval 50 (shown in the show / program type) of the watched program 52 is depicted. Under previous systems, viewing habits were analyzed globally (ie over the entire interval). In the example shown, the single time interval is from January to March. During this three month period, viewer 36 watched a total of 80 programs 54 and is classified as shown. However, as mentioned above, such global analysis is not always accurate because the viewing habits can change rapidly over time. For example, during a time interval 50 the viewer watched two opera related programs. Such viewing habits represent only 2/80 or 2.50% of the total watched programs. Because this percentage is so low, it is extremely unlikely that opera-related programs will be recommended to viewers in the future. This ignores the fact that opera-related programs can actually represent 100% of all watched programs during January, and by themselves may be worth recommending in the future.

To eliminate such mistakes, fragmentation system 26 will “fragment” or split time interval 50 into smaller time windows, as shown in FIG. 2B. In particular, the three month time interval 50 is fragmented into three time windows 60A to 60C of programs 62A to 62C, with each window 60A to 60C representing a month's time. As depicted, viewer 36 watched 30 sitcom programs during the January time window 60A (eg, 10 FRASIER, 8 SEINFELD, and 12 DARMA & GREG). During the time window 60B in February, viewer 36 watched a total of 15 programs 64B, including one baseball program, ten basketball programs, and four sitcom programs. In addition, during the time window 60C in March, viewer 36 watched a total of 35 programs 64C, including 12 drama programs, 10 sitcom programs, 11 basketball programs, and 2 opera programs. By fragmenting a single time interval 50 into a smaller time window, more accurate viewing habits can be determined.

It should be understood that the fragmentation system 26 may be programmed to fragment any single time interval into any number of time windows in any manner. For example, the time interval 50 may be fragmented into several weekly windows (as opposed to monthly windows).

Once time interval 50 is fragmented into smaller time windows 60A-60C, probability system 28 (FIG. 1) relates to each program 62A-62C in each time window 60A-60C. Will determine the conditional probability. As used herein, conditional probability refers to the percentage of time a particular program was watched during a particular time window 60A, 60B, or 60C. In particular, to calculate the conditional probabilities for a particular program, the amount of time Qp the program was watched must be divided by the total amount Qt of the program watched during each time window 60A-60C (Qp / Qt). For example, the conditional probability for a basketball program is 0/30 or 0.00 during the time window 60A in January and the conditional probability for the basketball program is 10/15 or 66.6% during the time window 60B in February. The conditional probability for the basketball program during the time window 60C in March, 11/35 or 31.4%. Thus, basketball-related programs are worth recommending to viewers 36 during February and March.

In determining what program should be recommended to viewer 36, threshold system 30 will apply a local noise threshold to determine a recommendation based thereon. In particular, the noise threshold will apply to the conditional probabilities of each program for a particular month. The noise threshold is usually some minimum level at which conditional probabilities must be at least equal in order for the program associated with it to be recommended. For example, if the noise threshold is 4%, a basketball-related program would be recommended based on the viewing habits of February and March, as only these two windows 60B, 60C would have at least 4% (ie 66.6 each). Basketball conditional probabilities such as% and 31.4%). Conversely, basketball is below the noise threshold during time window 60A in January, representing 0% of the program watched.

It is to be understood that the 4% noise threshold used herein is exemplary only and any noise threshold may be implemented. In addition, in recommending a program, any known algorithm can be implemented. For example, a recommendation can be made based on the analysis of the previous month. For example, a viewing recommendation during April may include a drama program, a sitcom program, a basketball program, and an opera program (ie, the conditional probability of an opera program during the March time window 60C is only 2/35 or 5.71). Because it is%}. Alternatively, recommendations can be made about the same time window for the next year. For example, a recommendation based on analysis of the March time window 60C may be made during March of the next year. In any case, the present invention analyzes locally as opposed to making viewing habits global.

If the conditional probability of the program is at least equal to the noise threshold, the program may be added to the user profile of the viewer 36 by the profile system 32. In particular, as discussed above, many consumer devices allow viewer 36 to set up a user profile for storage (eg, in database 22). Such a profile may refer to personal information such as the viewer's name, age, and may refer to programming information such as the viewer 36's preferred programs, actors, networks and / or genres. Under the present invention, the profile system 32 will update the user profile of the viewer 36 based on the locally analyzed viewing habits. This may be particularly useful if the viewer 36's preferences change but the user profile is not updated. For example, if viewer 36 never expressed a preference for a basketball-related program, but the current viewing behavior when analyzed locally indicates that preference, the viewer's user profile is automatically updated to indicate that preference. Can be done.

In any case, regardless of whether the user profile is updated, output system 34 will output any recommendation to viewer 36. As mentioned above, recommendations can be made in any known manner. To this extent, recommendations may have general or specific properties. For special properties, certain programs can be recommended. For example, if the viewer 36 has shown a strong tendency to watch basketball-related programs, a particular program may be output, such as "NBA Finals 7 at 7pm Saturday night broadcast on the XYZ network." To this extent, the recommendation can be made in the form of a display on the viewer's television screen or in any alternative manner.

As mentioned above, the present invention can be similarly applied regardless of whether the programs 62A to 62C are program types (as depicted in FIG. 2B) or special shows. If the program is a special show, recommendations based on the conditional probabilities of a particular show may be made with respect to the same show or similar shows. For example, if viewer 36 saw DARMA & GREG with a conditional probability of 50% during the March time window 60C, subsequent broadcasts of DARMA & GREG may be recommended. Alternatively, other sitcoms (eg FRASIER) may be recommended. The exact form of the recommendation is not intended to be limiting.

Referring now to FIG. 3, a method flow diagram 100 in accordance with the present invention is shown. As depicted, the first step 102 is to fragment a single time interval of the watched program into a plurality of time windows of the watched program. Once fragmented, the second step 104 is to determine the conditional probabilities for each watched program in each time window. The third step 106 is then to apply a noise threshold to each program within each time window to identify the recommended program.

It is understood that the present invention can be realized in hardware, software or a combination of hardware and software. Any kind of computer / server system (s)-or other apparatus adapted to perform the methods described herein-is suitable. A typical combination of hardware and software can be a general purpose computer system with a computer program, which controls the recommendation system 10 to perform the methods described herein when loaded and executed. Alternatively, a special purpose computer may be used that includes specialized hardware for carrying out one or more functional tasks of the present invention. The invention may also be embedded in a computer program product, which includes all the features that enable the implementation of the methods described herein, and when loaded into a computer system may perform these methods. A computer program, software program, program or software in this context may cause a system having an information processing capacity to directly or (a) convert to another language, code or notation and / or (b) reproduce in different material forms. Then, any expression in any language, code, or notation of a set of instructions intended to perform a particular function.

The foregoing description of the preferred embodiment of the present invention has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and obviously, many modifications and variations are possible. Such modifications and variations which are apparent to those skilled in the art are intended to be included within the relevant field.

The present invention is applicable to methods, systems, and program products for locally analyzing viewing behavior.

Claims (22)

As a method of locally analyzing the viewing habits, Fragmenting a single time interval of the watched program into a plurality of time windows of the watched program; Calculating conditional probabilities for each of the watched programs in the plurality of time windows; And Comparing each conditional probability and a noise threshold to determine a recommended program. The method of claim 1, further comprising providing a single time interval of the watched program prior to the fragmentation step. The method of claim 1, wherein each conditional probability is: Determining a program viewing amount for a program specifically watched during the particular time window; Determining a total amount of viewing for all watched programs during the particular time window; And And analyzing the viewing behavior, calculated by dividing the program viewing amount by the total viewing amount. The method of claim 1, wherein the watched program comprises a program type or a particular show. 2. The method of claim 1, wherein the conditional probability for a particular viewing program in a particular time window must be at least equal to a noise threshold for the particular program to be a recommended program for a particular time window. The method of claim 1, further comprising adding the recommended program to a user profile. The method of claim 1, further comprising outputting the recommended program. As a method of locally analyzing the viewing habits, Providing a single time interval of the watched program; Fragmenting the single time interval into a plurality of time windows of watched programs; Calculating conditional probabilities for each watched program in each of the plurality of time windows; And Locally applying a noise threshold to each conditional probability to determine a recommended program for each of the plurality of time windows, wherein the calculated conditional probability for a particular watched program for a particular time window is determined by the specific program. And the program must be at least equal to the noise threshold for the program to be a recommended program for a particular time window. 10. The method of claim 8, further comprising adding the recommended program to a user profile. 10. The method of claim 8, further comprising outputting the recommended program. 10. The method of claim 8, wherein said locally applying comprises locally analyzing a viewing habit, comparing noise thresholds and respective conditional probabilities to determine a recommended program for each of the plurality of time windows. How to. 9. The method of claim 8, wherein each conditional probability is: Determining an amount of program viewing for a particular viewing program during the particular time window; Determining a total amount of viewing for all watched programs during a particular time window; And And analyzing the viewing behavior, calculated by dividing the program viewing amount by the total viewing amount. As a system for analyzing viewing habits locally, A fragmentation system that fragments a single time interval of the watched program into a plurality of time windows of the watched program; A probability system for calculating conditional probabilities for each watched program of the plurality of time windows; And And a threshold system that compares the noise threshold with each of said conditional probabilities to determine a recommended program. The system of claim 13, wherein each conditional probability is based on an amount of time a particular program is watched during a particular time window and an amount of time all programs are watched during the particular time window. . 15. The method of claim 13, wherein the conditional probability for a particular viewing program during a particular time window is analyzed locally for viewing behavior that must be at least equal to a noise threshold for the particular program to be a recommended program for a particular time window. system. 14. The system of claim 13, further comprising a profile system for adding the recommended program to a user profile. 18. The system of claim 13, further comprising an output system for outputting the recommended program. A program product stored on a recordable medium for local analysis of viewing behavior, wherein the program product is executed when: Program code for fragmenting a single time interval of the watched program into a plurality of time windows of the watched program; Program code for calculating a conditional probability for each watched program of the plurality of time windows; And And program code for comparing each threshold conditional probability with a noise threshold to determine a recommended program. 19. The program product of claim 18, wherein each conditional probability is based on the amount of time a particular program is watched during a particular time window and the amount of time all programs are watched during the particular time window. 19. The program product of claim 18, wherein the conditional probability for a program specifically watched during a particular time window must be at least equal to the noise threshold for the particular program to be a recommended program for the particular time window. 19. The program product of claim 18, further comprising program code for adding the recommended program to a user profile. 19. The program product of claim 18, further comprising program code for outputting the recommended program.