Market analysis, demand forecasting or surveying

Method of and apparatus for optimal scheduling of television programming to maximize customer satisfaction

4745549

Abstract

In a method and apparatus for extracting programs suitable for individual subscriber taste from among all available television programs and for editing an individual subscriber television program list, objective data is statistically processed by linear programming. The processed results are input to a computer and are stored on a hard disk. The storage contents are read out from the hard disk and are printed out. Subscriber complaints about the program list are periodically fed back to improve prediction precision, thereby providing an automatic controller attuned to subscriber taste when the individual subscriber program list is used to automatically control a TV or VTR.


Claims

What is claimed is:

1. An apparatus for providing an individual television program table wherein television channels and corresponding time intervals are listed for each individual, comprising;

record reading means for reading subscriber taste data obtained from a questionnaire and in response developing subscriber taste data signals;

means for evaluating television programs in accordance with predetermined criteria to obtain program evaluation data;

linear programming means responsive to said subscriber taste data signals and said program evaluation data for generating optimal program list data;

printer means for printing from said optimal program list data a program list on a recording medium;

feedback means for inputting complaints on the program list from a subscriber and, in response, modifying said program evaluation data to improve prediction reliability of a subsequent program list to be printed by said printer; and

setting means for setting a television or video cassette recorder in accordance with the program list.

2. In a device for printing out an individual television program table, said table comprising a list in units of time intervals and channel numbers in a form optimal for each subscriber by selecting optimal programs from among all available television programs, a method of editing the program table, comprising the steps of;

evaluating questionnaires showing subscriber taste in television programs to obtain taste evaluation questionnaires;

evaluating television programs to obtain television program evaluations;

applying a linear program to the subscriber taste evaluations and television program evaluations to obtain optimal individual subscriber program data;

storing the optimal individual subscriber program data;

printing out the stored optimal individual subscriber program data in the form of an individual subscriber program list;

receiving complaints from a subscriber when the subscriber makes complaints on the printed program list;

feeding back the complaints to be included in the linear program to increase prediction reliability of a subsequent program list; and

reading the program list to automatically set a television set or video tape recorder.

3. In a device for printing out an individual television program table, said table comprising a list in units of time intervals and channel numbers in a form optimal for each subscriber by selecting optimal programs from among all available television programs, a method of editing the program table, comprising the steps of;

evaluating questionnaires showing subscriber taste in television programs to obtain taste evaluation questionnaires;

evaluating television programs to obtain television program evaluations;

applying a linear program to the subscriber taste evaluations and television program evaluations to obtain optimal individual subscriber program data;

storing the optimal individual subscriber program data;

printing out the stored optimal individual subscriber program data in the form of an individual subscriber program list;

receiving complaints from a subscriber when the subscriber makes complaints on the printed program list;

feeding back the complaints to be included in the linear program to increase prediction reliability of a subsequent program list; and

editing the printed individual subscriber program list to control a program rating of a television set or a video tape recorder.


Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to editing of an individual television program according to linear programming, a method of automatically controlling a program rating, and an apparatus therefor.

2. Description of the Prior Art

In order to implement such an apparatus according to conventional techniques, a program editor selects suitable programs reflecting a customer's taste from among all available television programs by utilizing subjective knowledge according to the results of questionnaires filled by end users (customers). In this case, the programs are selected on at least a once-per-week basis. The programs are listed and printed out in a tape-like table (to be referred to as a table hereinafter) according to the channel numbers and time intervals. The TV or VTR is then automatically set according to this table. Japanese Patent Publication Nos. 41-8542, 42-1882, 53-40340, and 59-21114 issued to the present applicant are concerned with the above conventional system.

SUMMARY OF THE INVENTION

It is an object of the present invention to edit an individual television program list which is formed on the basis of objective decision and which satisfies a customer.

It is another object of the present invention to provide a method and apparatus for editing individual television program lists in a short period of time, even if many customers simultaneously order the lists.

It is still another object of the present invention to eliminate subjective opinions of an editor when he selects programs suitable for the tastes of customers from among all available television programs.

It is still another object of the present invention to provide a method employing linear programming for questionnaires and program evaluation and adopting a technique for feeding back complaints against the evaluation results, thereby improving satisfaction of the customers in the individual television programs listed by the editor.

It is still another object of the present invention to edit a control tape for controlling an automatic controller incorporated in a TV or VTR according to the program list.

In order to achieve the above objects of the present invention, the present invention is constituted by a method comprising the steps of: statistically processing objective data according to linear programming; inputting processed results in a computer, storing them on a hard disk, and printing out the contents therein; and periodically feeding back complaints to be compared against the processing results to improve reliability of the data. The preferred embodiment also includes editing of a control tape. The functional system diagram is shown in FIG. 2.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a conventional system;

FIG. 2 is a block diagram of a system according to the present invention;

FIG. 3 is a diagram showing a computer program according to the present invention;

FIG. 4 is a table showing the relationship between time intervals, days, and television channel numbers according to the results of questionnaires;

FIGS. 5A and 5B are a flow chart for processing questionnaire information (subscriber data registration);

FIG. 6 is a flow chart of television program evaluations;

FIG. 7 is a flow chart for editing television programs according to linear programming;

FIG. 8 is a table showing an output of a television program table;

FIGS. 9A and 9B are a flow chart for processing a complaint; and

FIG. 10 is a diagram of components associated with a computer according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

The present invention is constituted by a means for evaluating questionnaires collected from end users, a means for evaluating all available television programs, a means for calculating the evaluation results according to linear programming and editing data (a television program list for time intervals and channel numbers in a form suitable for each customer), a means for inputting the data in a computer and causing the computer to calculate the data, a means for temporarily storing an output from the computer, a means for printing out each individual television program list (to be referred to as a program list hereinafter) according to the storage contents, a means for feeding back complains against the first printed-out program list, and a means for automatically controlling a TV or VTR according to the program list. The constitution of the present invention described above is expressed as a program diagram in FIG. 3.

Although hardware associated with the above operations is added to constitute the entire system, a control tape is finally edited to control an automatic controller incorporated in a TV or VTR. This is known in conventional systems to those skilled in the art, and since it is not directly associated with the present invention, a detailed description thereof will be omitted. The structures and operations of the means described above will be described below:

(1) Subscriber Questionnaire Result Input Means

FIG. 4 shows one form of questionnaire used as a questionnaire result input means for extracting and representing individual tastes. Referring to FIG. 1, the leftmost column represents days of the week, and the uppermost row represents time intervals (in units of 15 minutes). Television channel Nos. are respectively filled at intersections of corresponding rows and columns. A basic questionnaire for editing the list described above is exemplified below: 

    ______________________________________
    Subscriber Questionnaire
    ______________________________________
    Q1: Are you married or single?
    1.        Single
    2.        Married without children
    3.        Married with children in the age group
              of 0 to 6 years old
    4.        Married with children in the age group
              of 7 to 12 2 years old
    5.        Married with children in the age group
              of 13 years old or more
    Q2: What is your age group?
    6.        25 years old or less
    7.        35 years old or less
    8.        45 years old or less
    9.        55 years old or less
    10.       56 years old or more
    Q3: Do you Watch NHK (Nihon Hoso Kyokai) programs?
    11.       Never
    12.       Only news or morning programs
    13.       Only educational programs
    14.       The same amount as commercial television
              programs
    15.       Usually
    Q4: Do you watch political and economic programs?
    16.       Always
    17.       Usually
    18.       Often
    19.       Sometimes
    20.       Never
    Q5: Do you watch science programs?
    21.       Always
    22.       Usually
    23.       Often
    24.       Sometimes
    25.       Never
    Q6: Do you watch historical programs?
    26.       Always
    27.       Usually
    28.       Often
    29.       Sometimes
    30.       Never
    Q7: Do you watch documentary programs?
    31.       Always
    32.       Usually
    33.       Often
    34.       Sometimes
    35.       Never
    Q8: Do you watch news programs?
    36        Always
    37.       At least once a day
    38.       Not if a desired program is in the same
              time interval
    39.       Sometimes
    40.       Never
    Q9: Do you like performing arts programs?
    41.       Like very much
    42.       Like somewhat
    43.       Indifferent
    44        Dislike somewhat
    45.       Dislike very much
    Q10: Do you like sports programs?
    46.       Like very much
    47.       Like somewhat
    48.       Indifferent
    49.       Dislike somewhat
    50.       Dislike very much
    Q11: Do you like "variety talk show" programs?
    51.       Like very much
    52.       Like somewhat
    53.       Indifferent
    54.       Dislike somewhat
    55.       Dislike very much
    Q12: Do you like quiz show programs?
    56.       Like very much
    57.       Like somewhat
    58.       Indifferent
    59.       Dislike somewhat
    60.       Dislike very much
    Q13: Do you like variety show programs?
    61.       Like very much
    62.       Like somewhat
    63.       Indifferent
    64.       Dislike somewhat
    65.       Dislike very much
    Q14: Do you like "rock`n` roll" and "foreign pops"
    musical programs?
    66.       Like very much
    67.       Like somewhat
    68        Indifferent
    69.       Dslike somewhat
    70.       Dislike very much
    Q15: Do you like "folk song" and "contemporary
    singer/song writers or equivalent" musical programs?
    71.       Like very much
    72.       Like somewhat
    73.       Indifferent
    74.       Dislike somewhat
    75.       Dislike very much
    Q16: Do you like "classical" musical programs?
    76.       Like very much
    77.       Like somewhat
    78.       Indifferent
    79.       Dislike somewhat
    80.       Dislike very much
    Q17: Do you like Samurai programs?
    81.       Like very much
    82.       Like somewhat
    83.       Indifferent
    84.       Dislike somewhat
    85.       Dislike very much
    Q18: Do you like "home drama" and comedy programs?
    86        Like very much
    87.       Like somewhat
    88.       Indifferent
    89.       Dislike somewhat
    90.       Dislike very much
    Q19: Do you like suspense and action dramas?
    91.       Like very much
    92.       Like somewhat
    93.       Indifferent
    94.       Dislike somewhat
    95.       Dislike very much
    Q20: Do you like foreign movies in Japan?
    96.        Like very much
    97.       Like somewhat
    98.       Indifferent
    99.       Dislike somewhat
    100.      Dislike very much
    ______________________________________


In the above questionnaire, five selection items are provided for each of 20 questions, so that a total of 20.times.5=100 items are provided. The 20 questions include questions on family members, age groups, time intervals in which customers do not watch television, and favorite dramas, as well as questions for determining whether the customers prefer NHK programs and news programs, and which program is the most interesting. Each question has five possible selection items. The items for linear programming (to be described later) are assigned with numbers as follows. The first question is assigned with P(1) to P(5); the second question with P(6) to P(10); the third question with P(11) to P(15); . . . ; and the twentieth question with P(96) to P(100). The flow chart of the subscriber data registration/retrieval (correction) is shown in FIG. 5.

(2) TV Program Evaluation Means

A TV program evaluation means has the same form as that of the questionnaire described in (1) Subscriber Questionnaire Result Input Means, and is associated with the contents of questions asked of the subscribers for respective TV programs. 

    ______________________________________
    TV Program Evaluation
    ______________________________________
    E1: Target age group of the TV program
    1.           Children of 6 years old or less
    2.           Children of 7 to 12 years old
    3.           Young people
    4.           Middle-aged people
    5.           The Elderly
    E2: Political and economic factor
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E3: Scientific factor
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E4: Historical and educational factor
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E5: Documentary factor
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E6: Factor of news or other information source
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E7: Factor of show business and gossip
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E8: Factor of sports program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E9: Factor of variety talk show
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E10: Factor of quiz show program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E11: Factor of variety show
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E12: Factor of rock`n` roll and pops program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E13: Factor of Japanese pops program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E14: Factor of classical music program
    1            Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E15: Factor of Samurai program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E16: Factor of home drama and comedy program
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E17: Factor of suspense and action drama
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    E18: Factor of foreign movie in Japan
    1.           Strong
    2.           Moderate
    3.           Slight
    4.           Very slight
    5.           None
    ______________________________________


The data registration for TV programs is given in the flow chart of FIG. 6.

(3) Linear Programming Means

The basic algorithm for linear-programming the evaluations in the above (1) and (2) will be described. The system of the present invention is based on the following algorithm: ##EQU1## where Pl(i): the degree of interest of one subscriber for the ith program (the degree is represented by a numeral; the degree given by numeral "0" represents no interest).

i: the ith program when the 21 hours from 5 a.m. to 2 a.m. are divided into 15-minute intervals.

P(j): representing that the subscriber circles the jth items in the questionnaire. If the subscriber circles item 3 in question 3, P(11)=P(12)=P(14)=P(15)=0 and P(13)=1.

n: 20 questions.times.5 selection items=100

T(i,j): representing that the jth item of the ith TV program is circled. i and j are given by 1 or 0.

Pc(j): an unknown coefficient for 0.ltoreq.Pc(j) (having nothing to do with the subscribers)

Equation (1) is a mathematical expression for evaluating a degree of commonness between the subscriber preferences and the TV programs. Coefficient Pc(j) is commonly given for all subscribers. The value of Pc(j) is accurately determined by the feedback of complaints from the subscribers.

In order to evaluate commonness between the subscriber preferences and the TV programs, linear programming is represented by an equation of the first degree. The coefficient Pc(j) is a common value for all subscribers. The value of Pc(j) is accurately calculated by feeding back (i.e., learning) complaints from the subscribers, and a detail description thereof will be made later.

(4) Data Storage Means and Printout Means

A data storage means according to the present invention is exemplified by a hard disk. Data input to the hard disk is immediately printed out, as shown in the flow chart of FIG. 7. If data for several thousands of subscribers is stored in the hard disk and code numbers are respectively assigned to the individual subscribers, an optical program table for each subscriber can be printed out, as shown in FIG. 8. According to a test, it took about 20 minutes to print out each program table after evaluation of the questionnaire if program language FORTRAN was used. The term "print out" does not mean that the program table is finally presented to the corresponding subscriber but that the table is confirmed in processing. The printed program table must be tested. By collecting complaints and feeding them back, a more complete program table can be prepared.

(5) Complaint Processing

As described last in the basic algorithm for the above data processing, the value of Pc(j) is accurately calculated by collecting complaints from the subscribers. An example of complaint reception is given as follows:

Calculation results are given for a given subscriber: ##EQU2## where TBS and NHK are Japanese TV broadcasting stations. A proposal for programming station TBS for Tuesday 19:30 is made. Assume that the given subscriber presents a complaint to this proposal in the following manner. ##EQU3## In this case, a difference between the degree of interest for Pl(i2) and that for Pl(i1) is represented by Yi. In other words, this is associated with an evaluation of the degree of importance of the complaint. At present, Yi=-1. In this case,

Pl(i2)-{Pl(i1)+Yi}.ltoreq.0

Substitution of equation (1) into the above inequality yields the following inequality: ##EQU4## Assuming that ##EQU5## the above inequality can be rewritten as: ##EQU6## The left-hand side is then substituted by Vi, so that: ##EQU7## Xj is determined to minimize Vi. Assuming only a sum V of positive values Vi (programs causing complaints), ##EQU8## where m is the total number of Vi components for Vi>0, as many as complaints possible must be received. A minimum V is then calculated to determine the accurate Xj. The above operation is given by the flow chart in FIG. 9.

Referring to FIG. 9, "Search Record" in Processing=1 means search of data for one subscriber. "Input Items Excluding Keyword" indicates the selection items excluding the subscriber name read in katakana characters and the date of birth. "New Input" indicates a new subscriber to be registered. However, if the corresponding record is found, i.e., YES, the correct input is entered to update the corresponding value. The flow then advances by selecting an alternative step. In short, in complaint processing, a series of steps from "Initialize Linear Programming" to "Find Countermeasures" are important in Processing=2. More specifically, linear programming initialization is a recalculation of Pc(j). The countermeasures indicate that a Pc(j) value different from the current value is calculated and updates the current value. These mathematical steps are the center of complaint processing, i.e., the learning function. The coefficient Pc(j) can be more accurate to improve prediction precision. Therefore, more suitable programs can be provided to the subscribers.

The overall system is shown in FIG. 10; identification of components is in Appendix I. The steps in the program can be represented by Appendix II.

The embodiment described above exemplifies TV program ratings. However, the method and apparatus of the present invention is not limited to such a particular application. Evaluations can be made according to questionnaires similar to that in the above embodiment. The results are linear-programmed to collect the individual complaints and then to feed them back for processing, thereby further improving prediction precision. In this manner, the present invention can also be applied to surveys other than the embodiment described in this specification. ##SPC1##

«Previous Next»
Method and system for effecting sensory evaluation of a smoking product System and method for forecasting bank traffic and scheduling work assignments for bank personnel