JP4305862B2 - Program ranking apparatus, program ranking method, and program ranking program - Google Patents

Description

  The present invention relates to a television receiver, and more particularly to a program ranking apparatus, a program ranking method, and a program ranking program that rank a program that suits viewer's preference from a large number of programs.

  In recent years, with the start of BS broadcasting and CS broadcasting and the digitization of television programs, users can select programs to view from a huge number of programs. However, on the other hand, it has become difficult to comprehensively understand what programs are being broadcast, and it has become difficult for users to select programs that suit their preferences. In addition, there is a greater possibility that the user will not notice that a program that suits his / her taste is being broadcast and will miss it.

  Therefore, for the purpose of assisting the user in selecting a program that suits his taste from a huge number of programs, the broadcasting station has the program title, broadcasting time, channel, genre, description, cast list, etc. The technology that multiplexes and transmits the program information including the above items to the video / audio signal has been put into practical use and is in operation.

Some television receivers have a function of assisting a user in selecting a program by, for example, configuring an electronic program guide using program information and presenting it on a screen.
As a more advanced technology of the receiver, the viewing history including the program information of the program viewed by the user is stored in the receiver, and the user's preference for the contents of the program is analyzed in advance. In addition, at the time of selecting a program, program information of a program scheduled to be broadcast in the future is acquired, and a preference suitability level indicating how much the user is interested in the program is obtained for each program. There is a patent proposal for an automatic program selection method that automatically selects a program that the user is interested in.

Japanese Patent No. 3340342 provides a television channel selection device that analyzes user preferences and automatically selects and displays programs that match the user preferences.
Japanese Patent No. 3340342

  In the television channel selection apparatus according to Japanese Patent No. 3340342, keywords included in the text data of each item such as program title, genre, performer list, contents description, and the like included in the program information are extracted as attribute values. The attribute value is used as an analysis element for preference analysis. However, since this method handles all extracted attribute values equally regardless of the type of item, there is a problem that the characteristics of each item and the user's preference tendency for the item are not reflected in the calculation of the preference preference degree. .

  For example, it is assumed that a user who is interested in the news program in general regardless of the content or the performer watches a news program in which “Talent Hanako” happens to be a guest. In this case, the attribute value of the genre item, “News Program”, and the attribute value of the performer list item, “Talent Hanako”, are treated equally as analysis elements of the preference analysis. Therefore, in the subsequent program selection, there is a high possibility that not only the news program that the user is interested in but also the variety program in which the “talent Hanako” that is not of interest is appearing is selected. turn into.

As described above, the television channel selecting apparatus disclosed in Japanese Patent No. 3340342 has a problem in that accuracy of preference analysis cannot be improved and accuracy of automatic program selection cannot be improved.
Therefore, the present invention divides the preference suitability for a program into items such as a genre and a performer list, and optimizes the degree to which the preference suitability for each item is reflected in the suitability suitability of the entire program based on the viewing history. Thus, it is an object to improve the accuracy of preference analysis and improve the accuracy of program ranking for automatic program selection.

Therefore, in order to solve the above problems, the present invention provides the following apparatus, method, and program.
(1) The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to determine each program scheduled to be broadcast in the future by the user. In a program ranking device that ranks in descending order of degree of suitability for preference,
(111) first storage means for storing a viewing history of a program viewed by the user;
(104) second storage means for acquiring and storing the program information associated with the program;
Viewing frequency calculation means for obtaining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means (202, S402);
At the time of program ranking, using the viewing history and the program information, preference preference degree calculation means for obtaining one or more types of preference preference degrees indicating how much each program fits the user's preference And (203, S302, 205, S502),
Of the two programs scheduled to be broadcast in the future to be used when ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation means at the time of past program ranking and the viewing history A determination criterion setting means for setting a determination criterion for determining which one suits the user's preference, and executing the analysis process for each reset instruction to reset the determination criterion (204, S602) ,
Program ranking means for ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria (206, S703),
Have
The determination criterion setting means sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards It is a criterion setting means for,
The program ranking means extracts any two programs within a predetermined scheduled broadcast period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times each program within the predetermined broadcast schedule period is determined to be suitable for the user's preference, and rank the programs in descending order. A program ranking apparatus characterized by the above.
(2) Analyzing the user's preference from the viewing history of the program accompanied by the program information, and using the program information of each program scheduled to be broadcast in the future, In a program ranking device that ranks in descending order of degree of suitability for preference,
The program information has a plurality of items, each of which is composed of one or more attribute values,
(111) first storage means for storing a viewing history of a program viewed by the user;
(104) second storage means for acquiring and storing the program information associated with the program;
Viewing frequency calculation means for obtaining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means (202, S402);
Preference analysis means for calculating the appearance frequency of the attribute value included in the program information accompanying each program in the viewing history for each calculation instruction and obtaining the preference level for each attribute value (203, S302);
At the time of program ranking, the preference degree calculation means for adding the preference degree corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future for each item of the program information and obtaining the preference degree of preference for each program And (205, S502),
Of the two programs scheduled to be broadcast in the future to be used when ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation means at the time of past program ranking and the viewing history A determination criterion setting means for setting a determination criterion for determining which one suits the user's preference, and executing the analysis process for each reset instruction to reset the determination criterion (204, S602) ,
Program ranking means for ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria (206, S703),
Have
The determination criterion setting means sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards It is a criterion setting means for,
The program ranking means extracts any two programs within a predetermined scheduled broadcast period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times each program within the predetermined broadcast schedule period is determined to be suitable for the user's preference, and rank the programs in descending order. A program ranking apparatus characterized by the above.
(3) Analyzing the user's preference from the viewing history of the program accompanied by the program information, and using the program information of each program scheduled to be broadcast in the future, In the program ranking method in the program ranking device that ranks in descending order of the degree of suitability for preference,
The program ranking device comprises:
A first storage step of storing a viewing history of a program viewed by the user;
A second storage step of acquiring and storing the program information associated with the program;
A viewing frequency calculation step for obtaining a viewing frequency of each program corresponding to the program information stored in the second storage step based on the viewing history stored in the first storage step;
At the time of program ranking, using the viewing history and the program information, preference preference degree calculation step for obtaining one or more types of preference preference degrees indicating how much each program is adapted to the user's preference When,
Of the two programs scheduled to be broadcasted in the future to be used for ranking the programs by analyzing the preference suitability of each program obtained in the preference suitability calculation step at the time of ranking of past programs and the viewing history A determination criterion setting step for setting a determination criterion for determining which is suitable for the user's preference, and performing the analysis process for each reset instruction to reset the determination criterion;
A program ranking step of ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
It is characterized by performing
In the determination criterion setting step, one of the previously viewed programs included in the viewing history is set as a viewed program, and the preference suitability and the viewing frequency of the viewed program are set as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Standards A criterion setting step of setting,
In the program ranking step, any two programs within a predetermined broadcast scheduled period are extracted from program information of each program scheduled to be broadcast in the future stored in the second storage step, and the programs of the two programs are extracted. The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. And counting the number of times determined to match the user's preference for each program within the predetermined broadcast scheduled period based on the determination result, and ranking the programs in descending order of the number of times. A program ranking method characterized by the above.
(4) Analyzing the user's preference from the viewing history of the program accompanied by the program information, and using the program information of each program scheduled to be broadcast in the future, In the program ranking method in the program ranking device that ranks in descending order of the degree of suitability for preference,
The program information has a plurality of items, each of which is composed of one or more attribute values,
The program ranking device comprises:
A first storage step of storing a viewing history of a program viewed by the user;
A second storage step of acquiring and storing the program information associated with the program;
A viewing frequency calculation step for obtaining a viewing frequency of each program corresponding to the program information stored in the second storage step based on the viewing history stored in the first storage step;
A preference analysis step of calculating an appearance frequency of an attribute value included in program information accompanying each program in the viewing history for each calculation instruction, and obtaining a preference level for each attribute value;
At the time of program ranking, the preference degree calculation step of adding the preference degree corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future for each item of the program information to obtain the preference degree of preference for each program When,
Of the two programs scheduled to be broadcasted in the future to be used for ranking the programs by analyzing the preference suitability of each program obtained in the preference suitability calculation step at the time of ranking of past programs and the viewing history A determination criterion setting step for setting a determination criterion for determining which is suitable for the user's preference, and performing the analysis process for each reset instruction to reset the determination criterion;
A program ranking step of ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
It is characterized by performing
In the determination criterion setting step, one of the previously viewed programs included in the viewing history is set as a viewed program, and the preference suitability and the viewing frequency of the viewed program are set as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations with programs that have been broadcast in the same time zone as the programs that have not been viewed, and the analysis process is executed for each reset instruction. Judgment criteria A criterion setting step of setting,
In the program ranking step, any two programs within a predetermined broadcast scheduled period are extracted from program information of each program scheduled to be broadcast in the future stored in the second storage step, and the programs of the two programs are extracted. The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. And counting the number of times determined to match the user's preference for each program within the predetermined broadcast scheduled period based on the determination result, and ranking the programs in descending order of the number of times. A program ranking method characterized by the above.
(5) Analyzing the user's preference from the viewing history of the program accompanied by the program information, and using the program information of each program scheduled to be broadcast in the future, In a program ranking program that ranks in order of the degree of suitability to preference,
A first storage function for storing a viewing history of a program viewed by a user in a first storage unit;
A second storage function for acquiring the program information accompanying the program and storing it in a second storage means;
A viewing frequency calculation function for determining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
A preference preference calculation function for obtaining one or more types of preference suitability indicating how much each program is adapted to the user's preference using the viewing history and the program information at the time of program ranking. When,
Of the two programs scheduled to be broadcast in the future used for ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation function in the past program ranking and the viewing history A determination criterion setting function for determining which is suitable for the user's preference, a determination criterion setting function for executing the analysis process for each reset instruction and resetting the determination criterion;
A program ranking function that ranks programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Is a program that causes a computer to execute
The judgment criterion setting function sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards Is a criterion setting function,
The program ranking function extracts any two programs within a predetermined broadcast schedule period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program more suitable for the user's preference from the preference preference degree and the viewing frequency by using the judgment criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times that each program within the predetermined broadcast scheduled period is determined to match the user's preference, and rank the programs in descending order. A program ranking program characterized by that.
(6) Analyzing the user's preference from the viewing history of the program accompanied by the program information, and using the program information of each program scheduled to be broadcast in the future, In a program ranking program that ranks in order of the degree of suitability to preference,
The program information has a plurality of items, each of which is composed of one or more attribute values,
A first storage function for storing a viewing history of a program viewed by a user in a first storage unit;
A second storage function for acquiring the program information accompanying the program and storing it in a second storage means;
A viewing frequency calculation function for determining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
A preference analysis function for calculating an appearance frequency of an attribute value included in program information associated with each program in the viewing history for each calculation instruction, and obtaining a preference level for each attribute value;
A preference degree calculation function that adds the preference level corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future at the time of program ranking and obtains the preference degree for each program. When,
Of the two programs scheduled to be broadcast in the future used for ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation function in the past program ranking and the viewing history A determination criterion setting function for determining which is suitable for the user's preference, a determination criterion setting function for executing the analysis process for each reset instruction and resetting the determination criterion;
A program ranking function that ranks programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Is a program that causes a computer to execute
The judgment criterion setting function sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations with programs that have been broadcast in the same time zone as the programs that have not been viewed, and the analysis process is executed for each reset instruction. Re-established criteria Is a criterion setting function,
The program ranking function extracts any two programs within a predetermined scheduled broadcast period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program more suitable for the user's preference from the preference preference degree and the viewing frequency by using the judgment criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times that each program within the predetermined broadcast scheduled period is determined to match the user's preference, and rank the programs in descending order. A program ranking program characterized by that.

As described above, the present invention has the following effects.
(Effect 1) In the program ranking apparatus, method, and program of the present invention, the preference suitability for each item such as the genre / performer list and the viewing history of the program broadcast in the past stored in the apparatus Based on the judgment result of program viewing that can be judged from the above, judgment criteria for each program scheduled to be broadcast in the future are set and updated.

  By ranking the programs according to this criterion, the preference suitability of the entire program is obtained from the preference suitability for each item according to the criteria optimized for the viewing tendency of the user, and reflected in the judgment of program ranking. To be able to.

  For example, if the user is watching many news programs in general that are distinguished from “report” by the attribute value of the genre item, regardless of the attribute value of the performer list or keyword item, the appearance by performing the neural network learning Regardless of the preference suitability of the item list and the keyword item, a criterion is formed such that the preference suitability of the entire program of the program having a high preference suitability of the genre item is high. For this reason, even a news program that does not include the attribute values of the interested performers list or keyword item ranks as much as the news program that contains the attribute values of the interested performer list or keyword items. The possibility of being attached becomes high.

Accordingly, it is possible to improve the accuracy of preference analysis and to improve the accuracy of program ranking for automatic program selection.
(Effect 2) In the further program ranking apparatus and method of the present invention, the preference suitability for each item such as the genre / performer list of the program broadcast in the past stored in the apparatus, the viewing frequency, Based on the judgment result of the program viewing that can be judged from the viewing history, a judgment standard for each program scheduled to be broadcast in the future is set and updated.

By ranking programs according to this criterion, in addition to Effect 1, the viewing frequency is also reflected in the determination of program ranking.
For example, when a user is watching a drama program decided every week, the viewing frequency of the drama program is reflected in the automatic program selection, so regardless of the performer list and keyword items included in the drama program. The possibility that the ranking of the drama program is high is increased.

  Accordingly, it is possible to further improve the accuracy of the preference analysis and obtain the effect of further improving the accuracy of program ranking for automatic program selection.

  Next, the best mode for carrying out the present invention will be described. In this embodiment, the program ranking apparatus of the present invention is used as an automatic program selection apparatus.

  In one embodiment of the present invention, a broadcast station multiplexes and transmits program information including a program title, a broadcast time, a channel, a genre, a description of contents, a performer list, etc. related to a program together with a video and audio signal, and receives it on the receiving side. A television broadcasting system is assumed in which a video / audio signal and program information are separated from a broadcast wave and a user can select a program to view using program information displayed on a receiver.

The “program” used in the following description is as follows.
For example, the first episode, the second episode,..., And the final episode broadcast on a fixed day of the week as in a series drama are treated as the same program. More specifically, assuming that there are all 11 episodes of the serial drama “Jiro Monogatari”, all 11 episodes are treated as the same program “Jiro Monogatari”. Of these programs, when the broadcast has already ended up to the second episode, the first episode and the second episode, which have ended, are referred to as “programs broadcast in the past” of the program “Jiro Monogatari”. Call. Then, the third and subsequent episodes that have not been broadcast yet and are scheduled to be broadcast from now on are called “programs scheduled to be broadcast in the future” of the program “Jiro Monogatari”.

  As another example, programs broadcast on the same channel on the same day every week with the same content are also treated as the same program. More specifically, assuming that there was a variety program called “Comedy Battle” broadcasted on 8 channels every Saturday at 8am, this “Comedy Battle” is the same for every “Smile Combat” broadcast on any Saturday. Treat as a program. And if today is Friday, of these programs, the “Comedy Battle” broadcast on Saturday before the last week that has already been broadcast at this time is a generic term for the “Comedy Battle” program that is broadcast in the past. It is called a “program”. The “comedy battle” scheduled to be broadcast on Saturday after tomorrow, which has not been broadcast yet, will be collectively referred to as the “program to be broadcast in the future” of the program “Comedy Battle”.

  Program information is information relating to a program created by a broadcasting station, and includes items such as program title, broadcast time, channel, genre, content description, performer list, and the like. The program ranking information is additional information related to the program generated by the program automatic selection unit, and includes keywords extracted from the program information, program viewing frequency, item-specific preference suitability, and the like.

Further, in this description, two types of items, “general value broadcast prescription items” and “attribute value receiver prescription items”, are defined uniquely. The definition contents are described below.
Commonly used program information items are the program title, broadcasting time, channel, genre, description of contents, performer list, etc., as described above, which are multiplexed with video and audio signals and transmitted. is there.

  The attribute value broadcast stipulated items are items in which the format of attribute values is determined, such as a genre / performer list, among the above-described program information items that are generally used. For example, the attribute value of the item of genre is described for each program, and “1” is assigned to “drama” and “2” is assigned to “sport”. In this description, specifically, “genre” and “performer list” are taken up as attribute value broadcast regulation items.

Attribute value receiver specified items are information items such as program titles / contents descriptions, etc., among the above-mentioned commonly used program information items, and have specific attribute values. It is an item newly created on the receiver side by extracting attribute values from certain items according to certain rules. In this description, an item named “keyword” having a keyword extracted from text data using morphological analysis as an attribute value is created.
[Overall configuration]
FIG. 1 is a block diagram showing an automatic program selection apparatus using an embodiment of the program ranking apparatus of the present invention. The program automatic selection apparatus of this embodiment has basic functions as a television reception apparatus such as reception, viewing, recording, and reproduction of television broadcasts, and further has an automatic program selection function.

  The broadcast receiving unit 101 includes a tuner and a demodulating / separating unit. When a channel to be received is instructed by the viewing control unit 102, a broadcast wave transmitted from a broadcasting station of the instructed channel via a terrestrial wave or a broadcasting satellite. Are separated into video and audio signals and program information. The video / audio signal is output to the recording control unit 103, and the program information is output to the program information storage unit 104.

  The program information storage unit 104 is a storage device such as a hard disk, and stores program information input from the broadcast receiving unit 101 and program ranking information input from the program automatic selection unit 105. The program information storage unit 104 has a function of reading and writing program information and program ranking information by the program automatic selection unit 105.

  The recording control unit 103 outputs the video / audio signal input from the broadcast receiving unit 101 to the output control unit 106, and stores the video / audio signal in the video / audio signal storage when the viewing control unit 102 instructs to record. Also output to the unit 107.

  The video / audio signal storage unit 107 is a storage device such as a hard disk, and stores the video / audio signal input from the recording control unit 103. In addition, the output control unit 106 has a function of reading a video / audio signal.

  The output control unit 106 outputs the video / audio signal input from the recording control unit 103 to the output unit 108 when the viewing control unit 102 instructs broadcast viewing, and the viewing control unit 102 plays back the recorded program. When instructed, the program specified by the viewing control unit 102 is read from the video / audio signal storage unit 107 and output to the output unit 108.

  The output unit 108 includes an image output device such as a cathode ray tube or a liquid crystal display and an audio output device such as a speaker, and reproduces a video / audio signal input from the output control unit 106. Also, an operation screen for recording reservation is displayed, a recorded program list stored in the video / audio signal storage unit 107 is displayed, or a recommended program list input from the program automatic selection unit 105 is displayed. It has a function.

  An input unit 109 is an input device including a remote controller and its light receiving unit, and the user selects a channel to be viewed, starts / stops recording of the program being viewed, reserves recording, plays / stops the recorded program, It has a function of instructing automatic selection, selecting a program to be reserved for recording from a recommended program list, and the like.

  When a channel to be viewed is instructed from the input unit 109, the viewing control unit 102 instructs the broadcast receiving unit 101 to receive the channel, and when the input unit 109 instructs to start / stop recording of the program being viewed, The recording control unit 103 is instructed to start / stop recording, and when recording reservation is instructed from the input unit 109, the recording management information (recording start / stop time, channel, etc.) is output to the recording management information storage unit 110 and input. When the playback / stop of the recorded program is instructed by the unit 109, the viewing control unit 102 is instructed to play / stop the recorded program accordingly. Further, all instructions input from the input unit 109 are output to the viewing history storage unit 111 as a viewing history. Further, the recording management information stored in the recording management information storage unit 110 is periodically inspected, and if it is time to start recording, the broadcast receiving unit 101 is instructed to receive the channel of the program to be recorded and the recording control is performed. The unit 103 is instructed to start recording, and the recording control unit 103 is instructed to stop recording if it is time to stop recording.

The recording management information storage unit 110 is a storage device such as a hard disk, and has a function of reading and writing recording management information by the viewing control unit 102.
The viewing history storage unit 111 is a storage device such as a hard disk, and stores the viewing history input from the viewing control unit 102. In addition, the automatic program selection unit 105 has a function of reading a viewing history. Further, it may have a function of automatically deleting the viewing history after a predetermined period.

  The automatic program selection unit 105 reads the program information stored in the program information storage unit 104 and writes the program ranking information, reads the viewing history stored in the viewing history storage unit 111, and the preference information storage unit 112. Has the function of reading and writing the preference information stored in the program and the criteria for judging the program ranking. When an automatic program selection is instructed from the input unit 109, a list of automatically selected programs is output to the output unit. Alternatively, if automatic program selection is set to start automatically at the specified time, the program is automatically selected at the specified time and recording management information is recorded as a recording reservation list. The information is output to the management information storage unit 110.

The preference information storage unit 112 is a storage device such as a hard disk, and has a function of reading and writing preference information and criteria for determining program ranking by the program automatic selection unit 105.
The program information storage unit 104, the video / audio signal storage unit 107, the recording management information storage unit 110, the viewing history storage unit 111, and the preference information storage unit 112 may be configured to share a storage device such as one hard disk. .
[Detailed configuration of program automatic selection section]
Next, the configuration and operation of the program automatic selection unit 105 will be described in more detail.

  FIG. 2 is a block diagram of the automatic program selection unit 105 in the automatic program selection apparatus using an embodiment of the program ranking apparatus of the present invention. FIG. 13 is a flowchart showing an overall flow of the program automatic selection process performed by the program automatic selection unit 105.

  The automatic program selection unit 105 includes a keyword extraction unit 201, a viewing frequency calculation unit 202, a preference analysis unit 203, a program ranking determination criterion learning unit 204, an item-specific preference suitable degree calculation unit 205, and a program selection unit 206. The

  When new program information is stored in the program information storage unit 104, the keyword extraction unit 201 reads the program information from the program information storage unit 104, and uses morphemes from text data of items such as a program title and content description included in the program information. A keyword is extracted using the analysis, and is written in the program information storage unit 104 as program ranking information.

  When new program information is stored in the program information storage unit 104, the program viewing frequency calculation unit 202 shows the program included in the viewing history read from the viewing history storage unit 111 and the viewing history read from the program information storage unit 104. Using the information, the viewing frequency of each newly stored program is obtained and written into the program information storage unit 104 as program ranking information.

  The preference analysis unit 203 refers to the clock 207, and when the preset time is reached, the viewing history read from the viewing history storage unit 111 and the program information and program ranking of the program to be viewed read from the program information storage unit 104 Based on the information, the preference level for each attribute value in each item such as genre, performer list, and keyword is obtained and written in the preference information storage unit 112. The preference analysis may be performed about once a day.

  The program ranking judgment criterion learning unit 204 refers to the clock 207, and when a preset time is reached, the program-specific preference suitability and program viewing frequency read from the program information storage unit 104, and a viewing history storage unit Using the viewing history read from 111, the program ranking judgment criteria for selecting a program are learned, and the obtained program ranking judgment criteria are written into the preference information storage unit 112. It should be noted that learning of the criteria for judging program ranking may be performed about once a day.

  The item-specific preference suitability calculation unit 205 reads the future from the program information storage unit 104 when automatic program selection is started by referring to an instruction from the input unit 109 or the time set in advance with reference to the clock 207. Based on the program information and program ranking information of the program scheduled to be broadcast and the preference value for each attribute value of each item read from the preference information storage unit 112, the preference preference for each program for all programs to be automatically selected is obtained. Then, the item-specific preference suitability is written in the program information storage unit 104 as program ranking information, and the item-specific preference suitability, program information, and program ranking information are output to the program selection unit 206.

The program selection unit 206 uses the item-specific preference suitability and the program ranking information when the item-specific preference suitability, program information, and program ranking information are input from the item-specific preference suitability calculation unit 205. Then, the program is selected according to the criteria for judging the program ranking read from the preference information storage unit 112, and the program information of the selected program is output to the output unit 108 or the recording management information storage unit 110 as a selected program list.
<Overall Flow of Program Automatic Selection Unit 105>
The operation of the program automatic selection unit 105 configured as described above will be described in time series with reference to the flowchart of FIG.

Program automatic selection processing is started (S101).
In step S102, it is confirmed whether or not the keyword of the program related to the viewing history is extracted by the keyword extraction unit 201. The program related to the viewing history refers to a program whose viewing history is stored in the viewing history storage unit 111 among the previously broadcasted programs whose program information is stored in the program information storage unit 104. And the programs of other channels that were broadcast at the same time as the respective programs. Program keywords related to viewing history should normally be extracted by the previous automatic program selection process, but for some reason, such as the first automatic program selection process or the previous automatic program selection process being turned off. In some cases, keywords are not extracted.

  When the keyword of the program related to the viewing history has not been extracted, the keyword extracting unit 201 performs keyword extraction processing (S103) on the program related to the viewing history.

Next, the preference analysis unit 203 performs an attribute value preference degree calculation process (S104).
In step S105, the viewing frequency calculation unit 202 confirms whether or not the viewing frequency of the program related to the viewing history has been calculated. If the viewing frequency of the program related to the viewing history has not been calculated, the viewing frequency calculation unit 202 performs a viewing frequency calculation process (S106) on the program related to the viewing history.

  In step S107, it is confirmed whether or not the item-specific preference suitability calculation unit 205 has calculated the item-specific preference suitability of the program related to the viewing history. When the item-specific preference suitability level of the program related to the viewing history is not calculated, the item-specific preference suitability degree calculation unit 205 performs item-specific preference suitability degree calculation processing (S108) for the program related to the viewing history. Do.

Next, the program ranking determination criterion learning unit 204 performs a program ranking determination criterion learning process (S109).
The processing so far is mainly processing for the viewing history, and may be performed at an arbitrary time about once a day.

  The subsequent processing is processing for a program scheduled to be broadcast in the future, and is executed when a user's instruction for automatic program selection execution is received, or at an arbitrary time that is about once a day. It is also possible to display the stored execution result only when the user instructs to automatically select and execute the program.

A keyword extraction process (S110) for extracting keywords from the program information of all the programs scheduled to be broadcast in the future stored in the program information storage unit 104 is performed.
Next, a viewing history of a program broadcast in the past corresponding to each program scheduled to be broadcast in the future, stored in the viewing history storage unit 111, and program information stored in the program information storage unit 104, Is used to perform viewing frequency calculation processing (S111) for calculating the viewing frequency of all programs scheduled to be broadcast in the future.

  Next, an item-specific preference suitability calculation process (S112) for calculating the item-specific preference suitability from the program information of all the programs scheduled to be broadcast in the program information storage unit 104 and the program ranking information is performed. Do.

Finally, a program automatic selection execution process (S113) for selecting a program that matches the user's preference from all the programs scheduled to be broadcast in the future stored in the program information storage unit 104 is performed.
Thus, the program automatic selection process ends (S114).
<Details of each component>
Next, the keyword extraction unit 201, the viewing frequency calculation unit 202, the preference analysis unit 203, the program ranking determination criterion learning unit 204, the item-specific preference suitability calculation unit 205, which are components of the program automatic selection unit 105, The operation of each of the program selection unit 206 will be described in detail.
<Details of Keyword Extraction Unit 201>
First, the flow of keyword extraction processing (S103) performed by the keyword extraction unit 201 will be described in time series using the flowchart of FIG.

The keyword extraction process is started (S201).
In the keyword extraction process (S202), the program information (S801) stored in the program information storage unit 104 is read, and the keyword (S203) is extracted from the text data of the program title and description included in the program information. The extracted keyword (S203) is used for the keyword (S205) in the storage process (S204) for use in the attribute value preference calculation analysis process (S104) and the item-specific preference suitability calculation process (S108 / S112). Is stored in the program information storage unit 104.

This completes the keyword extraction process (S206).
<Details of Program Viewing Frequency Calculation Unit 202>
Next, details of the program viewing frequency calculation process (S106 / S111) performed by the program viewing frequency calculation unit 202 will be described.

  There are daily broadcasts, daily broadcasts from Monday to Friday, weekly broadcasts, broadcasts only once, etc., but here, for programs broadcast every week, the frequency of viewing of the program by the user is detected. . Note that programs that are broadcast every day and programs that are broadcast every day from Monday to Friday can be regarded as being broadcast every week if viewed in weekly units, and are therefore subject to program viewing frequency calculation.

  In order to calculate the program viewing frequency, a function for discriminating whether or not the program p is watched for w weeks before the scheduled broadcast date of the program p is defined by the equation of FIG. However, it is determined whether or not the program p has been viewed for w weeks before based on the program information regarding the program p and the viewing history for w weeks before. More specifically, there is a history of viewing during broadcasting in the same channel and the same time zone as the program p of w weeks ago (with a width of about 1 hour before and after the broadcasting time of the program p), and the viewing When it can be determined that the program title of the program included in the history is the same as the program title of program p, it is determined that the same program was watched w weeks ago.

Next, an equation for obtaining the program viewing frequency P _f (p) of the program p is defined as shown in FIG. 4 using the equation of FIG. This formula is obtained by adding a cosine window function to the formula in FIG. 3 and normalizing it so that the minimum value is 0 and the maximum value is 1.

From this equation, the program viewing frequency P _f (p) of the program p is obtained as a real number between 0 and 1, and the larger the value, the higher the viewing frequency. Furthermore, by applying a window function by cosine, the value of the viewing frequency is strongly influenced as the history becomes more recent. However, Nw is an integer of 1 or more and is a constant determined by how many weeks ago the viewing history is used. For example, if a history up to 50 weeks ago (about 1 year ago) is used, Nw is 50.
<Flow of viewing frequency calculation process>
Next, the flow of the viewing frequency calculation process (S106 / S111) performed by the program viewing frequency calculation unit 202 will be described in time series with reference to FIG.

The viewing frequency calculation process is started (S401).
Using the program information (S801) stored in the program information storage unit 104 and the viewing history (S802) stored in the viewing history storage unit 111, the viewing frequency is calculated according to the formulas of FIG. 3 and FIG. A calculation process (S402) is performed, and the viewing frequency (S403) of the program broadcast in the past of the program title that can be determined to be the same as the program scheduled to be broadcast in the future is obtained. The obtained viewing frequency (S403) is stored in order to be used for attribute value preference calculation processing (S104), program ranking determination criterion learning processing (S109), and program automatic selection execution processing (S113). In S404), it is stored in the program information storage unit 104 as the viewing frequency (S405).

This completes the viewing frequency calculation process (S406).
<Details of the preference analysis unit 203>
Next, details of the user attribute value preference analysis process (S104) performed by the preference analysis unit 203 will be described.

  First, of all viewing histories, viewing histories that are likely to be “while viewing” are excluded. Programs that are subject to “while watching” tend to have higher program viewing frequency. Conversely, when a program with a low program viewing frequency is viewed, it is considered that the user has selected the program with interest. Therefore, the history of viewing a program during broadcasting is excluded from the preference analysis if the program viewing frequency of the corresponding program is greater than a certain threshold. The threshold used here is a real number between 0 and 1, and is set to an appropriate value such as 0.5. A program with high viewing frequency is a program that the user wants to habitually watch, and it is considered that the program should be selected especially when the recording of the automatically selected program is reserved for recording. Inclusion of the program viewing frequency in the information used for selection increases the probability that a program with a high viewing frequency is also selected.

Next, preference is analyzed using all the viewing histories remaining after eliminating the viewing histories that may be “while viewing”. The preference analysis is performed on receivers such as keywords included in program ranking information, attribute values of broadcast regulation items such as genres and performers lists included in program information, regarding programs included in each viewing history. The number of appearances of the attribute value of the prescribed item is obtained and normalized by each item such as genre, performer list, and keyword. Normalization is performed according to the equation shown in FIG. 9 when the number of appearances of the attribute value x is f (x). This equation obtains the value of P (x) by normalizing the value of f (x) from 0 to 1 using a logarithmic function said to be close to human sensitivity. However, fmax is the maximum value of the number of appearances regarding all the attribute values of each item including the attribute value x. A is an arbitrary constant and is determined in consideration of the curve shape of the logarithmic function in normalization. With this formula, the number of appearances of the attribute value x is normalized to a real number between 0 and 1. Let the normalized value be the preference level of the attribute value x. FIG. 10 shows an example of the attribute value preference for each item of program information.
<Flow of attribute value preference calculation processing>
Next, the flow of attribute value preference degree calculation processing (S104) performed by the preference analysis unit 203 will be described in time series with reference to FIG.

The attribute value preference calculation processing is started (S301).
Preference analysis processing (S302) using program information (S801), keyword (S205), viewing frequency (S405) stored in program information storage unit 104, and viewing history (S802) stored in viewing history storage unit 111 )I do.

  The preference analysis process (S302) receives each attribute value broadcast regulation item of the genre / performer list extracted from the program information (S801) and an attribute value named <keyword> having the keyword (S205) as an attribute value. The attribute value preference level (S303) for each item indicating how much each attribute value included in the machine definition item matches the user's preference is calculated using the formula shown in FIG.

  The program information (S801) used for the preference analysis process (S302) includes viewing history (S802) by various types of operations such as viewing a broadcast program, viewing a program recording reservation, viewing a recorded program, and the like. Contains program information. However, with respect to program viewing during broadcasting, since there are so-called “while watching” programs that are not particularly interesting to the user and watch without much attention, preference analysis processing is performed on all program information (S801). When used in (S302), there is a problem that the accuracy of preference analysis cannot be improved. In order to avoid this problem, with respect to program viewing during broadcasting, referring to the viewing frequency for each program (S405), as described above, a program having a higher viewing frequency than the predetermined threshold may be a “watching” program. The preference analysis process (S302) is performed using the program information after determining that the program is high and excluding the program.

  The attribute value preference level (S303) calculated in the preference analysis process (S302) is used in the item-specific preference level calculation process (S108 / S112), so that the attribute value preference level ( In step S305, the preference information storage unit 112 stores the preference information. The attribute value preference (S303) is normalized within each item.

This completes the attribute value preference calculation processing (S306).
<Details of the item-specific preference suitability calculation unit 205>
Next, the item-specific preference suitable degree calculation unit 205 (S108 / S112) will be described.

  The calculation of the degree of preference for each item is performed by receiving attribute value broadcast stipulated items such as genres and performers list included in the program information stored in the preference information storage unit 112 and attribute values such as keywords included in the program ranking information. This is performed for each item such as genre, performer list, keyword, etc. using the degree of preference of each attribute value included in the machine-defined item.

  As for the genre, since only one genre attribute value is included in one program, the preference level of the genre attribute value included in the program information is set as the preference degree of the genre of the program. As for the performer list and keywords, there is a possibility that a plurality of attribute values may be included in one program, so that the sum of the preference levels of all attribute values included in one program is normalized for each item. Degree. Specifically, for example, the attribute values of the performer list of a certain program are set to x1, x2,... XN, and the sum Psum of the preference values of the performer list attribute values is obtained using the formula shown in FIG. This expression normalizes the value of Psum from 0 to 1 using a logarithmic function that is said to be close to human sensitivity.

Furthermore, the preference degree sum Psum is normalized using the formula shown in FIG. 12, and the preference preference degree P of the performer list of the program is obtained. Nmax is the maximum number of performers for all programs scheduled to be broadcast in the future. A is an arbitrary constant and is determined in consideration of the curve shape of the logarithmic function in normalization. That is, as A increases, the degree of convexity on the logarithmic function increases, and the distribution of the degree of preference preference after normalization tends to be larger. Conversely, the larger A is, the smaller the degree of convexity on the logarithmic function is, and the preference suitable degree distribution is averaged. The value of A that gives the best result of automatic program selection is determined by experiment. Similarly for the keyword, the preference for each item is obtained using the formula of FIG. 11 and the formula of FIG.
<Flow of item preference suitability calculation processing>
Next, the flow of the item-specific preference suitability calculation processing (S108 / S112) performed by the item-specific preference suitability calculation unit 205 will be described in time series with reference to FIG.

The item-specific preference suitable degree calculation process is started (S501).
Using the attribute value preference degree (S305) stored in the preference information storage unit 112 and the program information (S801) and keyword (S205) stored in the program information storage unit 104, the equation of FIG. Based on the above formula, an item-specific preference suitability calculation process (S502) for each program is performed to obtain an item-specific preference suitability (S503) for each program. The obtained degree of preference for each item (S503) for each program is stored in the storage process (S504) for use in the program ranking determination criterion learning process (S109) and the program automatic selection execution process (S113). It is stored in the item-specific preference suitable degree calculation unit 205 as the different preference appropriate degree (S505).

This completes the item-specific preference suitability calculation process (S506).
<Details of Program Ranking Judgment Criteria Learning Unit 204>
Next, the program ranking for selecting programs using the preference preference degree by item (S505), the program viewing frequency (S405), and the viewing history (S802), which is performed by the program ranking judgment criterion learning unit 204. Details of the determination criterion learning process (S109) will be described. Note that this processing corresponds to a determination criterion setting unit in the claims.

  First, a neural network used for learning a program ranking judgment criterion will be described. A neural network is an information processing model that imitates the human brain and is widely used in fields such as artificial intelligence. FIG. 5 is a schematic diagram of neurons constituting the neural network. The neuron receives a plurality of values xi (i = 1, 2,..., N) as inputs, and adds a weight wi (i = 1, 2,..., N) to each input to add a weight sum Σxiwi. When the threshold value is exceeded, the weight sum is output as output y. The neural network is a hierarchical structure of these neurons as shown in FIG. 6, and one or a plurality of outputs are obtained from a plurality of inputs. In general, as shown in FIG. 6, the input layer ii (i = 1, 2,..., 9), the intermediate layer ni (i = 1, 2,..., 5), the output layer oi (i = 1, 2, 3). By making this neural network perform learning called backpropagation using a large number of training input / output pairs, the weight value and threshold value of each neuron that returns a correct output for a given input Can be requested.

Next, a method for learning a program ranking judgment criterion using a neural network will be described.
FIG. 7 is a configuration diagram of a neural network used to select a program using the preference suitability for each item and the program viewing frequency and to learn the criteria for determining the program ranking. In this neural network, the preference suitable degree for each item such as genre, performer list, keyword, etc. and program viewing frequency for any two programs are input, and 0 or 1 indicating which of the two programs the user likes. (The output of o1 is 1 if the program 1 is preferred, the output of o2 is 0, the output of o1 is 0, and the output of o2 is 1. To do). The neurons (n1, n2,...) Used for the intermediate layer may be configured with an arbitrary number in consideration of calculation time and accuracy.

Learning the criteria for judging program ranking uses the characteristic that when selecting a viewing program by a user, it is common to select a program to view by comparing multiple programs broadcast in the same time period. To do. Specifically, with respect to each viewing history stored in the viewing history storage unit using the neural network of FIG. 7, the program included in the viewing history and other programs that were broadcast while viewing the program The combination of the genre, performer list, keyword, etc. for each item and preference frequency for each item and the program viewing frequency are input, and the output is included in the viewing history. Learning by back-propagation is performed as a program having a high degree of preference is selected). More specifically, in the example of the program broadcast table of FIG. 8, if the program D is viewed during broadcasting, the combination of (program 1, program 2) in FIG. 7 is (program D, program A) (program D). , Program B) (program D, program F) (program A, program D) (program B, program D) (program F, program D), and the previous three sets output (o1, o2). The output of (o1, o2) is (0, 1) in the following three sets (1, 0). If the program G is viewed during broadcasting, the combination of (program 1, program 2) in FIG. 7 is (program G, program C) (program G, program E) (program C, program G) (program E , Program G), the output of (o1, o2) is (1, 0) for the previous two sets, and the output of (o1, o2) is (0, 1) for the latter two sets. In addition, recording reservation is the same as viewing during broadcasting. As for the viewing history of viewing the recorded program, for example, if the recorded program H is viewed between 20:00 and 21:00 in FIG. 8, the combination of (program 1, program 2) in FIG. H, Program C) (Program H, Program E) (Program H, Program G) (Program C, Program H) (Program E, Program H) (Program G, Program H) The output is 0, and the latter three sets have an output of 1. Learning of the neural network is performed by using combinations of such programs regarding all viewing histories and output values corresponding to the combinations as input / output pairs for training.
<Flow of program ranking judgment criteria learning process>
Next, the flow of the program ranking judgment criterion learning process (S109) performed by the program ranking judgment criterion learning unit 204 will be described in time series with reference to FIG.

  The criteria for determining the program ranking is obtained by a neural network that receives the viewing frequency (S405) and the item-specific preference degree (S505) of any two programs and outputs the selection result of one of the programs. .

The program ranking determination criterion learning process is started (S601).
As shown in FIG. 7, the preference for each item (S 505) and the viewing frequency (S 405) of a program viewed in the past in which the neural network can be referenced from the viewing history (S 802) are set as one input (program 1). The preference degree by item (S505) and the viewing frequency (S405) of the program that was broadcast at the same time as the viewing program and was not viewed as the other input (program 2), and the program that was viewed, that is, the use A program that can be determined that the user has selected with interest and preference is used as an output result, and learning by back-propagation is performed in the program ranking determination criterion learning process (S602). This program ranking determination criterion learning process (S602) is a combination of a predetermined number of programs stored as a viewing history (S802) and programs that were broadcast at the same time but were not viewed. Accordingly, the program ranking judgment criterion (S603) that matches the user's preference can be generated and updated. The generated and updated program ranking judgment criteria (S603) are used for the next program automatic selection execution processing (S113), so that the program ranking is used as the program ranking judgment criteria (S605) in the storage processing step (S604). Is stored in the determination criterion learning unit 204.

This completes the program ranking determination criterion learning process (S606).
<Details of the program selection unit 206>
Next, a program automatic selection execution process (S113) for selecting a program that suits the preference from programs scheduled to be broadcast in the future performed by the program selection unit 206 will be described.

  First, with respect to all combinations for selecting any two programs from among programs scheduled to be broadcast in the future, the degree of preference for each item such as genre, performer list, and keywords related to those programs (S505), and program viewing frequency Using (S405) as an input, the output of the neural network of FIG. 7 in which the program ranking judgment criterion (S605) held in the program ranking judgment criterion learning unit 204 is set is obtained. At this time, among the two input programs, the program determined to be more suitable for the user's preference (program 1 in the case of output 0 in FIG. 7 and program 2 in the case of output 1) is determined to meet the preference. Count the number of times played.

  Next, the programs are arranged in descending order of the number of times determined to match the preference, and a preset number of programs are selected from the higher programs. When the selected program is displayed as a recommended program, the number of selections is set so as to be convenient for displaying the recommended program list on the screen. FIG. 20 shows an example of a display screen when the selected program is displayed as a recommended program.

In addition, when recording a selected program automatically, the time zone overlaps when the time zones of multiple upper programs overlap so that multiple programs scheduled to be broadcast in the same time zone are not selected. You may perform processing such as selecting a program that does not overlap in a lower time zone instead of a lower-level program, or a program that displays a warning on a screen and records a program in which the time zones overlap A mechanism may be provided so that the user can select. If the capacity of the video / audio signal storage unit is not sufficient, the number of programs that can be stored in the video / audio signal storage unit may be selected.
<Flow of processing during automatic program selection>
Next, the flow of the program automatic selection execution process (S113) performed by the program selection unit 206 will be described in time series with reference to FIG.

Automatic program selection is started (S701).
Two arbitrary programs that are scheduled to be broadcast in the future are extracted, and the viewing frequency (S405) and the preference for each item (S505) of the two programs (program 1 and program 2) are extracted into the neural network shown in FIG. As an input, this neural network selects a program using a program ranking judgment criterion (S605) held in the program ranking judgment criterion learning unit 204. This program selection process is executed for all combinations of programs to be selected (S702). The program that is selected more frequently in this selection process is the program to be automatically selected. As a recommended program in the descending order of the number of selections, an automatic selection result display (S703) is displayed on the screen, or a program with the largest number of selections is automatically recorded.

Thus, the program automatic selection execution process ends (S704).
In the above-described embodiment, an example has been described in which the “viewing while watching” program is excluded during the preference analysis, and the accuracy of the program ranking determination criterion is improved using the viewing frequency and the item-specific preference suitability. Even with an automatic program selection apparatus that improves the accuracy of the program ranking criteria using only the viewing frequency and the item-specific preference suitability, it is possible to perform automatic program selection with higher accuracy than in the prior art. (Corresponding to claims 1 to 6)
In addition, although the present embodiment has been described mainly assuming digital broadcasting, it can also be realized by analog broadcasting by using program information provided by an electronic program guide service provider represented by a G guide or the like.

The present invention also includes a program for causing a computer to realize the functions of the program receiving and recording apparatus described above. These programs may be read from a recording medium and loaded into a computer, or may be transmitted via a communication network and loaded into a computer.

It is a block diagram which shows the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a block diagram of the program automatic selection part in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. This is an expression of a discriminant function as to whether or not the program p has been viewed w 'weeks before the scheduled broadcast date of the program p. It is a formula for obtaining the program viewing frequency of the program p. It is a schematic diagram of the neuron which comprises a neural network. It is a schematic diagram which shows a structure of a neural network. It is a block diagram of the neural network used for the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is an example of a program broadcast table. It is an expression for normalizing the number of appearances of attribute values of program information. It is an example of the preference value of the attribute value of program information. This is a formula for calculating the sum of preference values of attribute values of performers list or keywords included in a certain program. This is an expression that normalizes the sum of attribute values of performers or keywords. It is a flowchart which shows the whole flow of the program automatic selection process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the keyword extraction process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the attribute value preference degree calculation process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the viewing frequency calculation process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the item suitable preference suitable degree calculation process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the judgment reference | standard learning process of the program ranking in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a flowchart which shows the flow of the program automatic selection execution process in the program automatic selection apparatus using one Example of the program ranking apparatus of this invention. It is a display example of a recommended program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 201 Keyword extraction part 202 Viewing frequency calculation part 203 Preference analysis part 204 Judgment reference learning part of program ranking 205 Item preference suitability calculation part 206 Program selection part S216 Keyword extraction process step S205 View frequency calculation process step S106 Preference analysis process step S111 Program Ranking Judgment Criteria Learning Processing Step S208 Item-Specific Preference Suitable Level Calculation Processing Step S212 Program Automatic Selection Determination Processing Step

Claims (6)

The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking device that ranks in descending order
First storage means for storing a viewing history of a program viewed by a user;
Second storage means for acquiring and storing the program information associated with the program;
Viewing frequency calculation means for obtaining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
At the time of program ranking, using the viewing history and the program information, preference preference degree calculation means for obtaining one or more types of preference preference degrees indicating how much each program fits the user's preference When,
Of the two programs scheduled to be broadcast in the future to be used when ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation means at the time of past program ranking and the viewing history A determination criterion setting unit for determining which one is suitable for the user's preference, a determination criterion setting unit configured to execute the analysis process for each reset instruction to reset the determination criterion;
Program ranking means for ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Have
The determination criterion setting means sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards It is a criterion setting means for,
The program ranking means extracts any two programs within a predetermined scheduled broadcast period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times each program within the predetermined broadcast schedule period is determined to be suitable for the user's preference, and rank the programs in descending order. A program ranking apparatus characterized by the above. The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking device that ranks in descending order
The program information has a plurality of items, each of which is composed of one or more attribute values,
First storage means for storing a viewing history of a program viewed by a user;
Second storage means for acquiring and storing the program information associated with the program;
Viewing frequency calculation means for obtaining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
Preference analysis means for calculating the appearance frequency of the attribute value included in the program information accompanying each program in the viewing history for each calculation instruction, and obtaining the preference level for each attribute value;
At the time of program ranking, the preference degree calculation means for adding the preference degree corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future for each item of the program information and obtaining the preference degree of preference for each program When,
Of the two programs scheduled to be broadcast in the future to be used when ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation means at the time of past program ranking and the viewing history A determination criterion setting unit for determining which one is suitable for the user's preference, a determination criterion setting unit configured to execute the analysis process for each reset instruction to reset the determination criterion;
Program ranking means for ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Have
The determination criterion setting means sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards It is a criterion setting means for,
The program ranking means extracts any two programs within a predetermined scheduled broadcast period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times each program within the predetermined broadcast schedule period is determined to be suitable for the user's preference, and rank the programs in descending order. A program ranking apparatus characterized by the above. The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking method in the program ranking device for ranking in descending order of
The program ranking device comprises:
A first storage step of storing a viewing history of a program viewed by the user;
A second storage step of acquiring and storing the program information associated with the program;
A viewing frequency calculation step for obtaining a viewing frequency of each program corresponding to the program information stored in the second storage step based on the viewing history stored in the first storage step;
At the time of program ranking, using the viewing history and the program information, preference preference degree calculation step for obtaining one or more types of preference preference degrees indicating how much each program is adapted to the user's preference When,
Of the two programs scheduled to be broadcasted in the future to be used for ranking the programs by analyzing the preference suitability of each program obtained in the preference suitability calculation step at the time of ranking of past programs and the viewing history A determination criterion setting step for setting a determination criterion for determining which is suitable for the user's preference, and performing the analysis process for each reset instruction to reset the determination criterion;
A program ranking step of ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
It is characterized by performing
In the determination criterion setting step, one of the previously viewed programs included in the viewing history is set as a viewed program, and the preference suitability and the viewing frequency of the viewed program are set as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Standards A criterion setting step of setting,
In the program ranking step, any two programs within a predetermined broadcast scheduled period are extracted from program information of each program scheduled to be broadcast in the future stored in the second storage step, and the programs of the two programs are extracted. The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. And counting the number of times determined to match the user's preference for each program within the predetermined broadcast scheduled period based on the determination result, and ranking the programs in descending order of the number of times. A program ranking method characterized by the above. The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking method in the program ranking device for ranking in descending order of
The program information has a plurality of items, each of which is composed of one or more attribute values,
The program ranking device comprises:
A first storage step of storing a viewing history of a program viewed by the user;
A second storage step of acquiring and storing the program information associated with the program;
A viewing frequency calculation step for obtaining a viewing frequency of each program corresponding to the program information stored in the second storage step based on the viewing history stored in the first storage step;
A preference analysis step of calculating an appearance frequency of an attribute value included in program information accompanying each program in the viewing history for each calculation instruction, and obtaining a preference level for each attribute value;
At the time of program ranking, the preference degree calculation step of adding the preference degree corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future for each item of the program information to obtain the preference degree of preference for each program When,
Of the two programs scheduled to be broadcasted in the future to be used for ranking the programs by analyzing the preference suitability of each program obtained in the preference suitability calculation step at the time of ranking of past programs and the viewing history A determination criterion setting step for setting a determination criterion for determining which is suitable for the user's preference, and performing the analysis process for each reset instruction to reset the determination criterion;
A program ranking step of ranking programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
It is characterized by performing
In the determination criterion setting step, one of the previously viewed programs included in the viewing history is set as a viewed program, and the preference suitability and the viewing frequency of the viewed program are set as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations with programs that have been broadcast in the same time zone as the programs that have not been viewed, and the analysis process is executed for each reset instruction. Judgment criteria A criterion setting step of setting,
In the program ranking step, any two programs within a predetermined broadcast scheduled period are extracted from program information of each program scheduled to be broadcast in the future stored in the second storage step, and the programs of the two programs are extracted. The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. And counting the number of times determined to match the user's preference for each program within the predetermined broadcast scheduled period based on the determination result, and ranking the programs in descending order of the number of times. A program ranking method characterized by the above. The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking program that ranks in descending order,
A first storage function for storing a viewing history of a program viewed by a user in a first storage unit;
A second storage function for acquiring the program information accompanying the program and storing it in a second storage means;
A viewing frequency calculation function for determining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
A preference preference calculation function for obtaining one or more types of preference suitability indicating how much each program is adapted to the user's preference using the viewing history and the program information at the time of program ranking. When,
Of the two programs scheduled to be broadcast in the future used for ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation function in the past program ranking and the viewing history A determination criterion setting function for determining which is suitable for the user's preference, a determination criterion setting function for executing the analysis process for each reset instruction and resetting the determination criterion;
A program ranking function that ranks programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Is a program that causes a computer to execute
The judgment criterion setting function sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations of programs that have been broadcast in the same time zone and programs that have not been viewed, and the analysis process is executed for each resetting instruction. Resetting standards Is a criterion setting function,
The program ranking function extracts any two programs within a predetermined broadcast schedule period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times that each program within the predetermined broadcast scheduled period is determined to match the user's preference, and rank the programs in descending order. A program ranking program characterized by that. The user's preference is analyzed from the viewing history of the program accompanied by the program information, and the program information of each program scheduled to be broadcast in the future is used to adapt each program scheduled for future broadcast to the user's preference. In the program ranking program that ranks in descending order,
The program information has a plurality of items, each of which is composed of one or more attribute values,
A first storage function for storing a viewing history of a program viewed by a user in a first storage unit;
A second storage function for acquiring the program information accompanying the program and storing it in a second storage means;
A viewing frequency calculation function for determining the viewing frequency of each program corresponding to the program information stored in the second storage means based on the viewing history stored in the first storage means;
A preference analysis function for calculating an appearance frequency of an attribute value included in program information associated with each program in the viewing history for each calculation instruction, and obtaining a preference level for each attribute value;
A preference degree calculation function that adds the preference level corresponding to the attribute value included in the program information of each program scheduled to be broadcast in the future at the time of program ranking and obtains the preference degree for each program. When,
Of the two programs scheduled to be broadcast in the future used for ranking the programs by analyzing the preference suitability of each program obtained by the preference suitability calculation function in the past program ranking and the viewing history A determination criterion setting function for determining which is suitable for the user's preference, a determination criterion setting function for executing the analysis process for each reset instruction and resetting the determination criterion;
A program ranking function that ranks programs that match the user's preference from the programs scheduled to be broadcast in the future using the criteria.
Is a program that causes a computer to execute
The judgment criterion setting function sets one program among previously viewed programs included in the viewing history as a viewed program, and uses the preference suitability and the viewing frequency of the viewed program as first information. The watched program is viewed with the preference preference degree and the viewing frequency of one program among the programs that have not been viewed that are broadcast in the same time zone as the watched program as second information. The process of analyzing the first to third information with the selection result as the third information is processed for the predetermined number of the viewed programs included in the viewing history. The judgment criteria are set for all combinations with programs that have been broadcast in the same time zone as the programs that have not been viewed, and the analysis process is executed for each reset instruction. Re-established criteria Is a criterion setting function,
The program ranking function extracts any two programs within a predetermined broadcast schedule period from program information of each program scheduled to be broadcast in the future stored in the second storage means, and The process of determining which program is a program that better matches the user's preference from the preference preference degree and the viewing frequency by using the determination criteria is a combination of all programs within the predetermined broadcast scheduled period. , And based on the determination result, count the number of times that each program within the predetermined broadcast scheduled period is determined to match the user's preference, and rank the programs in descending order. A program ranking program characterized by that.

Images