JPH10162028A - Method and device for information retrieval - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily retrieve information that a group desires to have, from a large number of programs by determining the priority of information for the group consisting of plural users according to basic choice preferences characteristic of the users. SOLUTION: Character data for individual IDs of specific users allocated to a group with group seal data are stored in a user model storage and utilization part 14 together with the group seal data. On a monitor screen, characters and figures representing a maximum mode, a simple mean mode, and a minimum mode are displayed so that a desired mode can be selected and specified. A group user model formation part 53 generates a priority table of genres of group users with the selected mode and sends it out to an information file part 19, and then EPG data are rearranged according to the genre priority data to display a program guide on the display screen of the monitor device 4.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[Table of Contents] The present invention will be described in the following order. Technical Field to which the Invention pertains Prior Art Problems to be Solved by the Invention Means for Solving the Problems Embodiments of the Invention (1) Overall Configuration of Satellite Broadcasting Reception System (FIG. 1) (2) Configuration of Reception Decoding Device ( (FIGS. 2 and 3) (3) Information search by receiving / decoding device (3-1) Construction of general user model (FIGS. 2 to 17) (3-2) Formation of specific user model (FIGS. 18 to 28) 3-3) Use of specific user model (FIG. 29) (3-4) Formation and use of group user model (FIG. 3)
(0 and FIG. 31) (4) Operation and Effect of Embodiment (5) Other Embodiments (FIGS. 32 to 35) Effects of the Invention

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval method and apparatus, for example, a program retrieval method for retrieving a program required by a viewer from a large number of television programs distributed via broadcast satellites. And the apparatus is suitable.

[0003]

2. Description of the Related Art In a satellite broadcasting system in which a television program is distributed to a viewer via a broadcasting satellite, a television signal is digitized and an enormous number of programs are distributed simultaneously. In such a system, the number of programs selected by the viewer is significantly increased.

In a system in which various information is provided from a host to a computer terminal via a telephone line or a dedicated line, a user of the terminal selects necessary information from a vast amount of information and transmits the selected information to the host. Will be required.

[0005]

When a viewer or a user attempts to select such a television program or information using a computer, the user must search for a desired program or information from a vast amount of programs or information. To this extent, the search operation of the viewer or the user is complicated, and there is a problem that it is difficult to easily select a required program or information.

The present invention has been made in view of the above points, and it is an object of the present invention to propose an information retrieval method and an information retrieval apparatus which can greatly facilitate the retrieval of an individual or a group of viewers or users.

[0007]

According to the present invention, there is provided a general user group classified according to user attributes and / or general user selection preference data comprising classification items and information contents based on information use status. And forming a specific user model unique to the specific user based on the basic information selection preference data of the specific user, registering the specific user model formed for each of the plurality of users in association with each user, Based on one or more specific user models among the specific user models, information that matches one or more specific user models is searched.

According to the present invention, the general user selection preference data is corrected by the basic selection preference data of a specific user, and the basic user preference data is corrected by a plurality of basic selection preference data obtained for a plurality of specific users. Thus, by forming selection preference data of a group composed of a plurality of users,
A search for information suitable for the group-specific characteristics is performed.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

(1) Overall Configuration of Satellite Broadcasting Receiving System In FIG. 1, reference numeral 1 denotes a satellite broadcasting receiving system as a whole, and a broadcast signal received by a parabolic antenna 3 is a receiving / decoding device (IRD: Integrated Receiver / Decoder) 2 Are demodulated and compression decoded. A video / audio signal SV1 obtained as a result is converted to a VCR (Video Cassett
e Recorder) 6.

The VCR 6 records the video / audio signal SV1 on a video tape loaded therein, or sends the video / audio signal SV1 as it is from an output line to the monitor device 4 to display it on a monitor. I do.

When the viewer operates the remote commander 5, a command corresponding to the operation is converted into an infrared signal IR and transmitted to the receiving / decoding device 2. Reception decoding device 2
Switches the channel based on the command, registers / reads user data, and connects each device (VCR6, VCR7, DVD8 and MD9) connected to the receiving / decoding device 2.
And various operations such as transmission of a control signal CONT to the controller. The control signal CONT is sent to the VCR 6 via the control line. When the VCR 6 is designated as a control target by the control signal CONT, the control signal COT
The VCR 6 is controlled by the NT. On the other hand, as a control target of the control signal CONT, a device (VCR 7 of 8 mm system, a digital video disk player (DVD: Digital Video Player) sequentially connected to the VCR 6 via a control line.
deo Disc) 8, Mini Disc Player (MD: Mini Dis)
c) When any one of 9 and the monitor device 4) is designated, the VCR 6 sends the control signal CONT to the subsequent 8 mm system VCR 7 as it is.

When the VCR 7 receives the control signal CONT, it determines the device specified by the control signal CONT. When the result of this determination is VCR7, VC
R7 performs the operation specified by control signal CONT. This instruction is, for example, 8mm
If the instruction is to reproduce the video tape, the VCR 7 reproduces the video tape and sends the reproduced video signal SV3 to the monitor device 4 to display the reproduced video signal SV3. If the instruction by the control signal CONT is an instruction to record the broadcast signal (video / audio signal SV1) received and decoded by the receiving and decoding device 2 in the VCR 7, the VCR 7 sends the VHS signal from the receiving and decoding device 2 to the VHS system. V
The video / audio signal SV1 input via the CR 6 and the monitor device 4 is recorded. On the other hand, the control signal CONT
Is not the VCR 7, the VCR 7 sends the control signal CONT to the subsequent DVD 8 as it is.

When the control signal CONT is input to the DVD 8, the DVD 8 determines the device specified by the control signal CONT. When the discrimination result is DVD8, DV
D8 executes the operation specified by the control signal CONT. If this instruction is for reproducing video and / or audio from a disc loaded in a DVD, for example,
The VD 8 reproduces the disc and sends out the video / audio signal SV4 to the monitor device 4 to display it. On the other hand, when the control target of the control signal CONT is not the DVD 8, the DVD 8
The NT is transmitted to the subsequent MD9 as it is.

Upon receiving the control signal CONT, the MD 9 determines the device specified by the control signal CONT. When the discrimination result is MD9, MD9
Performs the operation specified by control signal CONT. When the instruction is, for example, an instruction to reproduce a disk loaded in the MD 9, the MD 9 reproduces the disk and sends an audio signal SV5 to the monitor device 4 to display the audio audibly. Further, the instruction by the control signal CONT causes the broadcast signal (video / audio signal SV1) received and decoded by the reception / decoding device 2 to be transmitted to the MD.
9, the MD 9 records the audio signal of the video / audio signal SV1 input from the receiving / decoding device 2 via the VHS system VCR 6 and the monitor device 4. On the other hand, the control target of the control signal CONT is M
If it is not D9, MD9 sends the control signal CONT to the monitor device 4 that continues. At this time, the monitor device 4 executes the operation specified by the control signal CONT.

(2) Configuration of Receiving / Decoding Apparatus In FIG. 2, the receiving / decoding apparatus 2 transmits an RF signal output from an LNB (Low Noise Block downconverter) 3A of a parabolic antenna 3 to a tuner 2 of a front end 20.
1 and demodulated. The output of the tuner 21 is Q
The signal is supplied to the PSK demodulation circuit 22 and QPSK demodulated.
The output of the QPSK demodulation circuit 22 is supplied to an error correction circuit 23, where an error is detected and corrected, and corrected if necessary.

A CAM (Conditional Access) constituted by an IC card comprising a CPU, a ROM and a RAM.
Module 33 stores a key necessary for decrypting the encryption together with a decryption program. Since the signal transmitted via the broadcast satellite is encrypted, a key and a decryption process are required to decrypt the code.
Therefore, the CA via the card reader interface 32
This key is read from M33 and demultiplexer 2
4 is supplied. The demultiplexer 24 uses this key to decrypt the encrypted signal.

The demultiplexer 24 receives a signal output from the error correction circuit 23 of the front end 20, supplies a decoded video signal to an MPEG video decoder 25, and supplies a decoded audio signal to an MPEG audio decoder 26.

The MPEG video decoder 25 stores the input digital video signal in the DRAM 25A,
The decoding process of the video signal compressed by the PEG method is executed. The decoded video signal is NTS
The luminance signal (Y), the chroma signal (C), and the composite signal (V) of the NTSC system are supplied to the C encoder 27.
Is converted to The luminance signal and the chroma signal are output as S-video signals via buffer amplifiers 28Y and 28C, respectively. The composite signal is output via a buffer amplifier 28V.

The MPEG audio decoder 26 stores the digital audio signal supplied from the demultiplexer 24 in the DRAM 26A, and executes a decoding process of the audio signal compressed by the MPEG system. The decoded audio signal is supplied to the D / A converter 3
At 0, digital / analog conversion is performed, the left channel audio signal is output via a buffer amplifier 31L, and the right channel audio signal is output via a buffer amplifier 31R.

The RF modulator 41 converts a composite signal output from the NTSC encoder 27 and an audio signal output from the D / A converter 30 into an RF signal and outputs the RF signal. The RF modulator 41 passes an RF signal of the NTSC system input from another device and outputs the signal to another device as it is. In this example,
These video signals and audio signals are supplied to the VCR 6 via the AV line.

The CPU 29 executes various processes according to a program stored in the ROM 37. Also, C
The PU 29 controls the AV device control signal transmission / reception unit 2A, outputs a predetermined control signal to another device via a control line, and receives a control signal from another device.

A predetermined command can be directly input to the CPU 29 by operating an operation button switch on the front panel 40. When an operation key of the remote commander 5 is operated, an infrared signal is output from the IR transmitting section of the remote commander 5, the infrared signal is received by the IR receiving section 39, and the light receiving result is supplied to the CPU 29. Therefore, the remote commander 5
By operating, a predetermined command can be input to the CPU 29.

The CPU 29 is provided with the demultiplexer 2
4 fetches, for example, EPG (Electronic Program Guide) information other than the video signal and audio signal output from
EPG data will be created from this, and SRAM (Static Ra
ndom Access Memory) 36 for storage. EP
The G information includes information (for example, a channel of a program, a broadcast time, a title, a genre, a program description, and the like) of a program of each broadcast channel from the current time to several tens hours later. Since this EPG information is transmitted frequently, the SR
The AM 36 always holds the latest EPG information.

An EEPROM (Electrically Erasable Pro)
The gramable read only memory (38) stores data (for example, rewritable reception histories of the past four weeks of the tuner 21 and data of the knowledge base unit 11 to be described later) to be held even after the power is turned off. Also, the CPU 29
The time information output by the calendar timer 35 is compared with the time stamp output by the demultiplexer 24 from the received signal, and the MPEG video decoder 25 performs decoding at the correct timing according to the result of the comparison. And the MPEG audio decoder 26.

Further, the CPU 29 executes a predetermined OSD (On-
(Screen Display) When data is to be generated, the MPEG video decoder 25 is controlled. The MPEG video decoder 25 generates predetermined OSD data in accordance with the control, writes the generated OSD data into the DRAM 25A, and further reads out and outputs the data. Thus, predetermined characters, figures, and the like can be output to the monitor device 4 and displayed.

FIG. 3 shows a block diagram of the information search function unit for searching for a program desired by the user based on the EPG information in the reception / decoding device 2 of FIG. That is, in FIG. 3, the knowledge base unit 11, the user model storage / utilization unit 14, and E
EEPR of the reception decoding device 2 as the PG data storage unit 18
The OM 38 (FIG. 2) is assigned (however, the CPU 29 (FIG. 2) is assigned to the utilization unit of the user model storage / utilization unit 14), and the user model formation unit 13, the group user model formation unit 53, and the information filter unit 19 are provided. CP
U29 (FIG. 2) is assigned and the EPG data input unit 1
The front end 20 (FIG. 2) is assigned as 7,
The remote controller 5 and the front panel 40 (FIG. 2) are assigned as the user profile input unit 12 and the user operation unit 15, the IR receiving unit 16 and the CPU 29 (FIG. 2) are assigned as the search request transmitting unit 16, and the recommendation information display unit The MPEG video decoder 25 (FIG. 2) is assigned as 50, and the NTSC encoder 27 (FIG. 2) is assigned as the display controller 51. Incidentally, the data of the knowledge base unit 11 is stored in the EEPROM 38 at the time of product shipment of the receiving / decoding device 2, and by taking in new data from outside via the satellite or the modem 34,
It can be updated. If the data of the knowledge base unit 11 cannot be updated, the data stored in advance as the data of the knowledge base unit 11 may be stored in the ROM 37. The EPG data taken in through the front end 20 (FIG. 2) and the modem 34 (in the case of the Internet) is stored in the SRAM 36.

When the operation key of the program guide is selected by the user operation section 15 of FIG. 3, the CPU 29 of FIG.
The EG video decoder 25 (recommended information display unit 50) is controlled to display a broadcast program selection screen on the monitor device 4. The user can select and specify a desired program by moving the cursor to a desired program position on this screen and clicking on the remote commander 5.

The CPU 29 can update the data stored in the EEPROM 38 (that is, the knowledge base unit 11 in FIG. 3) with the latest data externally fetched through the modem 34. By the way, EEP
As a method of taking in the data in the ROM 38 from an external device (floppy disk, card-shaped recording medium, etc.),
In addition to communication using a modem, an output line dedicated to data may be provided.

The CPU 29 includes an EEPROM 38
The data stored inside (ie, the suitable base unit 11) is stored in an external storage device via the modem 34 or an output line (not shown) provided for exclusive use, and is taken in as necessary. Thus, the data of the knowledge base unit 11 before being updated can be used again.

(3) Information Retrieval by Receiving / Decoding Apparatus (3-1) Construction of General User Model FIG. 3 is a functional block diagram of a part relating to information retrieval of the receiving / decoding apparatus 2 described above with reference to FIG. The unit 11 stores in advance “user viewing behavior” obtained based on the result of the viewer survey as a “general user model”.

The "general user model" is a model in which users who are viewers are classified according to various characteristics, and the programs viewed by each of the classified users are indicated by a rating for each genre. The user's information usage status includes audience ratings by genre for each scene of life, well-viewed images (genre),
Weekly viewing time, etc.

As a user classification method, first, there is a classification based on age and gender. In this classification, the users are classified into 12 items of 10-year-olds from the age of 10 to 50-year-olds, and 6 items of the age of 60 or older, classified by gender. As described above, the genre audience rating for each life scene classified by age and gender is provided as a data table TB1. This data is
As shown in FIG. 4 (A), “at breakfast”, “at lunch”,
It is an audience rating table of each program genre for every life scene, such as "at dinner", "when relaxing on weekdays", and "when relaxing on holidays". For example, the audience rating of a plurality of programs at "breakfast time" is tabulated for each age and gender.

FIGS. 5 to 10 show specific examples of data tables in which these tabulated data are summarized for each of a plurality of living situations. FIG. 5A shows the audience rating by program genre for each life scene of a teenager male, and FIG. 5B shows the audience rating by program genre for each living scene of a 20-year-old male. FIG. 6A shows the audience rating by program genre for each life scene of a male in his 30s, FIG. 6B shows the audience rating by program genre for each life scene of a male in his 40s, and FIG. ) Shows the audience rating by program genre for each life scene of a male in his 50s, and FIG. 7B shows the audience rating by program genre for each life scene of a male in his 60s.

FIG. 8 (A) shows the audience rating by program genre for each life scene of a teenage female, and FIG.
9A shows the audience rating by program genre for each life scene of a 0-year-old woman, FIG. 9A shows the audience rating by program genre for each living scene of a 30-year-old woman, and FIG. Figure 1 shows the audience rating by program genre for each living scene of women in their teens.
0 (A) shows the audience rating by program genre for each living scene of a woman in her 50s, and FIG. 10 (B) shows the audience rating by program genre for each living scene of a woman in her 60s.

Thus, as shown in FIG. 11, a data table composed of audience ratings for each age, gender, life scene, and program genre, and an average data table obtained by averaging the audience ratings of each program genre among the data tables are shown. By storing the information in the knowledge base unit 11 (FIG. 3), the user can select a program. It can be stored in advance as user data. Therefore, the viewing tendency appearing for each age and gender of the general user can be obtained from the data table TB1.

Second, as a user classification method, as shown in FIG. 4B, there is a classification based on a life stage. This classification includes “advancement” as a user's personal growth stage that takes into account the user's relationship with the family and society,
Classify by gender into stages such as "employment", "marriage", "child care" and "retirement". The genre audience rating for each life scene of each user thus classified is provided as a data table TB2. This data is an audience rating table of each program genre for each life scene such as "at breakfast", "at lunch", "at dinner", "when relaxing on weekdays", and "when relaxing on holidays". For example, at "breakfast time", the ratings of a plurality of programs are tabulated for each gender-specific life stage. At this time, for the stage until the married woman becomes unemployed at the age of 60, she separates working women from full-time housewives.

Thus, the data base comprising the audience rating for each life stage, life scene, and program genre, and the average data table obtained by averaging the audience rating of each program genre among the data tables, are respectively stored in the knowledge base unit 1.
1 (FIG. 3), as a feature for the user to select a program, the habit of determining the genre of the program to be viewed in relation to each life scene is defined in advance as general user data for each life stage. Can have. Therefore, the viewing tendency that appears for each general user's life stage can be obtained from the data table TB2.

Thirdly, as a user classification method, as shown in FIG. 4C, there is a classification based on attitude-value characteristics regarding television viewing. In this classification based on attitude-value characteristics, eight factors based on the results of factor analysis of opinions about watching television programs (television-viewing attitude values) from an overall significant number of market surveys covering viewers. Classify into. These factors are the active participation axis indicating whether or not to actively participate, the knowledge culture axis indicating whether or not to seek knowledge education, the affirmative / negative axis indicating whether or not to affirm television, and limiting the television. A lazy / limited axis that indicates whether or not to watch, a trend-oriented axis that indicates whether emphasis is placed on topicality, a whim / standard axis that indicates whether or not to watch on a whim, a comfort / immersion that indicates whether or not to watch casually The axis is determined by the loneliness / union axis, which indicates whether to look alone. For each of the factors classified in this way, a group that responds strongly is defined as an upper group of (+) and a lower group of (−), and 16% of users are picked up, and each factor (+) group, (−) ) Each group has a genre audience rating for each life scene as a data table TB3. This data is "at breakfast", "at lunch", "at dinner", "when relaxing on weekdays" and "when relaxing on holidays".
It is an audience rating table of each program genre for every life scene like this. For example, the audience rating of a plurality of program genres at "at breakfast" is tabulated for each factor.

Thus, the data base comprising the ratings for each factor, life scene, and program genre, and the average data table obtained by averaging the ratings for each program genre among the data tables are respectively stored in the knowledge base unit 11 (FIG. By storing the information in 3), as a feature for the user to select a program, the habit of determining the genre of the program to be viewed in relation to each life scene is generally determined for each factor (ie, television viewing attitude values). It can be stored in advance as user data. Therefore, the viewing tendency that appears for each value of the general user can be obtained from the data table TB3.

As a classification method based on the attitude and value characteristics related to television viewing, there is a classification based on clusters in addition to the classification by factor shown in FIG. 4C. In the classification by this cluster, among the eight factors described above with reference to FIG. 4C, for example, three factors relating to television viewing attitude and behavior (active participation axis, lazy / limited axis, easy / immersive axis) are analyzed. By doing so, eight user clusters are obtained.

That is, as a first cluster, a “balanced group” who likes a slightly affirmative television program, a wobbling program, and a storyless program with a very average viewing, and as a second cluster, he keeps watching television in a lazy manner. "Daradara Zapping", who prefers easy-going, easy-going TV programs, and as a third cluster, prefers programs with storylines and dramas, rather than those who watch the most lazy and actively participate in TV. As the fourth cluster, "Daradara appreciation tribe," as the fourth cluster, "Story program immersion tribe," who prefers the feeling of story and drama and tries to limit programs instead of actively participating in TV , Actively participate in TV and enjoy everyone, I like stories and drama-like moving programs, and I tend to watch somewhat lazy "TV is a friend As objective group ", the sixth of the cluster,
The "anti-TV (TV) family", who is negative and passive about television and chooses only the programs they want to watch and sees easy-going programs without a story. As an eighth cluster, "Busy people who like but can't see" regardless of whether they're positive, participatory, drama-inspiring or easy-going
They prefer easy-going, low-profile programs that have no storyliness, positively and actively participate in television, and are classified as "laughable tribes" that select only programs that they want to watch relatively.

Thus, the data table including the audience ratings for each cluster, life scene, and program genre, and the average data table obtained by averaging the audience ratings of each program genre among the data tables are respectively stored in the knowledge base unit 11 (FIG. By storing the information in 3), as a feature for the user to select a program, the habit of determining the genre of the program to be viewed in relation to each life scene is generally defined for each cluster (that is, television viewing attitude values). It can be stored in advance as user data. Therefore, the viewing tendency that appears for each value of the general user can be obtained from the data table. Incidentally, FIG. 12 shows a specific example of the genre audience rating for each life scene of the first cluster.

As general user data stored in the knowledge base unit 11 (FIG. 3), there is information usage status data of a user. As the first information use status data, FIG.
As shown in FIG. 3A, a table of weekly audience rating data corresponding to each program genre for each age and gender, and a program genre frequently viewed for each age and gender as shown in FIG. Some are tabulated by audience rating.
FIG. 14 shows a specific example of the table in FIG.

As the second information use situation data,
A table of weekly audience rating data corresponding to each program genre for each life stage as shown in FIG. 15 (A), and a program frequently viewed for each life stage as shown in FIG. 15 (B) There is a table of janles by audience rating.

As the third information use situation data,
As shown in FIG. 16A, weekly audience rating data corresponding to each program genre is tabulated for each factor axis.
In addition, as shown in FIG. 16B, there is a table in which frequently viewed program genres are tabulated by audience rating for each factor axis. FIG. 16 shows the fourth information usage status data.
FIG. 16C shows a table of weekly audience rating data corresponding to each program genre for each attitude value cluster, and FIG. 16 shows a program genre frequently viewed for each attitude value cluster, as shown in FIG. Some are tabulated. FIG. 17 shows specific data of the above-mentioned eight clusters and the frequently viewed program genres corresponding to the respective clusters.

(3-2) Formation of Specific User Model In FIG.
To form a specific user model suitable for the specific user. That is, the user operates the remote commander 5 (FIG. 2) of the user profile unit 15 to input the user profile in an interactive manner with the screen displayed on the monitor device 4.

FIG. 18 shows the operation keys of the remote commander 5. The cursor movement key group K10 includes movement keys K14 and K12 for moving the cursor up, down, left and right.
K15 and K13 and an enter key K11 for deciding an item pointed by the cursor are provided. The remote commander 5 has a guide key K21 for requesting a program guide, an explanation key K20 for displaying various explanations on the monitor screen, a menu display key K22 for displaying various menus on the monitor screen, and a group user. Mode selection key K23 for performing a group user, and a family mode key K2 for extracting data of all users registered when forming a group user
4 and the like are provided.

As shown in FIG. 19, the user operates the "menu" display key K22 of the remote commander 5 in step SP11 of the processing procedure, as shown in FIG. FIG. 2) displays a menu screen on the display screen 4A of the monitor device 4. In the following step SP12, the user operates the cursor movement key K10 of the remote commander 5 to move the cursor on the screen to a position for instructing profile registration, and selects this with the enter key.

As a result, a personal ID character selection screen as shown in FIG. 20 is displayed on the monitor screen 4A, and the user inputs the user's personal ID in an interactive manner with the screen at step SP13. In the case of FIG.
By designating the characters “A” to “E” with the cursor, the personal ID of the specific user who is currently performing profile input to any of these five characters
You can register as a character. When the registration of the personal ID character is completed, another interactive screen is displayed on the monitor screen 4A, and the user proceeds to step S.
At P14, the user profile is input in the form of a question on the interactive screen. In step SP14, first, FIG.
Is displayed. 21 is a question screen corresponding to the general user data table TB1 (FIG. 4A) classified and set for each age and gender in the knowledge base unit 11, and the user moves the cursor on the screen to the remote commander 5 By operating the button and clicking the enter key K11, the age and gender specified by the cursor can be input.

When the input of age and gender is completed,
2 is displayed on the monitor device 4. FIG.
In the case of, the general user data table TB2 (FIG. 4) classified and set for each life stage of the knowledge base unit 11
9B is a question screen corresponding to (B)), in which the user moves the cursor on the screen by operating the remote commander 5;
By clicking the setting key, a rice stage item designated by the cursor can be input.

When the input of the life stage is completed, an interactive screen shown in FIG.
In the case of FIG. 23, the question screen is a question screen corresponding to the general user data table TB3 (FIG. 4C) classified and set for each factor of the attitude value of the knowledge base unit 11 relating to the television viewing. By moving the cursor by operating the remote commander 5 and clicking the setting key, the factor item designated by the cursor can be input.

When the input of the factor items is completed, the process proceeds to FIG.
4 is displayed on the monitor device 4. FIG.
Is a question screen corresponding to a general user data table (FIG. 12) classified and set for each cluster of attitude values regarding television viewing of the knowledge base unit 11, and the user moves the cursor on the screen to the remote commander 5 By operating the cursor and clicking the enter key K11, the cluster designated by the cursor can be input.

The profile data of the specific user input from the user profile input unit 12 shown in FIG. 3 is sent to the user model forming unit 13 (FIG. 3). The user model forming unit compares the user profile data input from the user profile input unit 12 with each data table of the knowledge base unit 11, specifies a user group from a data table that matches, and specifies this as a user model. I do. For example, when a user model of a man in his thirties is specified from a profile input of age and gender, FIG.
From each data table of the above-mentioned age sex / life scene by / genre-specific viewership data of 30 Saitai male table it is formed as a user model U _1A of the specific user for.

Similarly, a user model U _1B for each life stage, a user model for each factor, and a user model for each cluster are formed in accordance with the respective profile inputs.

The user model forming unit 13 _calculates the following equation from the age-sex user model U _1A and the life stage user model U _1B .

(Equation 1) Get a table U ₁ of "life scene genre audience, as seen from the demo graphic" represented Te Niyotsu.

Further, the user model forming unit 13 determines each factor of the genre audience rating for each life scene (FIG. 4C) viewed from the factor axis of the general user model stored in the knowledge base unit 11 in advance. from the process to profiles input data relating to factors specific user has (Fig. 23), obtaining a table U ₂ of life scene by genre audience viewed from factor axis related to the specific user. That is, Yotsute the profiles input data by factors items described above with reference to FIG. 21, factor value of the particular user at this time F _(n) is obtained for each factor of eight items, these eight items factors value F _{(n )} , The average value of the audience ratings for each of the program genres is tabulated for each of the living scenes.
_{Get two} .

Accordingly, the user model forming unit 13 determines the above-mentioned “life scene genre audience rating viewed from the demographic”.
From the table U ₁ and the table “Upper audience rating by genre from the viewpoint of factor axis” table U ₂ ,

(Equation 2) As a result, the data U of “Journal audience rating by life scene” of the specific user is obtained. As a result, as shown in FIG. 25, a “daily audience rating by life scene” table relating to the specific user is obtained, and stored in the user model storage / utilization unit 14 together with the personal ID character data described above with reference to FIG. Incidentally, the user model storage / utilization unit 14 is provided with a storage area for each of a plurality of specific users, and each area stores user data together with the corresponding ID character data.

Here, the specific user data of FIG.
In this case, the specific user is 2
The case of an eight-year-old man will be described. Knowledge base unit 11
Is a format (10, l, 2) in which each chronological data is provided with an interpolation control identifier "l" as an attribute of the chronological user group.
0, l, 30, l, 40, l, 50, l, 60, l) have attribute data. Therefore, the user profile input unit 1
When the data “28-year-old man” is input to the user profile forming unit 2, the user model forming unit 13
The interpolation control identifier “l” and the age and gender are detected from the data received from FIG.
6 is compared with the interpolation control identifier “l” of the data of the 20-year-old man and the 30-year-old man, and the arithmetic expression corresponding to each interpolation control identifier, here, the arithmetic expression of the linear interpolation is applied to the individual user. Of the user model data.

As a result, the age data, which is continuous data, can be compressed and stored in the knowledge base unit 11 at every tenth generation. Incidentally, as attributes of the user classification that effectively uses the interpolation control, in addition to the age attribute described above with reference to FIG. 26, there are a hobby of the user, years of experience in sports, and years of continuation of membership members, and the like.

The data of “Journal audience rating by life scene” concerning the specific user obtained in this manner is obtained based on the life scene of the general user model. In this case, the user model forming unit 13 has a life scene / time correspondence function as shown in FIG. 27, and this function is sent to the user model storage / utilization unit 14 so that the user can actually request the program guide. The genre preference value (viewing rate) corresponding to the life scene at the time n is calculated by the user model storage / utilization unit 14. In this regard, it can be determined that the higher the audience rating, the higher the degree of preference. The preference value can be obtained by directly assigning the audience rating or assigning a number corresponding to the ratio.

For example, the rating data at dinner is multiplied by a coefficient of “1” from 19:00 to 20:00, and the coefficient becomes “0” at 21:00. Between 20:00 and 21:00, coefficients are multiplied by linear interpolation. After 21:00, the data at the time of relaxation after dinner is multiplied by a coefficient “1”. At 20:00, the data at the time of relaxation is multiplied by a coefficient of “0”.

As a specific example, when a program guide is requested at 20:40, the user model storage / utilization unit 14
The genre preference value F at 20:40, where the genre preference value at dinner is F _y and the genre preference value when relaxing on weekdays is F _k.
20 is

(Equation 3) It is calculated by

By the way, the genre preference value obtained by the corresponding function is calculated by the corresponding function (parameters are constants) determined as the life scene of the general user model. In this case, the user model forming unit 13 matches the “daily audience rating by life scene” of the specific user obtained based on the user profile input with the lifestyle of the specific user based on the meal time data of the specific user. It is made to convert to a data table.

That is, the knowledge base unit 11 has a variable term that can be changed according to the lifestyle of the specific user with respect to the corresponding function, and the user model forming unit 13 inputs variable data to be substituted into the variable term to the user. Request. As this input request, the user model forming unit 13 displays an interactive screen as shown in FIG. Based on this screen, when the user inputs a normal dinner start time a, a normal dinner end time b, an early dinner start time c, and a late dinner end time d as variable data using a remote commander, the user model The forming unit 13 calculates a genre preference value (audience rating) at a time n [hour] at which there is a program guide request based on the input data.

That is, when the program guide request time n is time a
And b, the genre preference value F _yn relating to the life scene at the time of dinner at the time n is represented by the following equation, where the genre preference value at the general user at the time of dinner is F _y :

(Equation 4) Is calculated by

When the program guide request time n is between times c and a, the genre preference value F _yn relating to the life scene at the time of dinner at the time n is given by the following equation:

(Equation 5) Is calculated by

Further, the program guide request time n is changed to times b and d.
In the case where the value is between 0 and 0, the genre preference value F _yn regarding the life scene at the time of dinner at time n is expressed by the following equation:

(Equation 6) Is calculated by

If there are m types of life scenes and the genre preference values for all the life scenes at the program guide request time n are F _n , the following equation is obtained.

(Equation 7) Is represented by However, when F _n > 1, F _n = 1
Where the average value of the genre preference is F _a and the genre preference value F _n is _a fixed percentage of the average value Fa (for example, 1 of F _a ).
If / 2) less than, and F _n = 0.5F _a. The data obtained in this way is calculated every time there is a request from the user for the program guide, and is calculated by the information filter unit 19 (FIG. 3).
Sent to

Thus, the continuous time is compressed by the life scene data and is corrected to data suitable for the habit of the specific user. By the way, information usage information includes not only life scenes but also attribute data such as commuting time affecting media usage, information request frequency, movies, music related to news value, and years passed since the release of publications. Can be used.

(3-3) Utilization of Specific User Model When the user operates the program guide key K21 and the ID input key K25 of the remote commander 5 (FIG. 2) of the user operation unit 15 at the time of watching television, a search is performed from the remote commander 5. A program guide request command of the specific user specified by the ID is transmitted to the request transmission unit 16. The search request transmission unit 16 sends the day of the week data read from the internal clock and the filter data output request signal to the user model storage / utilization unit 14 in response to the program guide request.

[0072] The user model storing / utilization unit 14 calculates the genre preference value F _n from the viewership data table for a specific user model had it occurred designated ID corresponding to the day time when the program guide claims, thereby A genre priority table is created, and the group user model forming unit 53 is created.
To the information filter unit 19 via the.

The information filter unit 19 is configured to transmit a satellite broadcast wave to an EP.
The EPG (Electronic Program Guide) received by the G data information input unit 17 and stored in the EPG data storage unit 18
ide) EP corresponding to the guide request time in the data
The G data is rearranged according to the genre priority table, and sent to the recommendation information display unit 50.

The recommendation information display section 50 displays the EPG data rearranged according to the genre priority table on the monitor device 4 according to a predetermined display form. As a result, FIG.
As shown in FIG. 9, on the display screen 4A of the monitor device 4, the programs being broadcast at the time of requesting the program guide are displayed as a list with a combination of characters and / or graphics indicating the broadcast station or the like,
Of the list, only the list given the priority by the information filter unit 19 is surrounded by a thick frame. The bold frame is colored by genre, and the user selects a desired program from the programs surrounded by the bold frame by referring to the color by using a cursor.

Thus, the initial user model for each specific user is formed together with the ID character data based on the profile data input from the profile input unit 12, and a program desired by the user is selected based on the data. . Incidentally, when the user requests a program guide by designating a fixed time in the future, for example, the search request transmitting unit 16 sends the data of the time and date to the user model storage / utilization unit 14 to make the designation. It is possible to calculate the program selection information of the time and date that have been set.

(3-4) Formation and Utilization of Group User Model As described above, any one of a plurality of specific user models stored in the user model storage / utilization unit 14 is used as the personal ID character data. Identify using
This is a so-called personal program guide request mode for displaying a program guide according to the specified user model. On the other hand, from a plurality of specific user data stored in the user model storage / utilization unit 14, a predetermined number of user models are selected in accordance with designation of the user, and the data of the selected plurality of users are selected. There is a so-called group user program guide request mode in which a group user model is formed based on (specific user model) and a program guide is displayed based on the group user model.

In this group user program guide request mode, the user operates the mode selection key K of the remote commander 5.
23 is operated, and at this time, the CPU 2
9 (FIG. 2) displays the group member registration screen shown in FIG. That is, first, an interactive screen shown in FIG. 30A is displayed on the screen 4A of the monitor device 4. On this screen, a plurality of characters representing personal IDs are displayed together with corresponding symbols such as alphabets. The user designates a symbol, which is called a group seal, for specifying a group name by using a cursor, and decides by using a decision key K11 of the remote commander 5. As a result, the cursor moves below each character, and the user designates one of the characters by moving the cursor left or right. When the cursor is moved to a predetermined character position and the enter key K11 is operated, the character designated by the cursor is assigned the group seal designated by the cursor at this time.

After the group seal is assigned to one character, the cursor is further moved left and right to designate another character and the same operation is repeated, thereby obtaining a plurality of characters as shown in FIG. The same group seal can be assigned to characters. By the way,
The cursor is once again moved to the character to which the group seal has been assigned, and by operating the enter key K11 in this state, the group seal once assigned can be erased.

By allocating the same group seal to a plurality of characters in this way, the characters to which the same group seal is allocated can be set as one group. This group name is specified by the group seal data. The personal ID character data of each specific user assigned to the group by the group seal data is stored in the user model storage / utilization unit 14 together with the group seal data indicating the group name. As a result, the group model once registered in the group can easily display a program guide corresponding to the group user model on the monitor simply by inputting the group name and a group user model formation mode described later.

When the assignment of the group seal is completed,
When the user moves the cursor to the end position (FIG. 30) and operates the enter key K11, the registration of the group user is completed. The user model of each specific user corresponding to each character registered in the group as described above is stored in the user model storage / utilization unit 14 according to the group registration.
, And sent to the group user model forming unit 53 (FIG. 3).

When the group user registration is completed, a screen for selecting the group user model formation mode is displayed on the monitor screen. Characters and / or graphics representing the maximum mode, the simple average mode, and the minimum mode are displayed on this screen, and the user can select and specify a desired mode using a cursor.

Here, the group user formation mode will be described. As shown in FIG. 30, for example, assuming that there are two specific user models constituting a group user, a graph representing a preference value (viewing rate) for each genre of the first user model is represented by L1, and a graph of each of the second user models is represented by L1. A graph indicating a preference value (viewing rate) for the genre is L2. Of the preference values of these two user models, the mode that always selects the maximum value is called the maximum value mode, the mode that always selects the minimum value is called the minimum value mode, and the mode that always selects the two average values is the simple average mode. Call.

Therefore, when the maximum value mode is selected,
Of the genres preferred by each specific user constituting the group user model, the genre having the highest degree of preference is always selected. When the minimum value mode is selected, the genre having the lowest preference level is always selected from the genres preferred by each specific user constituting the group user model. If the simple average mode is selected,
Among the genres selected by each specific user constituting the group user model, a genre having an average preference level is always selected.

Thus, when the maximum value mode is selected, if any one of a plurality of specific users has a genre that the user particularly wants to see, the genre is selected. When the minimum value mode is selected, a genre that does not cause any dissatisfaction with any of the plurality of specific users is selected. Also, when simple average mode is selected,
A genre that all of the plurality of specific users prefer to some extent is selected.

Thus, the group user model forming section 53
Creates a priority table for the genre of the group user according to the selected mode and sends it to the information filter unit 19, whereby the EPG data is rearranged according to the genre priority data. As a result, the program guide described above with reference to FIG. 29 is displayed on the display screen 4A of the monitor device 4.

Incidentally, when generating a program guide with all the specific user models stored in the user model storage / utilization unit 14 as group users, the user operates the family key K24 of the remote commander 5 to The user model storage / utilization unit 14 registers all user models as group users, forms a group user model from these models, and displays a program guide.

(4) Operation and Effect of Embodiment In the above configuration, the general user model stored in advance in the knowledge base unit 11 is the result of analyzing the general user's television viewing style, and is not necessarily the specific user. Not true. Accordingly, by correcting the general user model based on the profile data of the specific user input from the user profile input unit 12 by the specific user, a user model unique to the specific user is formed.

At this time, the data input as the specific user profile include the age and gender of the specific user described above with reference to FIG. 21, the life stage of the specific user described above with reference to FIG. 22, and the “attitude and value of television viewing” described above with reference to FIG. The “Television Attitude and Attitude Values” cluster described above with reference to FIG. 24 is used. These profile data include specific user's basic preferences, values, and how to watch television programs and television programs. This is the basic data of a specific user that does not change in the long term.

Therefore, the specific user model formed based on the profile data is a unique basic model of the specific user that does not change in the long term. By grouping the specific user models formed in this way for a plurality of users and forming a group user model, a group model is formed from a model suitable for the basic characteristics of each user constituting the group.

By determining the priority of the program based on the group user model formed by combining the basic models specific to the specific user in this way, each user characteristic is not simply synthesized with the general user model. The program is searched according to the group user model having the reflected group-specific characteristics.

Thus, according to the above-described configuration, the characteristic unique to the group is obtained based on the basic selection preference unique to each specific user, and the program priority is determined by this. From among them, a program list based on their basic selection preferences can be easily obtained. By the way, the actual program selection list display example is shown in reference drawing 1.
And reference drawing 2.

(5) Other Embodiments (5-1) In the above-described embodiment, the program classifications are "news", "sports", "variety", "drama", "movie", "documentary", Although the case of setting genres such as "music", "culture" and "children's program" has been described, the present invention is not limited to this, and various other genres and further subdivided genres or performer classifications, etc.
Other classification methods may be used.

(5-2) In the above-described embodiment, the case where the initial user model is formed based on the classification by the age, sex, life stage, attitude value factor, and attitude value cluster has been described. The present invention is not limited to this, and one or two of them may be selected and formed.

(5-3) In the above-described embodiment, a case has been described in which discrete data of, for example, 10 years old is stored in the knowledge base unit 11 as age data, and this is linearly interpolated. Instead, various interpolation calculations such as trapezoidal interpolation can be applied.

(5-4) In the above embodiment, the display example of the program list displayed on the monitor device 4 is shown in FIG.
32, a case has been described where a plurality of program lists and a selected program list are simultaneously displayed. However, the present invention is not limited to this. For example, as shown in FIG. It may be displayed together with the sentence.

(5-5) In the above embodiment, a case has been described in which a program guide is provided for a specific group such as a family, but the present invention is not limited to this.
For example, the present invention can be applied to a case where a user registers as a member in a public scene. That is,
The user registers the member profile using the remote commander 5A as shown in FIG. The remote commander 5A includes movement keys K14, K11, K15, and K13 for moving a cursor on the monitor screen up, down, left, and right, a decision key K11 for deciding an item designated by the cursor, and numerals and symbols for inputting numbers and symbols. And a symbol key K9.

In this case, member registration is performed in the same procedure as the profile registration procedure described above with reference to FIG.
That is, when entering the personal ID character selection procedure in step SP3 of FIG. 19, the CPU 29 (FIG. 2) displays an interactive screen as shown in FIG. In accordance with this screen, the user inputs the member number using the number / symbol key K9 of the remote commander 5A, and also inputs the age and gender, and also the personal profile data shown in FIGS. Enter As a result, the profile of the user is registered together with the member number.

When a plurality of specific users registered in this way are grouped to form a group model, the user first displays a menu on the monitor screen, and selects an item for group registration. When the group registration is selected, an interactive screen as shown in FIG. 35A is displayed on the monitor screen. The user first registers the ID number of the group on the screen, and then sequentially inputs the member numbers of the users belonging to the group using the numeric / symbol keys K9. When the registration of a plurality of users is completed in this way, the cursor is moved to the end position and the enter key K11
By operating, the monitor screen is switched to a screen as shown in FIG.

On this screen, the members of the registered group are displayed. The user confirms the list display, and if there is an error, moves the cursor to the member number and operates the enter key K11 to renew the member number. Can be entered. If there is no correction, the determination key K11 is directly operated to terminate the group registration and confirmation processing.

Thus, in a public scene where an unspecified number of users exist, a group user model classified by group ID can be easily formed while maintaining confidentiality.

(5-6) In the above embodiment, the case where the specific user requests his / her program guide by inputting his / her own ID using the remote commander 5 has been described. The invention is not limited thereto, and various collation methods such as voice recognition or fingerprint collation can be applied.

(5-7) In the above-described embodiment, the case where the maximum mode, the minimum mode, and the simple average mode are set as the group user model formation mode has been described. However, the present invention is not limited to this. Various formation techniques can be applied, such as weighting the specific user model according to the life scene and forming a group user model at a ratio corresponding to the weight.

(5-8) In the above embodiment, the case where the present invention is applied to an apparatus for retrieving digital satellite broadcast programs has been described. However, the present invention is not limited to this. The present invention can be widely applied to various information retrieval apparatuses such as information retrieval, retrieval of package information such as a mini disk (MD) and a digital video disk (DVD). Digital video disk device 8
In the case where the present invention is applied to the
The digital video disk (DVD) 8 and the mini disk (MD) 9 each store a plurality of disks,
It has a so-called disk changer for selecting a disk in accordance with a user's designation, and the package information (disk title and its genre, etc.) of each stored disk is transmitted via a control line to the SRA of the reception decoding device (IRD) 2.
M36 (FIG. 2), and in the same manner as in the case described above with reference to FIG. 3, a disc suitable for a specific user is selected from a plurality of discs based on the basic selection preference unique to the user, and the monitor screen is displayed. Guide display. As a result, the user selects a disc that suits him from the list displayed on the monitor screen, and responds accordingly to the IRD.
2 is sent to the digital video disk device 8 or the mini disk device 9, and the corresponding disk is selected. By the way, the package information of each disk (the title of the disk and its genre) is recorded on each disk in advance, and this is read by the disk technician, or when the user loads each disk into the disk technician, May be input.

In addition to selecting a disc suitable for the user from a plurality of discs, all programs or songs recorded on each disc are stored in advance in the SRAM 36 together with the genre, so that a specific user can be stored. A search may be made for a combined program or song. As a result, the disc containing the program or the program selected by the user is selected.

When searching for Internet information, the Internet information (such as the title of each homepage and its genre, etc.) input via a communication line is stored in S
The information may be stored in the RAM 36 and searched for based on the user's basic selection preferences.

(5-9) Further, in the above-described embodiment, the case where the information retrieval device is provided in the satellite broadcast receiving / decoding device (IRD) 2 has been described. However, the present invention is not limited to this. IRD2, VCR6, 7, DVD8 and M from computer built-in software and the computer
In the system for controlling D9, a system may be configured such that the personal computer selects a program or a program based on the user's basic selection preferences. In this case, a dedicated device (having at least the information search block shown in FIG. 3) may be provided instead of the personal computer.

[0107]

As described above, according to the present invention, the priority of information as a group composed of a plurality of users is determined based on a basic selection preference unique to a plurality of users, so that an enormous number of programs can be obtained. , It is possible to realize an information search method and apparatus capable of easily searching for information desired by the group.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a satellite broadcast receiving system using an information search device according to the present invention.

FIG. 2 is a block diagram showing a configuration of a reception decoding device including an information search device.

FIG. 3 is a block diagram showing an information search function block of the receiving and decoding apparatus.

FIG. 4 is a schematic diagram showing a data table centered on daily audience ratings by life scene.

FIG. 5 is a schematic diagram showing a specific example of a genre audience rating by life scene.

FIG. 6 is a schematic diagram illustrating a specific example of a genre audience rating for each life scene.

FIG. 7 is a schematic diagram illustrating a specific example of a genre audience rating for each life scene.

FIG. 8 is a schematic diagram illustrating a specific example of a genre audience rating by life scene.

FIG. 9 is a schematic diagram illustrating a specific example of a genre audience rating by life scene.

FIG. 10 is a schematic diagram illustrating a specific example of a genre audience rating by life scene.

FIG. 11 is a schematic diagram illustrating a configuration of an age-sex / life-scene-specific / genre-specific audience rating data table.

FIG. 12 is a schematic diagram illustrating a genre audience rating by life scene according to a first cluster.

FIG. 13 is a schematic diagram illustrating a user model based on “viewing rate data” and “well-viewed video data”.

FIG. 14 is a schematic diagram illustrating a specific example of video data often viewed by age.

FIG. 15 is a schematic diagram illustrating a user model based on “viewing rate data” and “frequently viewed video data”.

FIG. 16 is a schematic diagram illustrating a user model based on “viewing rate data” and “frequently viewed video data”.

FIG. 17 is a schematic diagram illustrating a specific example of video data that is often viewed for each cluster.

FIG. 18 is a plan view showing an operation key arrangement of the remote commander.

FIG. 19 is a flowchart showing a profile registration processing procedure.

FIG. 20 is a schematic diagram illustrating a personal ID character selection screen.

FIG. 21 is a schematic diagram illustrating an interactive screen.

FIG. 22 is a schematic diagram illustrating an interactive screen.

FIG. 23 is a schematic diagram illustrating an interactive screen when a profile relating to an attitude value factor is input.

FIG. 24 is a schematic diagram illustrating an interactive screen when a profile relating to an attitude value cluster is input.

FIG. 25 is a schematic diagram illustrating a data table of a specific user.

FIG. 26 is a schematic diagram illustrating knowledge base data with an identifier for linear interpolation;

FIG. 27 is a schematic diagram for explaining data conversion by a corresponding function;

FIG. 28 is a schematic diagram illustrating an interactive screen.

FIG. 29 is a schematic diagram illustrating a display example of a selected program.

FIG. 30 is a schematic diagram illustrating a screen display example at the time of group member registration.

FIG. 31 is a characteristic curve diagram for explaining selection of a group user model formation mode.

FIG. 32 is a schematic diagram illustrating another display example of a selected program.

FIG. 33 is a plan view showing an example of an operation key arrangement of a remote commander.

FIG. 34 is a schematic diagram illustrating a user member registration screen in a public scene.

FIG. 35 is a schematic diagram illustrating a group registration screen in a public scene.

[Explanation of symbols]

2 reception decoding device, 4 monitoring device, 5 remote commander, 11 knowledge base unit, 12 user profile input unit, 13 user model forming unit, 14
... User model storage / utilization unit 17, EPG data input unit 19, information filter unit 50, recommended information display unit 53, group user model formation unit

Claims (12)

[Claims] 1. A general user selection preference data comprising information items and a classification item based on a general user group and / or information use status classified according to a user attribute, and basic information selection preference data of a specific user. Forming a specific user model specific to the specific user; registering the specific user model formed for each of a plurality of users in association with each user; and one or more specific users of the plurality of specific user models An information search method characterized by searching for information suitable for one or more specific user models based on a model. 2. The information retrieval method according to claim 1, wherein the information retrieval method forms a group user model based on the plurality of specific user models, and retrieves the information based on the group user model. Information retrieval method. 3. The information search method according to claim 2, wherein the information search method stores the group user model and searches for the information by reading the group user model in response to a user request. 4. The information search method according to claim 2, wherein the method of forming the group user model is a maximum value mode for selecting a maximum value of the plurality of specific user models. 5. The information search method according to claim 2, wherein the method of forming the group user model is a minimum value mode for selecting a minimum value of the plurality of specific user models. 6. The information search method according to claim 2, wherein the method of forming the group user model is an average mode for selecting an average value of the plurality of specific user models. 7. A storage means for storing general user selection preference data comprising general user groups classified according to user attributes and / or classification items based on information usage status and information contents, and said general user preference data read from said storage means. Specific user forming model means for forming the specific user model specific to the specific user based on the general user selection preference data and the basic information selection preference data of the specific user; and the specific user model formed for each of a plurality of users. Registering means for registering corresponding to each user, searching for information suitable for the one or more specific user models based on one or more specific user models among the plurality of specific user models An information retrieval device characterized by performing: 8. The information retrieval apparatus further comprises: group user model forming means for forming a group user model based on the plurality of specific user models registered by the registration means, and based on the group user model. The information retrieval apparatus according to claim 7, wherein the information is retrieved by using the information retrieval apparatus. 9. The information retrieving apparatus comprises a group user model storage means for storing the group user model, and reads the information by reading the group user model from the group user model storage means in response to a user request The information search device according to claim 8, wherein the search is performed. 10. The information retrieval apparatus according to claim 8, wherein said group user model forming means selects a maximum value of said plurality of specific user models to form a group user model. 11. The information retrieval apparatus according to claim 8, wherein said group user model forming means selects a minimum value of said plurality of specific user models to form a group user model. 12. The information retrieval apparatus according to claim 8, wherein said group user model forming means selects a mean value of said plurality of specific user models to form a group user model.