JP2006121611A - Telephone system, telephone system management apparatus, advertisement content distribution method, advertisement content distribution program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a telephone system, telephone system management apparatus, and advertisement content distribution method in which advertisement contents can also be effectively distributed to telephone terminals utilized by a number of unspecific persons and communication privacy can also be protected. <P>SOLUTION: A telephone system management apparatus 101 controls communication between telephone terminals A, B connected to a communication network 102, selects any advertisement contents out of a plurality of advertisement contents A, B, C registered in advance and distributes the selected advertisement contents to the telephone terminals A, B in use. The telephone system management apparatus comprises a video identification unit 107 for estimating, from video images transmitted from the telephone terminals A, B, user attribute elements of users utilizing the telephone terminals A, B and an audio identification unit 108 for estimating the user attribute elements from audio transmitted from the telephone terminals A, B. Estimated results of both the identification units are used to determine the user attribute elements and the advertisement contents are selected in accordance with the user attribute elements. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a telephone system for distributing advertising content to a telephone terminal, a telephone system management apparatus and an advertising content distribution method for configuring the telephone system, an advertising content distribution program for the telephone system, and a computer-readable recording medium. About.

  In recent years, incentive advertising services have been provided in which advertising content is distributed to a telephone terminal that is being used, and the advertiser bears communication costs according to the time when the user of the telephone terminal views the advertising content. In this type of incentive advertising service, the advertiser is required to effectively view the advertising content to be distributed to the advertising target person, and the user is required to distribute useful advertising content. Therefore, various telephone systems have been proposed in order to realize these requirements.

  For example, there is a telephone system that distributes advertising content to each videophone terminal constituting the multipoint video conference system via a multipoint connection unit (MCU) (see Patent Document 1). However, in the telephone system described in Patent Document 1, since the same advertising content is sent to each videophone terminal by the MCU, it becomes a so-called dripping advertisement, and the advertising content is hardly noticed by the participants of the conference, and the advertising effect is sufficiently expected. Can not.

  Therefore, conventionally, a telephone system has been proposed in which one of advertisement contents is selected from a plurality of advertisement contents registered in advance according to a predetermined standard, and the selected advertisement content is distributed to a telephone terminal.

  For example, a plurality of advertisement contents and personal information of a user associated with a telephone number are registered in advance, and the telephone system management device refers to the personal information of the user corresponding to the telephone number by the calling / called number In addition, there is a telephone system that selects any advertisement content according to the personal information and distributes it to the telephone terminal having the telephone number (see Patent Document 2, Patent Document 3, and Patent Document 4). However, when distributing advertising content to telephone terminals used by an unspecified number of people, such as telephone terminals shared by family members or companies, the telephone system as described in Patent Document 2 Since there are many cases where users who actually use telephone terminals do not match, advertising content cannot be effectively streamed. Further, in the telephone system as described in Patent Document 2, since it is necessary to register personal information in advance, it is not possible to provide an incentive advertising service to users who use public telephone terminals.

  Furthermore, as a telephone system having another configuration, a voice identification unit for identifying the content of a call between telephone terminals is provided, and a plurality of pre-registered advertising contents are selected according to the content of the call estimated by the voice identification unit. Some advertisement contents are selected (see Patent Document 5 and Patent Document 6).

JP 2002-16897 A JP-A-11-55408 JP 2002-290565 A JP 2003-060791 A JP 2002-165193 A JP 2002-271507 A

  However, in the telephone system described in Patent Document 5, advertising content can be effectively distributed to telephone terminals used by an unspecified number of people, but the content of the call is estimated by the voice identification unit. Therefore, there is a problem that the secret of communication is not protected and the privacy of the user is infringed. If the voice identification unit is installed on the telephone terminal side, the burden on the user who purchases the telephone terminal increases, and the user needs to update the recognition dictionary necessary for voice identification. However, the voice identification rate of the telephone system gradually decreases, and there is a possibility that appropriate advertising content cannot be distributed.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a telephone system and a telephone system management apparatus capable of effectively distributing advertisement contents to telephone terminals used by an unspecified number of persons and protecting the communication secret. And an advertising content distribution method, and a program and a recording medium thereof.

  As a first means for solving the above-mentioned problems, the present invention provides a plurality of telephone terminals, a telephone system management apparatus that controls communication between these telephone terminals, and connects the telephone system management apparatus and the telephone terminal. In a telephone system comprising a communication network, wherein the telephone system management apparatus selects any one of a plurality of pre-registered advertising contents and distributes the selected advertising contents to the telephone terminal in use The telephone system management device includes a voice identification unit that estimates a user attribute element of a user who uses the telephone terminal from the voice of the user, and the advertisement is determined according to the user attribute element estimated by the voice identification unit. A configuration for selecting content was adopted.

  Here, the user attribute element is a characteristic element possessed by the user who uses the telephone terminal, and includes, for example, sex, age, race, nationality, and the like.

  According to the configuration of the first means, the telephone system management device includes a voice identification unit that estimates a user attribute element of a user who uses the telephone terminal from the voice of the user, and is estimated by the voice identification unit. Since the advertising content is selected according to the user attribute element, the advertising content targeting the owner of the estimated user attribute element is distributed to the telephone terminal without estimating the content of the call. For example, the advertising content attribute element of the advertising subject is registered in association with the advertising content, and the telephone system management apparatus selects the advertising content in which the advertising user attribute element closest to the estimated user attribute element is associated. You just have to do it.

  In addition, as a second means for solving the above-described problem, the present invention includes a plurality of telephone terminals, a telephone system management apparatus that controls communication between these telephone terminals, the telephone system management apparatus, and the telephone terminal. A telephone comprising a communication network to be connected, wherein the telephone system management device selects any one of a plurality of pre-registered advertising contents and distributes the selected advertising contents to the telephone terminal in use In the system, the telephone system management device includes a video identification unit that estimates a user attribute element of the user from the video when the video that shows the user who uses the telephone terminal is transmitted from the telephone terminal. A configuration is adopted in which the advertisement content is selected according to the user attribute element estimated by the video identification unit.

  According to the configuration of the second means, since the telephone system management device estimates the user attribute element of the user who uses the telephone terminal from the video of the user, not only gender, age, race, but also, for example, User attribute elements that are difficult to identify by voice identification, such as makeup preference and ornament preference, are also estimated regardless of the content of the call, and the advertising content targeting the owner of the user attribute element thus estimated is the telephone terminal Delivered to. For example, the advertising content attribute element of the advertising subject is registered in association with the advertising content, and the telephone system management apparatus selects the advertising content in which the advertising user attribute element closest to the estimated user attribute element is associated. You just have to do it.

  Furthermore, as a third means for solving the above problem, in the second means, the telephone system management device estimates a user attribute element of a user who uses the telephone terminal from the voice of the user. It is possible to employ a configuration in which the user attribute element is estimated using both the voice identification unit and the video identification unit. According to this configuration, since the user attribute element is estimated by using both the voice identification unit and the video identification unit, the number of types of user attribute elements that can be estimated by the telephone system is increased, which is more effective for the telephone terminal. Advertising content is delivered. In addition, the user attribute elements that can be estimated redundantly in both identification units, for example, the gender, age, and race estimation results and the user attribute elements possessed by the actual user are improved in coincidence. The video may be either a moving image or a still image, and the execution order of the processing by the audio identification unit and the processing by the video identification unit is not limited.

  Furthermore, as a fourth means for solving the above-described problem, in any one of the first to third means, the telephone system management device is connected via a network including at least one of an IP network and a PSTN. The advertisement content can be acquired from the connected advertisement content holding device. According to this configuration, the advertising content holding device can be installed under the control of an advertiser or an advertising agency, and the advertiser or the like can update the advertising content to be distributed without passing through the telephone system maintainer. Can be done. For this reason, an advertiser etc. can update advertisement content according to sudden events, such as a time sale of a holiday, and can distribute it via a telephone system.

  Further, as a fifth means for solving the above problem, in the third or fourth means, the telephone system management apparatus performs an estimation process by the voice identification unit, an estimation process by the video identification unit, and the voice. It is possible to adopt a configuration in which any estimation process is selected and executed according to the terminal environment of the telephone terminal among the estimation processes using the identification unit and the video identification unit in combination. According to this configuration, since the telephone system automatically performs estimation processing according to the terminal environment, it is possible to distribute appropriate advertising content according to the terminal environment.

  Furthermore, as a sixth means for solving the above-described problem, in any one of the first to fifth means, the telephone system management device estimates the user attribute element for each of the telephone terminals that perform communication exchange, A configuration in which the advertisement content is selected for each of these telephone terminals can be employed. According to this configuration, the user attribute element is estimated for each telephone terminal, and the advertising content is selected accordingly, so that the advertising content is effectively distributed to any telephone terminal.

  Further, as a seventh means for solving the above-described problem, the present invention controls communication between telephone terminals connected to a communication network, and selects any one of advertisement contents from a plurality of advertisement contents registered in advance. In the telephone system management apparatus that selects and distributes the selected advertisement content to the telephone terminal that is in use, the telephone system management apparatus includes a voice identification unit that estimates a user attribute element of the user who uses the telephone terminal from the voice of the user. A configuration is adopted in which the advertisement content is selected according to the user attribute element estimated by the identification unit. According to this configuration, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the first means.

  In addition, as an eighth means for solving the above-described problem, the present invention controls communication between telephone terminals connected to a communication network, and selects any one of advertisement contents from a plurality of advertisement contents registered in advance. In the telephone system management apparatus that selects and distributes the selected advertisement content to the telephone terminal that is being used, when a video showing the user using the telephone terminal is transmitted from the telephone terminal, the user attribute element of the user The video identification unit for estimating the advertisement content is selected from the video, and the advertisement content is selected according to the user attribute element estimated by the video identification unit. According to this configuration, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the second means.

  Furthermore, as a ninth means for solving the above-mentioned problem, in the eighth means, a voice identification unit for estimating a user attribute element of a user who uses the telephone terminal from the voice of the user is provided. And the video identification unit can be used in combination to estimate the user attribute element. According to this configuration, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the third means.

  Further, as a tenth means for solving the above problem, in the ninth means, the estimation process by the voice identification unit, the estimation process by the video identification unit, and the combined use of the voice identification unit and the video identification unit A configuration in which any one of the estimation processes is selected and executed according to the terminal environment of the telephone terminal can be employed. According to this configuration, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the fifth means.

  Furthermore, as an eleventh means for solving the above-mentioned problem, in any one of the seventh to tenth means, the user attribute element is estimated for each of the telephone terminals that perform communication exchange, and for each of these telephone terminals, the user attribute element is estimated. A configuration for selecting advertisement content can be employed. According to this configuration, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the sixth means.

  As a twelfth means for solving the above-mentioned problem, the present invention provides a plurality of telephone terminals, a telephone system management apparatus that controls communication between these telephone terminals, the telephone system management apparatus, and the telephone terminal. The telephone system management device includes a communication network to be connected, and the telephone system management device selects any one of a plurality of advertisement contents registered in advance, and distributes the selected advertisement contents to the telephone terminal that is using the advertisement contents. In the advertising content distribution method using the telephone system, the telephone system management apparatus estimates the user attribute element of the user who uses the telephone terminal from the voice of the user, and User attribute element of this user when a video showing the user using this telephone terminal is transmitted from the telephone terminal According to the terminal environment of the telephone terminal, one of a step of video identification estimated from the video and a step of estimating the user attribute element by using both the voice identification and the video identification is selected. I tried to run. According to this method, the advertising content is effectively distributed to the telephone terminal in use by the same effect as the fifth means.

  As a thirteenth means for solving the above problem, the present invention can be an advertising content distribution program for causing a telephone system to execute the advertising content distribution method described in the twelfth means.

  As a fourteenth means for solving the above problem, the present invention can be a computer-readable recording medium on which the advertising content distribution program of the thirteenth means is recorded.

  As described above, according to the present invention, since the advertising content is distributed to the telephone terminal without identifying the content of the call, the advertising content can be effectively applied to the telephone terminal used by an unspecified number of persons. Can be distributed, and a telephone system, a telephone system management device, and an advertisement content distribution method capable of protecting the secret of communication can be provided.

  Embodiments of a telephone system, a telephone system management apparatus, and an advertisement content distribution method according to the present invention will be described with reference to the drawings. Each drawing conceptually shows only the part related to the present invention among the configurations of the telephone system, the telephone system management apparatus and the advertisement content distribution method according to the present invention.

  As shown in FIG. 1, the telephone system connects a plurality of telephone terminals A and B, a telephone system management apparatus 101 that controls communication between these telephone terminals, and a telephone system management apparatus 101 and telephone terminals A and B. Communication network 102.

  The telephone terminals A and B are provided with a television camera and a display screen, generate video data by copying a user in use, transmit this together with audio data, and reproduce externally input video data and audio data. Has a telephone function. As the telephone terminals A and B, for example, a home videophone, a personal computer terminal having an IP (Internet Protocol) videophone function, a mobile phone terminal, and a PHS terminal can be used. As the telephone terminals A and B, normal telephone terminals capable of only voice calls can be used.

  The telephone terminals A and B can be used when logging in to the telephone system management apparatus 101. As this login method, a user ID or password registered in advance in the telephone system management apparatus 101 is input and transmitted to the user from the telephone terminals A and B, and this is performed by the telephone system management apparatus 101 authenticating. Alternatively, the terminal-specific information can be automatically transmitted from the telephone terminals A and B, and the telephone system management apparatus 101 can authenticate the terminal-specific information.

  A PSTN (Public Switched Telephone Network) or IP (Internet Protocol) network is used as the communication network 102, and the telephone terminals A and B perform IP communication. The communication network 102 can be a satellite communication network, a mobile line switching network, or a network in which these are integrated, and the communication protocols of the telephone terminals A and B can be changed as appropriate.

  The telephone system management apparatus 101 has a function of controlling communication between the telephone terminals A and B and a plurality of advertisement contents A, B, and C registered in advance in the advertisement content holding apparatus 104 connected via the IP network 103. It has a function of selecting any one of the advertisement contents from among them and distributing the selected advertisement contents to the telephone terminals A and B that are in use. The advertisement content holding device includes an external storage or the like. In this embodiment, the content holding device 104 is provided outside the telephone system, both conceptually and physically, but the content holding device 104 may be provided in the telephone system management device 101. Each advertisement content includes moving image content, still image content, audio content, content combining these, and the like. Further, a network including any of an IP network, a PSTN network, or a dedicated line can be used for connection between the telephone system management apparatus 101 and the advertisement content holding apparatus 104. In particular, it is preferable to use a network including at least one of an IP network and a PSTN in that an existing telephone system that is widely spread can be used. Of course, it is possible to use a network including both an IP network and a PSTN.

  The telephone system management apparatus 101 conceptually includes a communication control unit 105, a user authentication unit 106, a video identification unit 107, a voice identification unit 108, a main control unit 109, a content control unit 110, and a mixing unit 111. And a user management unit 112 and an incentive information unit 113.

  The communication control unit 105 has a communication exchange function for exchanging communication between the telephone terminals A and B, and a function for controlling communication between the telephone system management apparatus 101 and the telephone terminals A and B.

  The user authentication unit 106 performs authentication based on login information such as user IDs and passwords received from the telephone terminals A and B.

  The video identification unit 107 has a function of estimating a user attribute element of a user from the video when a video showing a user using the telephone terminal is transmitted from the telephone terminals A and B, and a telephone for communication exchange For each of the terminals A and B, user attribute elements that can be identified from the video of the user such as sex, age, race, makeup preference, and ornament preference are estimated.

  In the video identification unit 107, the input video is identified by the principal component identification (hereinafter referred to as PCA) of face shape features that has been conventionally used, and user attribute elements are estimated. For example, IEICE Technical Report, PRMU2003-142, pp. 13-18, Nov. By this method, likelihood calculation using a Gaussian mixture model (hereinafter referred to as GMM) is performed, and the likelihood of the gender of the user is obtained. In addition to gender, user attribute elements for which likelihood calculation is performed include, for example, preference for cosmetics required from video identification elements such as age, lipstick color, eye shadow color, presence of glasses, earrings, necklace There is a tendency of preference for ornaments required from video identification elements such as coloring, and the race and nationality of users required from video identification elements such as skin and hair color. Further, the likelihood of the user attribute element based on the upper limit probability can be obtained from the composite likelihood probability of these video identification elements. The user attribute element estimation processing in the video identification unit 107 will be described below by taking as an example the case of estimating the gender of a user based on likelihood calculation.

  First, the video identification unit 107 obtains the likelihood of the user's gender according to Equation 1 shown below.

  y is a face shape feature parameter identified by PCA, and v is a GMM parameter. In the case of gender likelihood calculation, the video identification unit 107 obtains the likelihood ν ′ in each of the male GMM parameter and the female GMM parameter, from which the probability m (m) that is male and the probability m (m that is female). f) The probability m (m · f) of indistinguishment is calculated. The basic probability calculation can be obtained, for example, from the likelihood distribution diagram of the Demster-Shafer theory. Further, the video identification unit 107 integrates the basic probabilities obtained from the male and female GMM parameters using Equation 2 shown below.

  Finally, the video identification unit 107 determines the gender of the user by comparing the probabilities obtained by Equation 2. Note that the likelihood calculation using Equation 1 and Equation 2 is an example, and other likelihood calculation methods can be used.

  The voice identification unit 108 has a function of estimating a user attribute element of a user who uses the telephone terminals A and B from the voice of the user. For each telephone terminal A and B to be exchanged, sex, age, race, etc. User attribute elements that can be identified from the voices of the users are estimated.

  In the voice identification unit 108, the input voice is identified by a conventionally used acoustic identification algorithm or the like, and a user attribute element is estimated. For example, Kakutani et al., “Changes in pitch frequency and formant frequency of Japanese vowels by age and gender”, Acoustical Society 24, 6 (1968), extracts acoustic features, and based on acoustic models The likelihood of user attribute elements such as gender and age is required. The user attribute element estimation process in the speech identification unit 108 will be described below by taking as an example a case where the gender of the user is estimated based on likelihood calculation. The user's age and the like can be estimated in the same manner as described below.

  First, the voice identification unit 108 obtains the likelihood of the user's gender according to the following Equation 3.

  x is an acoustic feature parameter identified by PCA, and w is a hidden Markov model (hereinafter referred to as HMM) parameter. In the case of gender likelihood calculation, the speech identification unit 108 obtains the likelihood w ′ in each of the male HMM parameter and the female HMM parameter, from which the probability m (m) that is male and the probability m (m that is female). f) The probability m (m · f) of indistinguishment is calculated. The basic probability calculation can be obtained, for example, from the likelihood distribution diagram of the Demster-Shafer theory. Further, the voice identification unit 108 integrates the basic probabilities obtained from the male and female HMM parameters using the above-described Expression 2. Finally, the voice identification unit 108 determines the gender of the user by comparing the probabilities obtained from Equation 2.

  Further, the speech identification unit 108 specifies a language by “development of multilingual speech identification engine” (acoustics of the Acoustical Society of Japan, pp. 115-116, March 2002) or Japanese Patent Application No. 2002-261672, etc. You may make it estimate about a user's nationality.

  The main control unit 109 performs processing on the telephone system management apparatus 101 and the telephone terminals A and B and the advertising content holding apparatus, processing instructions in each part of the telephone system management apparatus 101, registration and delivery of processing results, calling of the telephone terminal, and the like. Run and control each configuration of the phone system.

  Further, the main control unit 109 selects any one of the advertisement contents A, B, and C registered in advance according to the user attribute element estimated by the video identification unit 107 for each telephone terminal. A function for selecting one of the advertisement contents A, B, and C registered in advance according to the user attribute element estimated by the voice identification unit 108, a video identification unit 107, A user attribute element is estimated by using the voice identification unit 108 in combination, and one of advertisements A, B, and C registered in advance in the advertisement content holding device 104 according to the estimated user attribute element And a function for selecting contents. Hereinafter, an example of advertisement content selection processing in the main control unit 109 will be described.

  First, the main control unit 109 calculates a basic likelihood probability (estimation result) for each user attribute element obtained by the video identification unit 107 and a basic likelihood probability (estimation) for each user attribute element obtained by the speech identification unit 108. Result) and the upper limit probability for the user attribute element is calculated for each identical user attribute element.

  That is, the main control unit 109 uses Equation 2 to calculate the basic likelihood probability for each user attribute element obtained in the video identification unit 107 and the basic likelihood probability for each user attribute element obtained in the speech identification unit 108. Then, based on the integrated likelihood probability, the upper limit establishment is calculated using Equation 4 below.

  The main control unit 109 determines each user attribute element such as the user's gender from the result of the upper limit probability obtained by Expression 4. Then, the main control unit 109 selects, from the advertising contents A, B, and C, the advertising content in which the attribute element of the advertising subject closest to the determined user attribute element is associated. Specifically, the advertisement contents A, B, and C are associated with each other as an attribute element of the advertising target by using or combining any one of the user attribute elements determined by the main control unit 109. Registered in the holding device 104, the main control unit 109 refers to the attribute element of the advertising target person of each advertising content based on each determined user attribute element, and is associated with the closest advertising target person attribute element Ad content is selected. For example, when the main control unit 109 determines that the “sex” is “male” and the “age” is “30s” among the user attribute elements of the telephone terminal A, the attribute element of the advertising target person of the advertising content A However, (male) is associated with “male”, (age) is associated with “20s”, and the attribute elements of the advertising target of advertisement content B are (male) as “male” and (age) as “30s” If the attribute elements of the advertising target of the advertising content C are associated with (female) as “female” and (age) as “30s”, the main control unit 109 The advertisement content B is selected as the advertisement content to be distributed.

  Here, the main control unit 109 uses an estimation process by the video identification unit 107, an estimation process by the audio identification unit 108, and a combination of the audio identification unit 107 and the video identification unit 108 according to the terminal environments of the telephone terminals A and B. Among the estimation processes, one of the estimation processes is selected and executed according to the terminal environment of the telephone terminals A and B. Specifically, the main control unit 109 determines the presence / absence of video transmission by the IP videophone function for each of the telephone terminals A and B, and automatically performs estimation processing by the video identification unit 107 for telephone terminals without video transmission. Skip to. In this case, the advertisement content to be distributed is a dedicated audio content, an audio information portion used in the moving image content, or the like. When the estimation process by the video identification unit 107 is selected, the main control unit 109 determines each user attribute element using the basic likelihood probability calculated by the video identification unit 107 as the above-described upper limit probability, and selects advertisement content. In addition, when the estimation process by the voice identification unit 108 is selected, each user attribute element is determined with the basic likelihood probability calculated by the voice identification unit 108 as an upper limit probability, and the advertisement content is selected.

  The content control unit 110 controls transmission / reception of each advertising content to / from the content holding device 104 and has a database (hereinafter referred to as DB) for storing each advertising content transmitted to the content holding device 104. The advertising content selected by the main control unit 109 in response to an instruction from the control unit 109 is sent to the mixing unit 111. The accumulation of advertisement content is executed when the advertisement content is updated in the content holding device 104.

  The mixing unit 111 performs a reproduction process of the advertisement content received from the content control unit 110, and sends the advertisement content subjected to the reproduction process in response to an instruction from the main control unit 109 to a distribution destination telephone terminal.

  The user management unit 112 conceptually includes user information 114, video identification information 115, audio identification information 116, and maximum likelihood information 117 as a DB.

  As conceptually shown in FIG. 2, the user information 114 includes a maintenance person registration unit registered by a telephone system maintenance person, a system automatic registration unit automatically registered by the telephone system management apparatus 101, and a user registration. A user registration unit.

  The maintenance person registration unit includes a user ID (U1), a password (U2), a URI (Uniform Resource Identifier) (U3), and a telephone number (U4). Each item becomes unique information for each user by the maintenance person after a contract for incentive advertising service by viewing advertisement contents is established between the maintenance person of the telephone system and the user who uses the telephone terminals A and B. Registered in

  The system automatic registration unit includes an IP address (U5), a distribution content identifier (U6), and a user status (U7). In the cell of the IP address (U5), the IP addresses of the telephone terminals A and B are registered at the start of use of the telephone terminals A and B. In the distribution content identifier (U6) cell, the content identifier (A, B, C) of the advertisement content distributed to the telephone terminals A and B is registered when the distribution of the advertisement content is completed. In the cell of the user status (U7), the states of the telephone terminals A and B such as “login”, “logout”, and “busy” are registered.

  The user registration unit includes a user name (U8) and a distribution request (U9). The name of the user who uses the telephone terminals A and B is registered in the cell of the user name (U8). In the cell of the distribution request (U9), the presence / absence of a distribution request for advertising content in the telephone terminals A and B is registered.

  As conceptually shown in FIG. 3, the video identification information 115 includes user ID (U1), gender likelihood (U11), age likelihood (U12), and nationality likelihood (U13). Items include a likelihood (element Z) for each attribute element and a content identifier (U14). In the user ID (U1) cell, a user ID of a user who uses the telephone terminal that is the transmission source of the video identified by the video identification unit 107 is registered. The cell of likelihood (element Z) for each user attribute element such as gender likelihood (U11), age likelihood (U12), and nationality likelihood (U13) was calculated by the video identification unit 107. Each likelihood is registered. In the content identifier (U14) cell, the content identifier of the advertising content selected by the main control unit 109 based on the estimation process by the video identification unit 107 is registered.

  As conceptually shown in FIG. 4, the voice identification information 116 includes user ID (U1), gender likelihood (U21), age likelihood (U22), and nationality likelihood (U23). Items include a likelihood (element Z) for each attribute element and a content identifier (U24). In the user ID (U1) cell, the user ID of the user who uses the telephone terminal that is the voice transmission source identified by the voice identification unit 108 is registered. The cell of likelihood (element Z) for each user attribute element such as gender likelihood (U21), age likelihood (U22), and nationality likelihood (U23) is calculated by the voice identification unit 108. Each likelihood is registered. In the cell of the content identifier (U24), the content identifier of the advertising content selected by the main control unit 109 based on the estimation process by the voice identification unit 108 is registered.

  As conceptually shown in FIG. 5, the maximum likelihood information 117 includes a user ID (U1), a gender maximum likelihood (U31), an age maximum likelihood (U32), and a nationality maximum likelihood ( The items include a maximum likelihood (element Z) for each user attribute element such as (U33) and a content identifier (U34). In the cell of the user ID (U1), the user ID of the user who uses the telephone terminal of the video and audio transmission source identified by the video identification unit 107 and the audio identification unit 108 is registered. In the cell of the maximum likelihood (element Z) for each user attribute element such as the maximum likelihood (U31) of the sex, the maximum likelihood of the age (U32), and the maximum likelihood of the nationality (U33), a voice identification unit 107 is provided. And the video identification unit 108 are combined, the upper limit probability calculated by the main control unit 109 is registered, and the main control unit 109 is registered in the cell of the content identifier (U34) based on the upper limit probability. The content identifier of the advertising content selected by is registered. Further, when the main control unit 109 selects the estimation process by the video identification unit 107, the basic likelihood probability calculated by the video identification unit 107 is registered as the above-described upper limit probability in the gender maximum likelihood (U31) and the like. Then, the content identifier of the content identifier (U14) of the video identification information 115 is registered in the cell of the content identifier (U34). The same applies when the main control unit 109 selects the estimation process by the voice identification unit 108.

  As conceptually shown in FIG. 6, the incentive information unit 113 includes the date (B10), the call ID (B11), the originating IP address (B12), the originating URI / phone number (B13), and the destination IP address. (B14), incoming URI / phone number (B15), start time (B16), end time (B17), content identifier (B18), content playback start time (B19), and content playback end time ( B20) is included in the item.

  In the date (B10) cell, the date when the call session was performed is registered. A unique call ID assigned to each call session is registered in the cell of the call ID (B11). In the cell of the calling IP address (B12), the IP address of the calling telephone terminal among the telephone terminals A and B that performed the call session is registered. In the cell of the calling URI / phone number (B13), the URI / phone number of the calling telephone terminal is registered. On the other hand, the IP address of the called telephone terminal is registered in the cell of the called IP address (B14), and the URI / phone number of the called telephone terminal is registered in the cell of the called URI / phone number (B15). Is registered. The time when the call session is started is registered in the cell of the start time (B16), and the time when the call session is ended is registered in the cell of the end time (B17). In the content identifier (B18) cell, the content identifier of the advertising content distributed to the telephone terminals A and B is registered. In the cell of the content playback start time (B19), the time when the distributed advertisement content starts to be played is registered, and in the cell of the content playback end time (B20), the time when the playback of the advertising content ends is registered. The

  An advertisement content distribution method using a telephone system conceptually having the above configuration will be described below. The advertisement content can be distributed before the start of communication, at the time of calling, or at the time of holding, but here, as an example, a case where the advertising content is distributed at the time of holding the telephone will be described. In this example, in the user information 114 of the user management unit 112, predetermined information is registered in advance in the items of the maintenance person registration unit (U1 to U4) and the user registration unit (U8, U9).

  First, when a video showing a user using the telephone terminal A is transmitted from the login of the telephone terminal A, advertisement content based on the video identification is obtained through video identification in which the user attribute element of the user is estimated from the video. The process until the selection is completed will be described with reference to FIG.

  The telephone terminal A sends login information such as a password (U2) to the telephone system management apparatus 101 (S1).

  The main control unit 109 sends the received login information to the user authentication unit 106 and instructs the user authentication unit 106 to authenticate the telephone terminal A (S2).

  The user authentication unit 106 performs authentication based on the login information from the telephone terminal A, and sends the authentication result to the main control unit 109 (S3).

  Upon receiving the authentication result from the user authentication unit 106, the main control unit 109, when the telephone terminal A is authenticated for login, displays the IP address (U5), user status (U7), etc. of the telephone terminal A as a user management unit. The user information 114 of 112 is registered (S4).

  When the main control unit 109 completes the registration necessary for the user information 114, the main control unit 109 instructs the telephone terminal A to transmit a video showing the user, and the telephone terminal A receives the instruction by voice guidance or the like. The user is prompted to take a user image (S5).

  When the video is transmitted from the telephone terminal A, the main control unit 109 sends the video to the video identification unit 107. In response, the video identification unit 107 determines whether the user of the telephone terminal A is one or more from the position of the user image in the video. As a result, if the user is determined to be one, the user attribute of the user is determined. When element estimation processing is performed and it is determined that there are a plurality of users of the telephone terminal A, a user attribute element of each user is estimated (S6). When there is no video transmission from the telephone terminal A, the main control unit 109 skips the estimation process (S6) by the video identification unit 107 for the telephone terminal A. This skip process is executed after a predetermined time has elapsed since the completion of the process of (S5).

  Next, the video identification unit 107 registers the estimation result (such as the calculated likelihood) of the user attribute element in the record of the user ID (U1) created in the video identification information 115. (S7). Here, when the user attribute element of each user is estimated in (S6), the video identification unit 107 integrates the likelihood of the user attribute element of each user for each identical user attribute element, and each user attribute element after the integration As a user of the telephone terminal A.

  By the way, in the said integration process, when there is variation in sex, age, etc. between users, it may happen that user attribute elements become unidentifiable. In such a case, the video identification unit 107 cancels the integration process described above, and registers the user attribute element of each user in the video identification information 115 in association with the user ID (U1). In addition, after the video identification unit 107 determines that there are a plurality of users, or after the above integration process is stopped, the user is selected according to a predetermined criterion (for example, a user closest to the center of the video). This user can also be the user of telephone terminal A.

  When the registration to the video identification information 115 is completed, the video identification unit 107 notifies the main control unit 109 that the video identification process (S6) of the telephone terminal A has been completed (S8).

  Based on the likelihood registered in the video identification information 115, the main control unit 109 selects advertising content to be distributed to the telephone terminal A, and registers the identifier of the advertising content in the content identifier (U14) of the video identification information 115. (S9). Here, when the user attribute element of each user is registered in (S7), the main control unit 109 selects the advertisement content for each user and registers it in the same manner as described above.

  Next, assuming that the processing of (S1) to (S9) has been completed for each of the telephone terminals A and B, the user of the user who uses the telephone terminals A and B from the incoming and outgoing calls between the telephone terminals A and B The process from the voice identification in which the attribute element is estimated based on the voice of the user until the selection of the advertisement content based on the voice identification is completed will be described with reference to FIG.

  In order to connect to the telephone terminal A, the telephone terminal B makes a call to the telephone terminal A using the telephone number of the telephone terminal A or a URI (S11).

  The main control unit 109 identifies the telephone terminal A by the telephone number or URI received from the telephone terminal B, and confirms the information registered in the user status (U7) of the user information 114 of the user management unit 112 ( S12).

  The main control unit 109 executes the calling process of the telephone terminal A except when the information registered in the user status U7 is “logout” or “busy” (S13).

  The telephone terminal A responds to the call from the main control unit 109 (S14), and the main control unit 109 that has received this response executes a response notification to the telephone terminal B (S15).

  When the main control unit 109 confirms the call between the telephone terminals A and B, the date (B10) of the incentive information unit 113, the call ID (B11), the calling IP address (B12), The originating URI / phone number (B13), the destination IP address (B14), the destination URI / telephone number (B15), and the start time (B16) are registered (S16), and the communication control unit 105 is connected to the telephone terminals A and B. The communication exchange is started (S17).

  When the telephone terminals A and B are in a call state, the main control unit 109 notifies the telephone terminals A and B of a voice identification instruction, and the telephone terminals A and B receive the voice identification by voice guidance or the like. Prompts for pronunciation (S18). In the video identification process (S7), when it is determined that there are a plurality of users of the telephone terminal A, the main control unit 109 prompts the telephone terminal A to pronounce one representative. Of course, the voice identification unit 108 may automatically determine that the first voice transmitted from the telephone terminals A and B belongs to the user of the telephone terminals A and B. In the video identification process (S7), when the user of the telephone terminal A is selected based on a predetermined standard (for example, the user closest to the center of the video), the main control unit 109 , To encourage the user to pronounce.

  The main control unit 109 transmits the voice transmitted from the telephone terminals A and B to the voice identification unit 108. In response, the voice identification unit 108 estimates the user attribute elements of the telephone terminals A and B (S19). In addition, during a call, the user attribute element may be estimated assuming that the voices separated by a certain time are those of the users of the telephone terminals A and B. Note that the voice can be divided by several keywords selected so as not to grasp the contents of the call. For example, the voice identification unit 108 identifies the pronunciations of “I” and “Masu”, and the voice from the phone terminal A to the word “I” is pronounced until the word “mas” is pronounced. The user attribute element is estimated as that of the user.

  The voice identification unit 108 registers the estimation result (calculated likelihood etc.) of the user attribute elements of the telephone terminals A and B in the record of the user ID (U1) created in the voice identification information 116 (S20).

  After completing the registration in the voice identification information 116, the voice identification unit 108 notifies the main control unit 109 of the completion of voice identification (S21).

  After that, the main control unit 109 selects advertisement content to be distributed to the telephone terminals A and B for each telephone terminal based on the likelihood registered in the voice identification information 116, and sets the identifier of the advertisement content in the voice identification information 116. The upper limit probability of each user attribute element is calculated using the likelihood registered in the content identifier (U24) and further registered in the video identification information 115 and the audio identification information 116, and the maximum likelihood information 117 is set in each item. The final advertisement content to be registered and distributed to the telephone terminals A and B is determined for each telephone terminal in accordance with the registration information of the maximum likelihood information 117 and registered in the content identifier (U34) of the maximum likelihood information 117 ( S22). When the video identification process (S6) is skipped, the main control unit 109 uses the advertising content identifier registered in the content identifier (U24) of the audio identification information 116 as the content identifier (maximum likelihood information 117). U34). When content identifiers are registered for each user in the video identification information 115, the main control unit 109 skips the calculation of the upper limit probability and registers these content identifiers for each user.

  Next, a description will be given based on FIG. 9 until the telephone terminals A and B change from the call state to the hold state and the distribution of the selected advertisement content is completed. In the following description, “A” of the advertising content A is registered in the content identifier (U34) of the telephone terminal A, and “B” of the advertising content B is registered in the content identifier (U34) of the telephone terminal B. Assuming that

  When the telephone terminals A and B are in a call state (S17), a user using the telephone terminal B performs a holding operation on the telephone terminal B, and a telephone holding request is notified from the telephone terminal B to the main control unit 109 (S31). ).

  In response to this, the main control unit 109 notifies the telephone terminal A of a telephone hold instruction (S32), searches the user information 114 of the user management unit 112, and browses the distribution request (U9) of the telephone terminal B. The presence / absence of the advertisement content distribution request is confirmed (S33).

  When the main control unit 109 confirms that the telephone terminal B has the advertisement content distribution request, the content identifier “A” registered in the content identifier (U34) of the maximum likelihood information 117 of the telephone terminal A is registered in the content control unit 110. ”And an advertisement content distribution instruction (content instruction) having this content identifier is notified (S34).

  In response to this, the content control unit 110 acquires the advertising content A from the advertising content holding device 104 (S35) and sends it to the mixing unit 111 (S36). Note that if the advertising content A is stored in the DB of the content control unit 110, (S35) is skipped.

  The mixing unit 111 performs the reproduction process of the received advertisement content A, and when the reproduction process is completed, the mixing unit 111 notifies the main control unit 109 of the completion of reproduction preparation (S37).

  In response to this, the main control unit 109 notifies the content reproduction instruction to the telephone terminal A (S38), and the content identifier (B18) of the incentive information unit 113, the content reproduction start time (B19) and the like. Registration processing of information accompanying reproduction is started sequentially (S39).

  Upon receiving the content reproduction instruction, the telephone terminal A enters a content reproduction state with the mixing unit 111, the telephone terminal A outputs the content A, and the distribution of the advertising content is completed (S40). If content identifiers are registered in the maximum likelihood information 117 for each user, the advertising content with those content identifiers is sequentially delivered to the telephone terminal A. When telephone terminal A issues a hold request, content B is distributed to telephone terminal B in the same manner as described above.

  Next, the process from the hold state to the disconnected state between the telephone terminals A and B will be described with reference to FIG.

  While the content playback state is established between the telephone terminal A and the mixing unit 111 (S40), a hold release request is notified from the telephone terminal B to the main control unit 109 (S41).

  In response, the main control unit 109 registers information at the content reproduction end time (B20) of the incentive information unit 113 (S42), and notifies the telephone terminal A of a hold release instruction (S43).

  Receiving this, the telephone terminal A ends the communication with the mixing unit 111 and re-establishes the communication state with the telephone terminal B (S44).

  The telephone terminal B notifies the main control unit 109 of a disconnection request when the call with the telephone terminal A ends (S45).

  In response, the main control unit 109 registers information at the end time (B17) of the incentive information unit 113, and the disconnection between the telephone terminals A and B is completed (S46).

  Note that the telephone system according to the embodiment can be realized by an advertising content distribution program that causes a computer constituting the telephone system management apparatus to execute the processing of the above-described advertising content distribution method. The advertisement content distribution program can be recorded on a computer-readable recording medium.

  In the embodiment, the video identification unit 107 and the audio identification unit 108 have a function of executing a plurality of types of identification methods, or the video identification unit 107 and the audio identification unit 108 are a plurality of different identification methods. It is also possible to select a suitable identification method or identification unit according to the type of user attribute element to be estimated. Thereby, the consistency between the estimation result of each identification unit and the user attribute element of the actual user is further improved. For example, the video discriminating unit 107 is composed of a discriminating unit discriminating unit specialized in gender discrimination and an discriminating unit discriminating unit specializing in age discrimination, and the video / audio transmitted from the telephone terminals A and B are processed in parallel. Each of the identification units may be identified.

  As described above, in this embodiment, information necessary for providing an incentive advertising service by distributing advertising content to users who use the telephone terminals A and B is recorded in the incentive information unit 113. Thereby, for example, it becomes possible to provide a discount service, a free service, other benefits, etc. at a rate of one person per person according to the sex and age of the advertising target person.

  Further, in this embodiment, since the reproduction processing is performed in the mixing unit 111 for each selected advertisement content, the advertisement content can be distributed for each telephone terminal.

  Further, in this embodiment, since the user attribute elements such as gender and age are estimated by the voice identification unit 108, analysis into the contents of a call during a call is not necessary, and user privacy can be protected. Also, in this embodiment, user attribute elements of nationality and race can be estimated by voice identification and video identification, so an incentive advertising service is provided internationally via a public telephone terminal installed at an international airport It becomes possible to do.

  Further, in this embodiment, since the advertising content is selected by the video identification process or the voice identification process, it is not necessary to register the user attribute element in the user information 114, and even in a public telephone terminal used by an unspecified person. It is possible to provide a similar incentive advertising service. Of course, in the case of a home videophone terminal, since the user is fixed to some extent, if personal information and user attribute elements are registered in advance in the user information 114 for each family member, video identification processing and voice identification processing are performed. By combining them, it is possible to distribute advertising content more effectively.

  In addition, when a public telephone terminal is connected to the telephone system according to this embodiment, a service can be provided in which a privilege is given to one person at a rate of one person when a user of the public telephone terminal falls under an advertising target. . Specifically, when the attribute element of the advertisement target person of the advertisement content is set to (sex) “female” and (age) “teens”, it is estimated that the user attribute element corresponding to these is included. The content is such that a present is provided to 100 winners by lottery among users of public telephone terminals. In this case, the personal information input unit of the winner is provided in the incentive information unit 113, and a message indicating that the present service has been won from the display screen provided on the public telephone terminal after the end of the call and necessary for providing the user with a present What is necessary is just to make it output the input guidance for registering and receiving personal information. Of course, presents may be provided without exception regardless of lottery.

  As described above, if a public telephone terminal is connected to the telephone system according to this embodiment and the incentive advertisement service is performed, the person who distributes the advertisement sample and the present is made to stand in the town as in the past, and the person's Compared to the method of distributing advertising samples after judging whether passersby are subject to advertising by subjectivity, in addition to obtaining the same advertising effect, greatly reducing labor costs At the same time, it becomes possible to collect personal information of the advertising target person.

The block diagram which showed notionally the schematic whole structure of the telephone system which concerns on embodiment, and a telephone system management apparatus. The conceptual diagram which shows the data structure of the user information provided in the user management part of a telephone system management apparatus. The conceptual diagram which shows the data structure of the video identification information. The conceptual diagram which shows the data structure of the audio | voice identification information. The conceptual diagram which shows the data structure of the maximum likelihood information. The conceptual diagram which shows the data structure of the incentive information part. The partial sequence figure which shows from the login of a telephone terminal to the completion of a video identification process among the flows of the advertising content delivery method using the telephone system which concerns on embodiment. The partial sequence diagram which shows from the call of the same telephone terminal to the completion of the advertisement content selection processing. The partial sequence diagram which shows from the hold process between the telephone terminals to completion of an advertisement content delivery process. The partial sequence figure which shows from the cancellation | release hold process between the telephone terminals to the completion of a cutting process.

Explanation of symbols

101 telephone system management device 102 communication network 103 IP network 104 advertising content holding device 105 communication control unit 106 user authentication unit 107 video identification unit 108 audio identification unit 109 main control unit 110 content control unit 111 mixing unit 112 user management unit 113 incentive information Unit 114 user information 115 video identification information 116 voice identification information 117 maximum likelihood information

Claims (14)

  A plurality of telephone terminals; a telephone system management apparatus that controls communication between these telephone terminals; and a communication network that connects the telephone system management apparatus and the telephone terminal. The telephone system management apparatus is registered in advance. In the telephone system that selects any one of the plurality of advertisement contents and distributes the selected advertisement contents to the telephone terminal that is using the telephone content, the telephone system management device is configured to A telephone system comprising: a voice identification unit that estimates a user attribute element from the voice of the user; and selecting the advertisement content according to the user attribute element estimated by the voice identification unit.   A plurality of telephone terminals; a telephone system management apparatus that controls communication between these telephone terminals; and a communication network that connects the telephone system management apparatus and the telephone terminal. The telephone system management apparatus is registered in advance. In the telephone system that selects any one of the plurality of advertisement contents and distributes the selected advertisement contents to the telephone terminal that is using the telephone contents, the telephone system management device transfers the telephone terminal from the telephone terminal. When a video showing a user to be used is transmitted, a video identification unit that estimates a user attribute element of the user from the video is provided, and the advertisement content is displayed according to the user attribute element estimated by the video identification unit. A telephone system characterized by selection.   The telephone system management device includes a voice identification unit that estimates a user attribute element of a user who uses the telephone terminal from the voice of the user, and uses the voice identification unit and the video identification unit together to determine the user attribute element. The telephone system according to claim 2, wherein the telephone system is estimated.   The telephone system management device acquires the advertising content from an advertising content holding device connected through a network including at least one of an IP network and a PSTN. The phone system described in Crab.   The telephone system management device performs any one of the estimation process among the estimation process by the voice identification unit, the estimation process by the video identification unit, and the estimation process by using the voice identification unit and the video identification unit in combination. 5. The telephone system according to claim 3, wherein the telephone system is selected and executed according to a terminal environment of the terminal.   The said telephone system management apparatus estimates the said user attribute element for every said telephone terminal which carries out communication exchange, and selects the said advertisement content for each of these telephone terminals, It is any one of Claim 1 to 5 characterized by the above-mentioned. Phone system.   A telephone that controls communication between telephone terminals connected to a communication network, selects any advertising content from a plurality of pre-registered advertising contents, and distributes the selected advertising content to the telephone terminals in use In the system management device, a voice identification unit that estimates a user attribute element of a user who uses the telephone terminal from the voice of the user is selected, and the advertisement content is selected according to the user attribute element estimated by the voice identification unit A telephone system management device.   A telephone that controls communication between telephone terminals connected to a communication network, selects any advertising content from a plurality of pre-registered advertising contents, and distributes the selected advertising content to the telephone terminals in use The system management apparatus includes a video identification unit that estimates a user attribute element of the user from the video when the video that shows the user using the telephone terminal is transmitted from the telephone terminal, and the video identification unit estimates The telephone system management apparatus, wherein the advertisement content is selected according to the user attribute element that has been set.   A voice identification unit that estimates a user attribute element of a user who uses the telephone terminal from the voice of the user, and the user attribute element is estimated by using the voice identification unit and the video identification unit in combination. The telephone system management apparatus according to claim 8.   Any one of the estimation process by the voice identification unit, the estimation process by the video identification unit, and the estimation process by the combined use of the voice identification unit and the video identification unit is performed according to the terminal environment of the telephone terminal. The telephone system management apparatus according to claim 9, wherein the telephone system management apparatus is selected and executed.   The telephone system management apparatus according to any one of claims 7 to 10, wherein the user attribute element is estimated for each telephone terminal that performs communication exchange, and the advertisement content is selected for each telephone terminal.   A plurality of telephone terminals; a telephone system management apparatus that controls communication between these telephone terminals; and a communication network that connects the telephone system management apparatus and the telephone terminal. The telephone system management apparatus is registered in advance. In the advertising content distribution method using the telephone system, the method of selecting any one of the plurality of advertising contents and the step of distributing the selected advertising content to the telephone terminal that is in use, The telephone system management device transmits a voice identification step for estimating a user attribute element of a user who uses the telephone terminal from the voice of the user, and a video showing the user who uses the telephone terminal from the telephone terminal. A video identification step for estimating the user attribute element of the user from the video, and the voice identification. Serial of the steps of estimating the user attribute element in combination with image identification, one of the steps, advertisement content distribution method, characterized in that selects and executes in response to the terminal environment of the telephone terminal.   An advertisement content distribution program for causing a telephone system to execute the advertisement content distribution method according to claim 12.   A computer-readable recording medium on which the advertising content distribution program according to claim 13 is recorded.

Images