JP2011170622A - Content providing system, content providing method, and content providing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an adequate content on the basis of voice characteristics such as a manner of speaking which is peculiar to a user. <P>SOLUTION: A content providing system 100 includes a content storage part 126 which makes voice characteristics associated with a content which is favorably presented to a user having the voice characteristics and stores them, a voice characteristics detection part 104 which compares the value of a voice element of the user which is computed on the basis of inputted user voice data with index data for determining a tendency to a standard to detect the voice characteristics peculiar to the user, a content-selecting part 106 which selects the content which is correlated with the voice characteristics detected by the voice characteristics detection part 104 from among the content stored in the content storage part 126, and a presentation part 108 which presents the selected content. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a content providing system, a content providing method, and a content providing program.

  Japanese Patent Application Laid-Open No. 2008-170820 discloses an emotion type content storage unit that stores an emotion type and content in association with each other, and a feature amount calculation unit that calculates a feature amount of a voice input from the voice input unit. An emotion type determining means for determining an emotion type based on the calculated audio feature quantity, a content reading means for reading the content stored in association with the determined emotion type, and the read content as a content A content providing system including a transmission unit that transmits to a reproduction unit is described. In the technique described in the document, as a simple feature quantity independent of a language or a speaker, a prosodic component such as a volume or a pitch is used, and a basic statistic for a predetermined time (for example, the past one second) of the feature quantity is obtained. The speaker is currently speaking. Then, the degree of each emotion is obtained from the deviation from the steady state of speaking (for example, the basic statistics for the past 5 seconds).

JP 2008-170820 A

  In the technique described in Patent Document 1, only a temporary emotion of the user is determined. However, for example, when it is desired to present an effective advertisement or the like as content to the user, it is better to present the content related to the user's unique property than the content related to the emotion generated temporarily to the user. Is likely to pull. In the method of presenting content related to emotions temporarily generated to the user, the content is changed each time the user's emotion changes, and the user's interest cannot be increased by repeatedly presenting the content. In the conventional technology, it is not possible to provide appropriate content based on voice characteristics such as a user-specific way of speaking.

  An object of the present invention is to provide a content providing system, a content providing method, and a content providing program that solve the above-described problem that it is impossible to provide appropriate content based on voice characteristics such as a user-specific way of speaking. It is to provide.

According to the present invention,
Content storage means for storing the audio feature and the content to be presented to the user having the audio feature in association with each other;
A voice feature detecting means for detecting a voice feature unique to the user by comparing a value of the voice component of the user calculated based on the input voice data of the user with index data for judging a tendency with respect to a standard; ,
Content selection means for selecting content associated with the audio features detected by the audio feature detection means from among the contents stored in the content storage means;
Presenting means for presenting the content selected by the content selecting means;
Is provided.

According to the present invention,
A content providing method using a computer system including a content storage unit that stores an audio feature and content to be presented to a user having the audio feature in association with each other,
A voice feature detection step of detecting a voice feature unique to the user by comparing a value of the voice component of the user calculated based on the input voice data of the user with index data for determining a tendency with respect to a standard; ,
A content selection step of selecting content associated with the audio feature detected in the audio feature detection step from the content stored in the content storage means;
A presentation step of presenting the content selected in the content selection step;
A content providing method is provided.

According to the present invention,
Computer
Content storage means for storing a voice feature and a content to be presented to a user having the voice feature in association with each other;
A voice feature detecting means for detecting a voice feature unique to the user by comparing the value of the voice component of the user calculated based on the input voice data of the user with index data for judging a tendency with respect to the standard;
Content selection means for selecting content associated with the audio feature detected by the audio feature detection means from among the content stored in the content storage means;
Presenting means for presenting the content selected by the content selecting means;
As a result, a content providing program is provided.

  It should be noted that any combination of the above-described constituent elements and a conversion of the expression of the present invention between a method, an apparatus, a system, a recording medium, a computer program, etc. are also effective as an aspect of the present invention.

  According to the present invention, it is possible to provide appropriate content based on voice features such as a user-specific way of speaking.

It is a block diagram which shows an example of a structure of the content provision system in embodiment of this invention. It is a figure which shows an example of a structure of the parameter | index data storage part in embodiment of this invention. It is a figure which shows an example of a structure of the content memory | storage part in embodiment of this invention. It is a block diagram which shows the network structure containing the content provision system in embodiment of this invention. It is a flowchart which shows the process sequence of the content provision system in embodiment of this invention. It is a block diagram which shows the other example of a structure of the content provision system in embodiment of this invention. It is a block diagram which shows an example of a structure of the content provision system in embodiment of this invention. It is a figure which shows an example of a structure of the audio | voice element value accumulation | storage part in embodiment of this invention. It is a figure which shows the other example of a structure of the parameter | index data storage part in embodiment of this invention. It is a figure which shows the other example of a structure of the parameter | index data storage part in embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, the same constituent elements are denoted by the same reference numerals, and the description thereof is omitted as appropriate.

(First embodiment)
FIG. 1 is a block diagram illustrating an example of a configuration of a content providing system according to the present embodiment.
The content providing system 100 includes a voice recognition unit 102 (voice recognition unit), a voice feature detection unit 104 (voice feature detection unit), a content selection unit 106 (content selection unit), a presentation unit 108 (presentation unit), An acoustic model storage unit 120, a language model storage unit 122, an index data storage unit 124 (index data storage unit), and a content storage unit 126 (content storage unit) are included.

  The acoustic model storage unit 120 stores an acoustic model. The language model storage unit 122 stores a language model. These acoustic models and language models can be generally used.

  The speech recognition unit 102 is based on the acoustic model stored in the acoustic model storage unit 120 and the language model stored in the language model storage unit 122, and the acoustic score of words that are candidates for speech recognition results for the input speech data, A language score and a total score based on the acoustic score and the language score are calculated. For example, the voice recognition unit 102 selects a word having a high total score as a voice recognition result of input voice data.

  The voice recognition unit 102 calculates values of various voice elements of the user based on the input voice data. The speech element can be information indicating how the user speaks, such as various parameters extracted from speech data used when speech recognition is performed, and the speech recognition result itself. The voice element can be, for example, voice power, speech speed, acoustic score, language score, recognition result reliability, filler rate in the recognition result, and the like. The voice recognition unit 102 can also detect whether the user is male or female. The speech recognition unit 102 can calculate the values of these speech elements based on the user's speech data input during a predetermined time.

  The index data storage unit 124 stores index data for determining a tendency with respect to a standard for various speech elements. The index data can be calculated based on the voices of a plurality of speakers. In the present embodiment, the index data storage unit 124 stores, as index data, standard values that are standard values calculated based on the voices of a plurality of speakers for various speech elements. FIG. 2 is a diagram illustrating an example of the configuration of the index data storage unit 124. Here, the index data storage unit 124 stores male and female standard values for each voice element.

  Returning to FIG. 1, the voice feature detection unit 104 compares the value of the user's voice element calculated by the voice recognition unit 102 with the standard value of the voice element stored in the index data storage unit 124 for each voice element. Thus, a voice feature unique to the user is detected.

  For example, in this embodiment, when the voice recognition unit 102 detects that the user is a woman, the voice feature detection unit 104 sets the value of each voice element of the user to “female” shown in FIG. Compare with standard values. For example, when the value of the voice power of the user is larger than the standard value “a2” of “sound power” of “female” shown in FIG. 2, the tendency of the voice power of the woman is “high voice power”. Become. Similarly, if the value of the filler rate in the recognition result of the user is smaller than the standard value “f2” of the “filler rate in the recognition result” of “female” shown in FIG. The tendency of the filler rate in the result is “low filler rate in the recognition result”. Here, each standard value is not a unique value, but may be a range of values. Therefore, for example, as a standard value “a2” of “sound power” of “female”, the standard range of female voice power can be stored. In this way, when the voice of the user is much higher or lower than the standard, the voice feature of the user can be “high voice power” or “low voice power”.

  In addition, the voice feature detection unit 104 can detect the amount of deviation from the standard value by comparing the value of each voice element of the user with the standard value of each voice element stored in the index data storage unit 124. When the amount of deviation from the standard value is large, the user has a high degree of having the voice feature.

  The content storage unit 126 stores the audio feature and the content to be presented to the user having the audio feature in association with each other. The content to be presented to the user can be selected on the basis of the audio characteristics, assuming the content related to what the user with such a tendency is not good at, the content related to what the user is good at, and the content that is likely to be interesting.

FIG. 3 is a diagram illustrating an example of the configuration of the content storage unit 126. The content storage unit 126 includes a content column and an audio feature column.
For example, in the content “A (CM of a speaking class)”, the voice features are “speech power low”, “low acoustic model score”, “high filler rate in recognition result”, “high speech rate”, “recognition result” “Reliability” is stored in association with each other.

  The voice feature of the user is “high filler rate” because the user's voice data contains many fillers such as “e” and “ano”. That is, it is assumed that a user having such a voice feature is not good at speaking, and is likely to be weak at himself. For this reason, CM (commercial) in the classroom for speaking can be associated with the content presented to the user having such voice characteristics.

  In addition, the content “B (CM of the karaoke store)” stores “audio power high” and the like as the audio feature in association with each other. It is assumed that a user with high voice power is likely to be good at karaoke. Therefore, CM (commercial) of a karaoke store can be associated with the content presented to the user having such voice characteristics.

  Returning to FIG. 1, the content selection unit 106 selects the content associated with the audio feature detected by the audio feature detection unit 104 from the content stored in the content storage unit 126. The presentation unit 108 presents the content selected by the content selection unit 106 to the user.

  For example, as described above, when the voice feature of the user is “high voice power”, the content selecting unit 106 selects the content “B (karaoke shop) associated with the voice feature“ high voice power ”shown in FIG. CM) ”is selected. The presenting unit 108 presents the content “B (karaoke shop CM)” selected by the content selecting unit 106 to the user.

  The audio feature detection unit 104 can also output and notify the audio selection of each user in association with the amount of deviation from the standard value of each audio element, etc. For example, when there are a plurality of audio features of the user, the content selection unit 106 selects the content associated with the audio feature having the largest deviation from the standard value from the content stored in the content storage unit 126. be able to. In addition, for example, the content selection unit 106 selects content associated with each audio feature from the content stored in the content storage unit 126 in descending order of deviation from the standard value, and sequentially presents it to the user. You can also do it.

FIG. 4 is a block diagram showing a network structure including the content providing system 100 in the present embodiment.
This network structure includes a content providing system 100 and a user terminal device 200 connected to the content providing system 100 via a network 150. The user terminal device 200 can be a user's PC or the like. The user terminal device 200 can be configured by a personal computer, for example. The user terminal device 200 can be configured to include voice input means such as a microphone and display means such as a display. When voice data is input via the voice input unit of the user terminal device 200, the voice data is input to the voice recognition unit 102 (see FIG. 1) of the content providing system 100 via the network 150. When the presentation unit 108 of the content providing system 100 provides content, the content is input to the user terminal device 200 via the network 150 and displayed on the display unit of the user terminal device 200. Note that the voice data of the user of the user terminal device 200 is input to the voice recognition unit 102 of the content providing system 100 via a network other than the network 150 such as a telephone line if the correspondence with the user terminal device 200 is achieved. It can also be set as the structure made.

  Next, in this embodiment, a procedure from when audio data is input until content is provided will be described. FIG. 5 is a flowchart showing a processing procedure of the content providing system 100 according to the present embodiment.

  When voice data is input (step S100), the voice recognition unit 102 performs voice recognition processing (step S102). This process can be a normal voice recognition process.

  Subsequently, the voice feature detection unit 104 compares the value of the voice element detected by the voice recognition unit 102 with the standard value of each voice element stored in the index data storage unit 124 to detect the voice feature unique to the user. (Step S104). The content selection unit 106 selects content associated with the audio feature detected by the audio feature detection unit 104 from the content storage unit 126 (step S106). The presentation unit 108 provides the content selected by the content selection unit 106 to the user terminal device 200 via the network 150 (step S108).

  According to content providing system 100 in the present embodiment, the content to be provided to the user is determined according to the user-specific audio characteristics detected based on the values of the various audio elements of the user. For example, content is provided in accordance with user-specific audio features such as “Eh” and “Anno”, which often contain fillers, fast speech, and good voice. Therefore, it is possible to increase the possibility of providing content that attracts the user's interest. In addition, since the user-specific voice characteristics do not change so much, the same content can be repeatedly provided, and the user's interest can be enhanced.

  Moreover, in this Embodiment, as shown in FIG. 6, the content provision system 100 can also be set as the structure further including the user information acquisition part 110 which acquires user information. Here, the user information can include, for example, the sex and age of the user. For example, when the user information acquisition unit 110 acquires the user's gender as the user information, the voice feature detection unit 104 selects either the male or female standard value in the index data storage unit 124 based on the input user's gender. Can be used to detect the voice characteristics of the user.

  In addition, the index data storage unit 124 can store not only standard values for each gender as shown in FIG. 2, but also standard values for each age, for example. For example, when the user information acquisition unit 110 acquires the user's age as the user information, the voice feature detection unit 104 uses the standard value of the corresponding age in the index data storage unit 124 based on the input user's age. Thus, the voice feature of the user can be detected.

(Second Embodiment)
FIG. 7 is a block diagram showing an example of the configuration of the content providing system in the present embodiment.
In the present embodiment, the content providing system 100 is a first embodiment in that the content providing system 100 further includes an average value calculation unit 112 (average value calculation unit) and an audio element value accumulation storage unit 128 (audio element value accumulation storage unit). And different. Here, a configuration in which the content providing system 100 also includes the user information acquisition unit 110 is shown.

  In the present embodiment, the user information acquisition unit 110 can acquire user identification information as user information. The average value calculation unit 112 identifies the user based on the user identification information, and stores the value of the voice element of the user in the voice element value storage unit 128 in association with the identification information of the user. The average value calculation unit 112 can store the value of the voice element in the voice element value accumulation storage unit 128 in association with the date and time when the voice data is input, for example. Further, the average value calculation unit 112 refers to the voice element value accumulation storage unit 128 and calculates the average value of the voice element values of the user accumulated at different times. In the present embodiment, the audio feature detection unit 104 compares the average value of each audio element calculated by the average value calculation unit 112 with the standard value of the audio element stored in the index data storage unit 124. , Voice features unique to the user can be detected.

The voice element value accumulation storage unit 128 accumulates the value of the voice element of the user in association with the user identification information. In addition, FIG. 8 is a diagram illustrating an example of the configuration of the audio element value accumulation storage unit 128.
Here, the voice element value data of the user whose user identification information (user ID) is “0002f” is shown. This user, for example, accesses the content providing system 100 at least “2010/01/03”, “2010/01/10”, and “2010/01/12”, and transmits audio data via the audio recognition unit 102. You are typing. For this reason, various audio element values calculated based on the audio data at the time of each access are stored in the audio element value storage unit 128.

  In this case, the average value calculation unit 112 stores, for example, data of the voice element “voice power” in association with “2010/01/03”, “2010/01/10”, and “2010/01/12”, respectively. The average value is calculated from the values “a10”, “a11”, and “a12”. The voice feature detection unit 104 compares the average value calculated by the average value calculation unit 112 with the standard value of the voice element “voice power” stored in the index data storage unit 124, and compares the user's voice element “voice power”. ”Is detected as a standard, larger than a standard, or smaller.

  Also in this embodiment, the same effect as that of the first embodiment can be obtained. In the present embodiment, since the average value of the user's voice element values accumulated at different times is used, the user-specific voice characteristics can be stably detected.

  In the above example, the user information acquisition unit 110 acquires user identification information, and the average value calculation unit 112 identifies a user based on the user identification information. However, the content providing system 100 may be configured not to include the user information acquisition unit 110, and the average value calculation unit 112 may identify the user by voice authentication based on the user's voice.

  Each component of the content providing system 100 illustrated in FIG. 1 is not a hardware unit configuration but a functional unit block. Each component of the content providing system 100 is centered on an arbitrary computer CPU, memory, a program for realizing the components shown in the figure loaded in the memory, a storage unit such as a hard disk for storing the program, and a network connection interface. It is realized by any combination of hardware and software. It will be understood by those skilled in the art that there are various modifications to the implementation method and apparatus.

  As mentioned above, although embodiment of this invention was described with reference to drawings, these are the illustrations of this invention, Various structures other than the above are also employable.

  In the first embodiment, an example is shown in which the standard value of each voice element is stored in the index data storage unit 124 by sex or age, but the standard value may be stored for each of various other classifications. it can. For example, focusing on one of a plurality of speech elements, a standard value can be stored for each classification according to the value of the speech element.

  An example is shown in FIG. Here, standard values such as speech speed, acoustic score, language score, recognition result reliability, filler rate in recognition results, etc., which are other speech elements, are stored for each of the speech power of large, medium and small An example is shown. When voice data is input to the voice recognition unit 102 and the value of each voice element is calculated by the voice recognition unit 102, the voice feature detection unit 104 determines the voice power of the user's voice based on the voice power value of the user. Judge whether it falls under large, medium or small. Then, using the standard values of other audio elements associated with the corresponding audio power, it is determined whether these other audio elements are larger or smaller than the standard. For example, if the voice power value of the input voice data is a1 to a2, the voice power of this user is large, and the standard value “b3” is used when determining the user's speech speed. Used.

  In addition, the index data storage unit 124 can also store an index value for determining a tendency with respect to a standard, instead of the standard value of each voice element, as index data. An example is shown in FIG. Here, index values for determining speech power and speech speed as large (high), standard (no feature), and small (low) are stored as speech elements. When voice data is input to the voice recognition unit 102 and the value of each voice element is calculated by the voice recognition unit 102, the voice feature detection unit 104 has a voice power value of the user stored in the index data storage unit 124, for example. It is determined whether the index value is included in (high), standard (no feature), or small (low), and a voice feature unique to the user is detected.

  In addition, the index data storage unit 124 can also store data indicating variation and degree of dispersion of each voice element value of the voices of a plurality of speakers as index data. Based on the value of each voice element of the input voice data, the voice feature detecting unit 104 determines in which direction the voice element of the user is deviated from, for example, the average value of the voice element values of the voices of a plurality of speakers. In addition, based on the degree of deviation, etc., the voice feature of the user can be detected.

  Also, the voice feature detection unit 104 feeds back the value of each voice element calculated based on the user's voice data input from the voice recognition unit 102 to the index value storage unit 124, and the index of the index value storage unit 124. You can also update the data. Thereby, the parameter number of speakers used for the index data in the index value storage unit 124 can be increased, and index data with high accuracy can be constructed.

  Moreover, in the above embodiment, an example in which content is only provided to the user terminal device 200 of the user who has input the voice data has been shown. However, when the voice feature of the user is detected, the user has the voice feature. It can also be set as the structure which matches the advertiser and the user who are recruiting users. For example, each content in the content storage unit 126 can also be stored in association with the necessity of notification to the advertiser. In this case, after obtaining the user's approval when the advertiser has registered as a voice feature such as “high voice power” and notification is required, or when it is detected that the user's voice power is high It can also be set to transfer the user information to the advertiser.

  Furthermore, for example, “similarity with a certain voice actor” or the like can be used as a voice element. In this case, the content providing system 100 can determine the degree of similarity with the voice feature of the voice actor and set the content related to the voice actor to be presented if the similarity is high.

DESCRIPTION OF SYMBOLS 100 Content provision system 102 Speech recognition part 104 Voice feature detection part 106 Content selection part 108 Presentation part 110 User information acquisition part 112 Average value calculation part 120 Acoustic model storage part 122 Language model storage part 124 Index data storage part 126 Content storage part 128 Voice element value storage unit 150 Network 200 User terminal device

Claims (6)

Content storage means for storing the audio feature and the content to be presented to the user having the audio feature in association with each other;
A voice feature detecting means for detecting a voice feature unique to the user by comparing a value of the voice component of the user calculated based on the input voice data of the user with index data for judging a tendency with respect to a standard; ,
Content selection means for selecting content associated with the audio features detected by the audio feature detection means from among the contents stored in the content storage means;
Presenting means for presenting the content selected by the content selecting means;
Content providing system including The content providing system according to claim 1,
The content providing system, wherein the index data of the value of the voice element is calculated based on voices of a plurality of speakers. The content providing system according to claim 1 or 2,
A content providing system further comprising index data storage means for storing the index data of the value of the audio element. The content providing system according to any one of claims 1 to 3,
Voice element value storage means for storing the value of the voice element of the user in association with the identification information of the user;
The user is identified, the value of the voice element of the user is stored in the voice element value storage unit in association with the identification information of the user, and the average of the voice elements of the user stored at different times An average value calculating means for calculating a value;
Further including
The content providing system for detecting an audio feature unique to the user by comparing the average value of the audio elements calculated by the average value calculating unit with a standard value of the audio element. A content providing method using a computer system including a content storage unit that stores an audio feature and content to be presented to a user having the audio feature in association with each other,
A voice feature detection step of detecting a voice feature unique to the user by comparing a value of the voice component of the user calculated based on the input voice data of the user with index data for determining a tendency with respect to a standard; ,
A content selection step of selecting content associated with the audio feature detected in the audio feature detection step from the content stored in the content storage means;
A presentation step of presenting the content selected in the content selection step;
A content providing method including: Computer
Content storage means for storing a voice feature and a content to be presented to a user having the voice feature in association with each other;
A voice feature detecting means for detecting a voice feature unique to the user by comparing the value of the voice component of the user calculated based on the input voice data of the user with index data for judging a tendency with respect to the standard;
Content selection means for selecting content associated with the audio feature detected by the audio feature detection means from among the content stored in the content storage means;
Presenting means for presenting the content selected by the content selecting means;
Content providing program to function as

Images