JP4808763B2 - Audio information collecting apparatus, method and program thereof - Google Patents

Description

  The present invention relates to a voice information collection apparatus that provides various information services, a method thereof, and a program thereof.

  Conventionally, when various information services (downloading, account creation, etc.) are used on a network such as the Internet, techniques such as image authentication and voice authentication are used. These technologies are authentication systems for preventing unauthorized access (such as random and continuous password entry) from a malicious program. In the image authentication, for example, an image obtained by processing a background or a character angle or an image that can be recognized by anyone is displayed, and text representing the image is input. In the voice authentication, for example, a voice is played and the heard voice is input as text.

On the other hand, there is a technology called voice recognition used for personal authentication and the like. Speech recognition is a technology that analyzes words spoken by humans using a computer and extracts spoken content as character data or recognizes an individual, and has attracted attention as a new input means. Since this voice recognition uses a statistical method, data recording a large amount of utterances is indispensable.
For example, in order to perform language learning, a technique is known in which conversation example data and misunderstanding example data are solicited and learning data adapted to a conversation scene is collected (for example, Patent Document 1).

JP 2003-84654 A

  However, in Japanese Patent Application Laid-Open No. 2004-260260, language learners are invited so that language learners are targeted. Therefore, the number of users who post data is limited, and there is a limit in collecting a large amount of data. In addition, collecting data in the form of public offering is often a burden on the user, and even if an incentive is given, it is difficult to promote data collection.

  An object of the present invention is to provide an audio information collecting apparatus, a method thereof, and a program thereof that can efficiently collect audio data.

  The voice information collection device of the present invention is a voice information collection device that collects voice information together with providing information services to terminal devices connected via a network, and includes text data and pronunciation information of the text data. The voice information storage means for storing in association and the text data stored in the voice information storage means are transmitted to the terminal device for display, and the voice information voice-input by the terminal device is acquired for the text data Voice information acquisition means that performs voice recognition of the acquired voice information, and information service provision means that starts providing the information service when the voice recognition result matches the pronunciation information of the transmitted text data The voice information acquisition means stores the voice information in the same format as the text data displayed on the terminal device. Voice input requesting means for causing the text data to be displayed to be displayed on the terminal device in parallel with the displayed text data and allowing the user to input voice to the text data to be collected. Features.

The present invention provides an information service by voice recognition using text data and its voice information, and can collect voice information for the text data. Here, the voice information is data obtained by digitizing voice uttered by a person from text data, and the pronunciation information is data indicating how to read the voice data.
The provision of the information service is performed using text data in which voice information is already stored in the voice information storage means. The collection target of the voice information is all text data stored in the voice information storage unit, but it is preferable to use text data in which no voice information is stored.

According to this invention, the voice information acquisition means and the information service provision means transmit the text data in which the voice information is already stored among the text data stored in the voice information storage means to the terminal device for display. The speech information input by speech to the text data is recognized by speech, and the speech recognition result is compared with the pronunciation information stored in the speech information storage means. If they match, an information service is provided. Further, the text input request means sends the text data stored in the voice information storage means to the terminal device together with the text data used for voice recognition, displays the text data in parallel, and causes the text data to be voice-inputted.
As described above, by performing voice input for using the information service and voice input for collecting voice information at the same time, it is possible to prompt the user to input voice without hesitation. As a result, it is possible to efficiently collect voice information of text data to be collected. Further, when using the information service, the user tries to input appropriate information, so that more appropriate audio information can be input.

  In the voice information collecting apparatus of the present invention, input information storage means for storing voice information voice-input to text data displayed by the voice input request means in association with the text data and storing it in the voice information storage means Is preferably further provided.

  In the present invention, since the voice information input to the text data to be collected is stored in the voice information storage means in the voice information collection device described above, more accurate voice information is collected in the voice information storage means. be able to.

In the voice information collection device of the present invention, input candidate storage means for storing the text data stored in the voice information storage means, the voice information, and the number of times the same voice information has been acquired, in association with each other; Registration determination means for determining that registration is possible when the number of times in the predetermined text data and the voice information is equal to or greater than a predetermined number, wherein the input information storage means is the voice that has been determined to be registerable. Information is preferably stored in the voice information storage means in association with the text data.

  According to this invention, the input candidate storage means stores the text data stored in the speech information storage means, and stores the number of times the same speech information is acquired for the text data. As the number of times increases, it can be said that the information is more well-known. Therefore, it is possible to register the voice information acquired a predetermined number of times or more by the registration determination unit, and the input information storage unit associates the voice information that can be registered with the text data. Since it is stored in the voice information storage means, more appropriate voice information for the text data can be collected. The data collected in this way is highly reliable.

  The voice information collection device of the present invention further comprises search history storage means for storing a search history of keyword search, wherein the voice input request means searches for the keyword whose search rate has rapidly increased based on the search history. It is preferable to extract from the history storage means and display it on the terminal device as text data to be collected voice information.

  In this invention, the voice input request means extracts the keyword whose search rate has increased rapidly based on the search history stored in the search history storage means, and sets it as text data to be collected voice information. The search history storage means stores keywords searched at a search site or the like, so that keywords with high topicality and fashion are stored quickly. Of these keywords, the keyword whose search rate has increased rapidly has higher topicality and fashionability. Therefore, it is possible to quickly and accurately collect voice information for keywords with high topicality and popularity.

The voice information collecting apparatus according to the present invention further comprises image conversion means for generating image data obtained by converting text data associated with the voice information into text data among the text data stored in the voice information storage means . It is preferable.
According to the present invention, the text data extracted from the voice information storage means is converted into image data, and each converted image data is transmitted to the terminal device for display. Since image data is difficult to analyze, unauthorized access and the like can be prevented, and security can be improved.

In the voice information collection device of the present invention, it is preferable that the text data stored in the voice information storage means is text data relating to an advertisement.
According to the present invention, the text data related to the advertisement is used as the text data. The text data related to the advertisement is, for example, information related to a product or service such as a product name, a manufacturer name, and a tagline, and the information is displayed in text. For this reason, every time the various information services are used, the advertisement text is noticed by the user, and the advertisement text is uttered by the user himself / herself, so that the advertising effect can be improved.

  The voice information collection method of the present invention is a voice information collection method for collecting voice information as well as providing an information service to a terminal device connected via a network. The voice information collection method includes text data and pronunciation information of the text data. Associating and storing the voice data in the voice information storage means, transmitting the text data stored in the voice information storage means to the terminal device for display, and obtaining the voice information inputted by voice from the terminal device to the text data When the acquired speech information is recognized, and the result of the speech recognition matches the pronunciation information of the transmitted text data, the provision of the information service is started, and the speech information acquisition means is the terminal device. The text data to be collected from the voice information is displayed in the same format as the text data to be displayed on the display. Text to display data in parallel with the terminal device, and wherein the to voice input to text data to be the collection object to the user.

According to the present invention, text data in which voice information is already stored among the text data stored in the voice information storage means is transmitted to the terminal device for display, and the voice information input by voice to the text data is displayed. The voice recognition result is compared with the pronunciation information stored in the voice information storage means, and if they match, an information service is provided. The text data stored in the voice information storage means is transmitted to the terminal device together with the text data used for voice recognition, displayed in parallel, and voice input is performed on the text data.
Thus, by performing voice input for using the information service and voice input for collecting voice information at the same time, it is possible to prompt the user to input voice without hesitation. As a result, it is possible to efficiently collect voice information of text data to be collected. Further, when using the information service, the user tries to input appropriate information, so that more appropriate audio information can be input.

The audio information collection program of the present invention causes a computer to execute the above-described audio information collection method.
The voice information collection program of the present invention is characterized by causing a computer to function as the aforementioned voice information collection device.
According to the present invention, the voice information collecting program causes the computer to carry out the voice information collecting method described above, so that the same operational effects as described above can be obtained with a simple configuration simply by installing the voice information collecting program. Highly useful.

[First Embodiment]
A first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is a block diagram showing a schematic configuration of a service providing system according to the first embodiment of the present invention, and FIG. 2 is a block diagram showing a schematic configuration of a service providing server in the first embodiment. FIG. 3 is a schematic diagram showing a state where the start page is displayed on the screen in the first embodiment. FIG. 4 is a flowchart showing the operation of the service providing server in the first embodiment.
[Service providing system configuration]
As shown in FIG. 1, the service providing system 10 includes a service providing server 100, a web server 200 connected to the service providing server 100 via the Internet 20, and a terminal device 300.

  The Internet 20 is the Internet based on a general-purpose protocol such as TCP / IP, but is not limited to this. For example, an intranet such as a LAN (Local Area Network), a network such as a communication line network or a broadcasting network in which a plurality of base stations capable of transmitting and receiving information via a wireless medium constitute a network, and a medium for directly receiving data Any configuration that transmits and receives data, such as the wireless medium itself, can be used.

The service providing server 100 has functions of providing information services and collecting voice information for text data. That is, the service providing server 100 functions as a voice information collecting device in the present invention.
As shown in FIG. 2, the service providing server 100 includes a voice information database 101 as voice information storage means, a voice candidate database 102 as input candidate storage means, and a search history database as search history storage means as storage means. 103.

  For example, as shown in Table 1 below, the voice information database 101 stores voice data representing the text data for each text data and reading data as pronunciation information indicating the reading of the text data as one record. It has a table structure. As voice data, data obtained by digitizing voice uttering text is stored in various data formats. Reading data is stored as katakana or hiragana text data. In the voice information database 101, there are voice data stored for each text data and those not stored.

  For example, as shown in Table 2 below, the speech candidate database 102 includes, for each text data, a speech data candidate of the text, a reading data candidate based on the speech data candidate, the same text data, and the same speech data. It has a table structure in which the number of times a candidate is input is stored as one record. As the text data, the same text data as the text data stored in the speech information database 101 is stored, and a plurality of speech data candidates and a plurality of reading data candidates may exist for one text data. As voice data candidates, voice data input by the user is stored, and as reading data candidates, katakana or hiragana text is stored. As the number of times of input, the number of times that the same speech data candidate is speech input for the same text is stored.

  For example, as shown in Table 3 below, the search history database 103 has a table structure in which the search frequency and search date of this keyword are stored as one record for each keyword. The keyword is stored as a keyword searched on a search site or the like.

  In addition, the service providing server 100 serves as an arithmetic processing unit, such as a voice input request unit 110, an image conversion unit 120, a voice information acquisition unit 130, a voice recognition unit 140, a registration determination unit 150, and an input information storage unit. Voice registration means 160, service provision determination means 170, service provision means 180 as information service provision means, and text extraction means 190. Moreover, although not shown in figure, you may provide the data transmission / reception means which transmits / receives data via the internet 20, the output means which outputs a web page as a screen display, input means, such as a keyboard, etc.

  The voice input request unit 110 transmits information related to the start page for starting use of the service to the terminal device 300 via the Internet 20. The information related to the start page includes authentication text data used for permitting use of the service and target text data used for collecting voice, and information such as a form forming the start page described later. . As the authentication text data, text data in which voice data and reading data are already stored is randomly extracted from the text data stored in the voice information database 101. As the target text data, text data stored in the voice information database 101 that is not stored with voice data and reading data is preferentially extracted.

  The image conversion unit 120 converts the authentication text data and target text data extracted by the voice input request unit 110 into image data. For example, font images of various characters are stored in advance in storage means (not shown), the corresponding font images are extracted one by one, and the extracted font images are combined to generate an image corresponding to the text data. Then, this image is converted into bitmap format image data (image data expressed by a dot pattern), and deformed by distorting it using a random number. Note that the method of converting text data into image data is not limited to this, and any method can be used.

FIG. 3 specifically shows the start page in which the information transmitted by the voice input request unit 110 and the image conversion unit 120 is displayed on the screen by the output unit of the terminal device 300 as described above.
As shown in FIG. 3, the start page 30 includes an authentication area 31 used for starting a service, a target area 32 used for voice collection, an OK button 33, a cancel button 34, and voice input. And an operation explanation unit 35 for displaying a sentence for prompting the user.

  In the authentication area 31, an authentication text display unit 311, a microphone display unit 312 displaying an image of a microphone indicating that voice input is possible, and voice input of the authentication text displayed on the authentication text display unit 311 are started. There are provided an input start button 313 and an input end button 314 for ending speech input of the authentication text displayed on the authentication text display unit 311. The authentication text display unit 311 displays image data obtained by converting authentication text data. In addition, voices input until the input start button 313 is pressed and then the input end button 314 is pressed are collected as authentication voice data.

  In the target area 32, the target text display unit 321, the microphone display unit 322 displaying a microphone indicating that voice input is possible, and the voice input of the text displayed on the target text display unit 321 are started. An input start button 323 for input and an input end button 324 for ending the voice input of the text displayed on the target text display unit 321 are provided. The target text display unit 321 displays image data obtained by converting the target text data into an image. In addition, voices input until the input start button 323 is pressed and then the input end button 324 is pressed are collected as target voice data.

  Thus, by making the display format of the authentication area 31 and the target area 32 for prompting the user to input voices, the user can input voices for the purpose of using the service without being aware of voice collection. Audio information can be collected efficiently.

When the OK button 33 is pressed, the acquired authentication voice data and target voice data are transmitted to the service providing server 100.
The cancel button 34 stops processing.
In the operation explanation section 35, an explanation sentence for prompting voice input is displayed. Sentences in which both the authentication text and the target text are input by voice are preferable. For example, as shown in FIG. 3, “Please input the following two readings by voice” is displayed.

  The voice information acquisition unit 130 acquires the authentication voice data and the target voice data input by voice on the start page 30 displayed on the output unit of the terminal device 300. Specifically, the voice data input by the terminal device 300 is transmitted to the service providing server 100 and received by the data transmitting / receiving means of the service providing server 100.

  The voice recognition unit 140 recognizes the authentication voice data and the target voice data acquired by the voice information acquisition unit 130 and converts them into text data (converted text). A method for converting voice data into text data is not particularly limited, and a known method can be used.

  The registration determination unit 150 determines whether or not the acquired target voice data can be registered in the voice information database 101. Specifically, the target text data displayed as an image on the target text display unit 321 of the start page 30, the acquired target voice data, and the converted text obtained by converting the target voice data into text by the voice recognition unit 140, The text stored in the speech candidate database 102 is associated with speech data candidates and reading candidates. For example, registration is possible if the number of inputs of the corresponding text data, speech data candidate, and reading candidate is two or more, and registration is not possible if the number of inputs is less than two. That is, registration is possible when the same speech data candidate and reading candidate are input three times or more for one text. Therefore, if the number of inputs is 1, the number of inputs in the speech candidate database 102 is updated to a value of plus 1, and if the corresponding text is not registered, it is newly stored in the speech candidate database 102.

  The voice registration unit 160 stores the target voice data determined to be registerable by the registration determination unit 150 in the voice information database 101. Since the corresponding text data is already stored in the voice information database 101, the voice data and the reading data thereof are stored. One or more voice data can be stored for one text data. If voice data registered by the same user by user authentication or similar voice data has already been registered, the registration may not be performed.

  The service provision determining unit 170 determines whether or not the service can be provided using the authentication text data displayed as the authentication text image on the start page 30 and the acquired authentication voice data. Specifically, the authentication text data, the authentication voice data, and the converted text obtained by converting the authentication voice data into text data by the voice recognition unit 140 are converted into text data, voice data, and reading data stored in the voice information database 101. Associate. Service can be provided if there is applicable data. Since the voice information database 101 stores a plurality of voice data for one text data, it may be matched with any one voice data. If there is no matching text data, the service cannot be provided.

  The service providing unit 180 provides a service when the service providing determining unit 170 determines that the service can be provided. Examples of services include various services such as download and creation of a new account. Information on the top page and service start page of each service is transmitted to the terminal device 300, and the contents displayed on these web pages Various services become available.

The text extraction unit 190 extracts a keyword whose search frequency has increased rapidly from keywords stored in the search history database 103 and stores the keyword in the voice information database 101. For example, since the search history database 103 stores the number of searches for each day, a keyword whose ratio of the number of searches between the previous day and the current day exceeds a predetermined value is extracted as a rapidly increasing word, and the voice information database 101 is extracted. Remember me. The frequency with which the text extraction unit 190 checks the search history database 103 is periodically performed, and can be adjusted as appropriate, for example, every hour, every three hours, every five hours, or the like.
Further, the voice input request unit 110 can extract text data from the voice information database 101, and can also extract a keyword whose search frequency has increased rapidly via the text extraction unit 190.

The web server 200 is a device that provides various types of information on a website. For example, a web page of a search site is registered.
Although not shown, the terminal device 300 requests a service desired to be used from the service providing server 100 as an arithmetic processing unit, and receives and transmits a web page of the requested service, and outputs the web page as a screen display. Output means and input means such as a mouse and keyboard capable of inputting characters and a microphone capable of inputting voice are provided. On the other hand, the storage means includes a database for storing form data related to various forms. Although it does not specifically limit as the terminal device 300, For example, a mobile telephone, a notebook personal computer, etc. are mentioned.
1 illustrates a configuration in which one terminal device 300 and one service providing server 100 are connected, but in practice, the service providing server 100 is connected to many terminal devices 300. .

[Service delivery system operation]
Next, the operation of the service providing system 10 will be described based on FIG. 1, FIG. 2, and FIG.
First, in order to access the service providing server 100, the user performs an input operation on the input unit of the terminal device 300, for example, starts a web browser, inputs an address, and requests an information service to be used, for example, a download service. .
In step S1, when the service providing server 100 receives a request for information service by a transmission / reception unit (not shown), the service providing server 100 proceeds to S2.
In step S <b> 2, the voice input request unit 110 transmits start page information such as authentication image data and target image data to be displayed as the start page 30 shown in FIG. 3 to the terminal device 300. The authentication image data is displayed on the authentication text display unit 311, and the text data in which the voice data is stored is randomly extracted from the text data stored in the voice information database 101, and is converted by the image conversion unit 120. The image is converted. The target image data is displayed on the target text display unit 321. Text data in which no voice data is stored is preferentially extracted from the text data stored in the voice information database 101, and image conversion is performed. The image is converted by means 120.

  The terminal device 300 receives start page information by a transmission / reception unit (not shown), and displays an image of the start page 30 by an output unit. Thereby, on the start page 30, after the user presses the input start button 313, the user reads the text displayed on the authentication text display unit 311 using input means such as a microphone and starts voice input. Then, the input end button 314 is pressed to end the voice input. Similarly, after the input start button 323 is pressed, the text displayed on the target text display unit 321 is read using input means such as a microphone and voice input is started. Then, the input end button 324 is pressed to end the voice input. When the user presses the OK button 33, the voice input authentication voice data and target voice data are transmitted to the service providing server 100.

In step S3, the voice information acquisition unit 130 of the service providing server 100 acquires the authentication voice data and the target voice data input from the terminal device 300, and the process proceeds to step S4. Specifically, authentication voice data and target voice data are received by the transmission / reception means.
In step S4, the voice recognition unit 140 recognizes the acquired authentication voice data and target voice data, converts them into text data, and then proceeds to step S5. Note that the converted text data is read data of voice data.
In step S <b> 5, the registration determination unit 150 determines whether the acquired target voice data can be registered in the voice information database 101. Specifically, the target text displayed as image data on the target text display unit 321 of the start page 30, the acquired target voice data, the reading data generated by text conversion of the target voice data, and the voice candidate database 102 are stored. The text data, the speech data candidate, and the reading data candidate are collated. If there is matching data, if the number of inputs is two or more, it is determined that registration is possible, and the process proceeds to step S6. On the other hand, if the number of times of input is less than 2, it is determined that registration is not possible because there is no predetermined number or more, and the process proceeds to step S7.

In step S6, the target voice data and the reading data that can be registered by the voice registration unit 160 are stored as the voice data and reading data of the corresponding text data stored in the voice information database 101, and the process proceeds to step S8. If the same voice data and reading data are already stored in the voice information database 101, no processing is performed. Then, the number of inputs of the corresponding text data, speech data candidate, and reading data candidate in the speech candidate database 102 is updated with a value obtained by adding one.
In step S7, the number of inputs of the corresponding text data, speech data candidate, and reading data candidate in the speech candidate database 102 is updated with a value obtained by adding one. On the other hand, if there are no voice data candidates and reading data candidates that match the acquired voice data and reading data in the voice candidate database 102, they are newly stored and the process proceeds to step S8.

In step S8, the authentication voice data acquired in step S3 by the service provision determination unit 170 and the text data converted in text in step S4 match the voice data and reading data of the corresponding text data stored in the voice information database 101. If it matches, the service can be provided and the process proceeds to step S9. On the other hand, if they do not match, the service cannot be provided, the process returns to step S2, and the start page 30 is transmitted again to request voice input. At this time, the authentication text image displayed on the authentication text display unit 311 may be changed to another authentication text image.
In step S <b> 9, the service providing unit 180 provides various services to the terminal device 300. In the present embodiment, as a download service, for example, a file to be downloaded is transmitted and displayed on the screen by the output means of the terminal device 300 or the like.

  Further, the text extraction means 190 periodically accesses the search history database 103, extracts a rapidly rising word in which the ratio of the previous day and today's search history is a predetermined value or more, and stores it in the voice information database 101 as text data. Yes.

[Effects of First Embodiment]
In the first embodiment described above, the following operational effects can be achieved.
The service providing server 100 determines whether or not the service can be provided based on the authentication text data image-displayed on the start page 30 by the service providing determination unit 170 and the authentication voice data input to the authentication text data. The target voice data corresponding to the target text data input by voice on the start page 30 by the means 150 and the voice registration means 160 is stored in the voice information database 101.
That is, the target voice data of the target text data can be collected every time the service is provided, and statistical data can be collected without making the user aware of the data collection. In addition, by displaying the authentication text data for voice input for service provision and the target text data for voice input for voice collection in parallel in the same format, voice input without making the user aware of voice collection Can be encouraged.

  Further, when the same voice data and the same reading data are input three times or more for one text data, the target voice data and the reading data are stored in the voice information database 101. High data can be collected. The voice information database 101 collected in this way can be widely used as a voice database and is highly useful.

  Further, the text extraction unit 190 periodically accesses the search history database 103, extracts a keyword whose search frequency has increased rapidly, and stores it in the voice information database 101. Since the keyword whose search frequency has increased rapidly represents the topic or trend at that time, it is possible to efficiently and quickly collect voice data for these keywords. Therefore, voice data for the latest keyword can be collected quickly and with high accuracy.

  Furthermore, in this embodiment, the authentication text display unit 311 and the target text display unit 321 convert the text data into an image by the image conversion unit 120 and display it as image data. Since image data is difficult to analyze, unauthorized access and the like can be prevented, and security can be improved.

[Second Embodiment]
A second embodiment of the present invention will be described with reference to FIG.
FIG. 5 is a block diagram showing a schematic configuration of a service providing server according to the second embodiment of the present invention.
[Service providing system configuration]
In the second embodiment, the configuration is the same as that of the first embodiment except that the authentication is performed using the advertisement text data stored in the advertisement information database 104. Hereinafter, a configuration different from the first embodiment will be described.

  For example, as shown in Table 4 below, the advertisement information database 104 has a table structure in which audio data that is advertisement attribute information regarding the advertisement image and its reading data are stored as one record for each advertisement text data for advertisement. It has become. Examples of the advertisement text data include the name of the product displayed in the advertisement image data, the manufacturer, the catchphrase, and the like, and these are stored as the text data. As the reading data, text data converted into text by analyzing voice data is stored in katakana. The reading data may be hiragana.

Further, since the operations of the voice input request unit 112 and the service provision determination unit 172 are different from those of the first embodiment by determining whether or not the service can be provided based on the advertisement text data, these will be described.
The voice input request unit 112 transmits information regarding the start page for starting use of the service to the terminal device 300 via the Internet 20. The information about the start page includes information such as a form forming the start page, in addition to an authentication image for determining whether to provide a service and a target image for collecting sound. As the authentication image, image data obtained by image conversion of the advertisement text data randomly extracted from the advertisement information database 104 by the image conversion means 120 is displayed. As the target image, among the texts stored in the voice information database 101, The audio data and reading data not stored are preferentially extracted, and the image data converted by the image conversion means 120 is displayed.

  The service provision determination unit 172 determines whether or not the service can be provided by using the advertisement text data displayed as an image on the start page 30 and the acquired advertisement voice data. Specifically, the advertisement text data displayed as an authentication image, the acquired advertisement voice data, and the reading data obtained by converting the advertisement voice data into text by the voice recognition means 140 are stored in the voice information database 101. Collate with data, voice data and reading data. If they match, the service can be provided. Since the advertisement information database 104 stores a plurality of voice data and reading data for one advertisement text data, it suffices to match any one of the voice data and reading data. If there is no matching data, the service cannot be provided.

  The service providing server 400 having the above configuration can collect voice data of text data stored in the voice information database 101 by the same operation as in the first embodiment. The advertisement information database 104 stores data based on advertisement information transmitted from an advertiser terminal (not shown) connected via the Internet 20.

[Effects of Second Embodiment]
In the second embodiment described above, the following operational effects can be obtained in addition to the operational effects of the first embodiment.
The service providing server 400 displays the advertisement text data on the start page 30 as an image. Accordingly, since the advertisement is displayed every time the use of the service is started, the advertising effect can be improved. In particular, displaying an advertisement on the start page 30 that is always displayed when using the service is easily noticeable by the user and has a high advertising effect. Further, since the user speaks the advertising text as it is, the effect of causing the user to recognize the product name, manufacturer name, catchphrase, etc. is enhanced, and the advertising effect can be improved.

[Modification]
In addition, this invention is not limited to embodiment mentioned above, In the range which can achieve the objective of this invention, the deformation | transformation shown below is also included, for example.
When there are a plurality of voice data for one text data in the voice information database 101, a voice data collection grouped for each text data can be created. As described above, since a plurality of voice data can be collected, it is possible to improve accuracy when used as a voice information database.

In the above embodiment, the target voice input information is stored in the voice information storage unit regardless of whether or not the service can be provided. However, the target voice input information may be stored only when the service can be provided. Since it is necessary to input an accurate voice in order to be able to provide a service, more reliable voice information can be collected by re-input. In addition, there is a high possibility that it is possible to eliminate clearly different voice data such as noise caused by erroneous operation and voice data of a person who does not know Japanese.
Further, the voice information storage means used in the above embodiment is configured to collect the reading data of the voice data. However, the present invention is not limited to this, and the items may be appropriately changed or increased / decreased depending on the information to be collected. You may let them.

  Furthermore, in the above embodiment, the text data acquired from the audio information database 101 by the image conversion means 120 is converted into image data. However, the text data is not converted but transmitted to the terminal device as it is for display. Good. According to this, the configuration of the service providing server 100 can be simplified and the processing can be speeded up.

In the above embodiment, the registration determination unit 150 determines that registration is possible in the voice information database 101 when the same target voice data is input three times with respect to the target text data displayed on the terminal device 300. However, the number of times to determine whether registration is possible is not limited to this. In order to increase reliability, it is preferable to increase the number of times.
In addition, since the user who has been permitted to use the service by the authentication text data and the authentication voice data has high reliability, the voice data with high reliability can be collected.

  The present invention can be used as a voice information collection device that provides various information services.

1 is a block diagram showing a schematic configuration of a service providing system according to a first embodiment of the present invention. The block diagram which shows schematic structure of the service provision server in the said 1st Embodiment. Schematic which shows the state by which the start page in the said 1st Embodiment was displayed on the screen. The flowchart which shows operation | movement of the service provision system in the said 1st Embodiment. The block diagram which shows schematic structure of the service provision server in 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Service provision system 20 Internet 30 Start page 31 Authentication area 32 Target area 100 Service provision server 101 Voice information database 102 Voice candidate database 103 Search history database 104 Advertisement information databases 110 and 112 Voice input request means 120 Image conversion means 130 Voice information acquisition Means 140 Voice recognition means 150 Registration judgment means 160 Voice registration means 170, 172 Service provision judgment means 180 Service provision means 190 Text extraction means 200 Web server 300 Terminal device

Claims (9)

A voice information collection device that collects voice information together with provision of information services to terminal devices connected via a network,
Voice information storage means for storing text data and pronunciation information of the text data in association with each other;
Voice information acquisition means for transmitting text data stored in the voice information storage means to the terminal device for display, and acquiring voice information voice-input by the terminal device for the text data;
An information service providing unit that performs voice recognition of the acquired voice information and starts providing the information service when the voice recognition result matches the pronunciation information of the transmitted text data;
The voice information acquisition means displays the text data to be collected of the voice information in the same format as the text data to be displayed on the terminal device by causing the terminal device to display the text data in parallel with the displayed text data. A voice information collection device comprising voice input requesting means for inputting voice to the text data to be collected. The voice information collection device according to claim 1,
Voice information input by voice to the text data displayed by the voice input request means is further provided with input information storage means for storing the voice information in the voice information storage means in association with the text data. Audio information collection device. The voice information collection device according to claim 2,
Input candidate storage means for storing the text data stored in the voice information storage means, the voice information, and the number of times the same voice information has been acquired, in association with each other;
Registration determination means for determining that registration is possible when the number of times in the predetermined text data and the audio information is a predetermined number of times or more;
The voice information collection device, wherein the input information storage means stores the voice information determined to be registerable in the voice information storage means in association with the text data. In the voice information collection device according to any one of claims 1 to 3,
It further comprises search history storage means for storing a search history of keyword search,
The voice input request means is
The voice information collection device, wherein the keyword whose search rate has rapidly increased based on the search history is extracted from the search history storage means and displayed on the terminal device as text data to be collected by the voice information. The voice information collection device according to claim 2 ,
An audio information collecting apparatus, further comprising: image conversion means for generating image data obtained by converting text data associated with the audio information among text data stored in the audio information storage means into an image. . In the voice information collection device according to any one of claims 1 to 3 and claim 5 ,
The voice information collecting apparatus, wherein the text data stored in the voice information storage means is text data relating to an advertisement. A voice information collecting method for collecting voice information together with providing information services to terminal devices connected via a network,
The text data and the pronunciation information of the text data are associated with each other and stored in the voice information storage means,
The text data stored in the voice information storage means is transmitted to the terminal device for display, and the voice information voice-input by the terminal device is acquired for the text data,
Performing speech recognition of the acquired speech information, if the speech recognition result matches the pronunciation information of the transmitted text data, start providing the information service,
The voice information acquisition means displays the text data to be collected of the voice information in the same format as the text data to be displayed on the terminal device by causing the terminal device to display the text data in parallel with the displayed text data. A speech information collecting method, comprising: inputting voice to the text data to be collected. A voice information collection program that causes a computer to execute the voice information collection method according to claim 7. A voice information collecting program for causing a computer to function as the voice information collecting device according to any one of claims 1 to 6.

Images