JP5502787B2 - Voice conversion support device, program, and voice conversion support method - Google Patents

Description

  Embodiments described herein relate generally to a speech conversion support apparatus, a program, and a speech conversion support method used for a service that converts a character such as a character into speech.

  In recent years, for example, on the Internet, a service for converting characters into speech has been started, and a speech synthesizer is used for this service.

  In general, a speech synthesizer converts a text (character string) of text input by a user into speech synthesis waveform data and outputs a speech signal or a speech file.

  By the way, in order to convert characters into speech, a speech synthesizer with high processing performance is required, and it is expensive to start this type of service. On the other hand, when a speech synthesizer with low performance is used, there is a problem that speech synthesis processing takes time and responsiveness is impaired.

JP 2004-117778 A

  In general, this type of service only returns certain voice data having no special characteristics to the text input by the user. Therefore, it is necessary to add some added value in order to give the service a characteristic.

  The problem to be solved by the present invention is to reduce costs while maintaining the responsiveness of character / speech conversion, and further, a speech conversion support apparatus, program and speech conversion support method capable of realizing a characteristic service Is to provide.

The speech conversion support device according to the embodiment is connected to a speech conversion device that converts text data into speech data via a network. The voice conversion support device includes a cache data storage unit, an index information storage unit, a search unit, and a data processing unit. In the cache data storage unit, voice data previously converted by the voice converter and text data from which the voice data is converted are stored as past data correspondingly. The index information storage unit ranks and stores a plurality of pieces of index information with different narrowing ranges for searching past data stored in the cache data storage unit in descending order of affinity with the user. The search unit searches a plurality of pieces of index information in the index information storage unit in the order of ranking based on text data input from a request source. The data processing unit as a result of search by the search unit, whereas the text data returns the audio data used in the past as indicated by the presence index information to the previous reading from the cache data storage unit Kiyo Motomemoto, either The text data that does not exist in the index information is sent to the voice converter, and the voice data converted by the voice converter is returned to the request source. The index information storage unit is a user registration dictionary uniquely registered by a user who has logged in to the device, a first usage history which is a usage history of the user, and a usage history of all users who have logged into the device. The search step first searches the user registration dictionary, and when the input text data does not exist in the user registration dictionary, the search step second searches the first usage history. If the text data does not exist in the first usage history, third, the second usage history is searched.

1 is a diagram illustrating an overall configuration of a speech synthesis system according to an embodiment. It is a block diagram of an application server. It is a figure which shows an example of index information. It is a flowchart which shows operation | movement of this speech synthesis system. It is a flowchart which shows the dictionary registration operation | movement for every user. It is a flowchart which shows the search operation | movement of the cache data in an application server.

Hereinafter, embodiments will be described in detail with reference to the drawings.
(First Embodiment) FIG. 1 is a diagram showing the configuration of a speech synthesis system according to a first embodiment.

  As shown in FIG. 1, the speech synthesis system of this embodiment includes a computer 1 (hereinafter referred to as “user PC 1”), which is a terminal device operated by a service user (hereinafter referred to as “user”), and a speech synthesis engine. It is composed of a voice synthesis server 3 which is a mounted computer, a computer 2 (hereinafter referred to as “application server 2”) as a voice conversion support device, and a network 4 which connects these devices.

  The speech synthesis server 3 converts the intermediate file (a set of text data and accent marks) transferred (input) from the application server 2 into speech data (hereinafter referred to as “speech file”) and returns it to the application server 2.

  The application server 2 is interposed between the voice synthesis server 3 and the user PC 1 and exchanges text data, intermediate files, and voice files.

  As shown in FIG. 2, the application server 2 includes a graphic user interface unit 21 (hereinafter referred to as “GUI unit 21”), a memory 22, an intermediate file generation unit 23, and an index information storage unit for managing cached data. 24, a cache data storage unit 25, a search unit 26, a communication processing unit 27, a data processing unit 28, a registration unit 29, and the like.

  The application server 2 searches the audio data cached (stored) in the cache data storage unit 25 using the index information in the index information storage unit 24 using the text data input from the user PC 1 as a keyword (search key). If there is a hit, the cached voice file is read out and returned to the requesting user PC 1 without requesting the voice synthesis server 3 for voice synthesis.

  In other words, the application server 2 checks whether or not it is cached in its own hard disk device, and if it is not cached, it outputs the text data input from the user PC 1 to the speech synthesis server 3, and As a response, the voice file converted (voice synthesized) by the voice synthesis server 3 is acquired and sent to the user PC 1.

  The GUI unit 21 displays a screen for logging in to the application server 2 from the user PC 1, a search screen, a registration screen, etc., accepts a voice synthesis request from the user PC 1, input of text data, and the like, and a voice file as a response to the request To the PC.

  That is, the GUI unit 21 realizes an input / output interface between the user PC 1 and the application server 2.

  The memory 22 stores a voice conversion dictionary. The voice conversion dictionary corresponds to the text data that is converted from the voice data converted (voice synthesized) by the voice synthesis server 3 and the accent symbols (accent information) that specify the strength of the sound when the text data is converted into voice. It is a dictionary for reference that is saved. This voice conversion dictionary is used by the intermediate file generation unit 23.

  The memory 22 is used as a work area when the data processing unit 28, the search unit 26, the registration unit 29, and the like execute their respective processes.

  The intermediate file generation unit 23 refers to the speech conversion dictionary by using the text data to be converted input from the user PC 1 as a key, reads the corresponding accent symbol from the speech conversion dictionary, and sets a set of the text data and the accent symbol. An intermediate file is generated and stored in the memory 22. The intermediate file is transmitted to the speech synthesis server 3 as original data for speech synthesis.

  The cache data storage unit 25 stores voice files, text data, intermediate files, and the like that have been converted to the previous (past) time.

  In the index information storage unit 24, a plurality of pieces of index information with different narrowing ranges for searching the past data cached (saved) in the cache data storage unit 25 in order of high affinity with the user are ranked. Is remembered.

  That is, the index information storage unit 24 stores index information for managing the data cached in the cache data storage unit 25. The index information will be specifically described with reference to FIG.

  The search unit 26 searches the registration dictionary 42 set for each user ID, the use history 43 for each user ID, and the server use history 44 in order of rank. That is, the search unit 26 searches the plurality of index information in the index information storage unit 24 in the order of ranking based on the text data input from the request source.

  The communication processing unit 27 exchanges data with the speech synthesis server 3 by TCP (HTTP) communication.

  As a result of the search by the search unit 26, the data processing unit 28 is cached from the past data storage location indicated by the index information where the text data (also referred to as a search key or keyword) exists, that is, from the cache data storage unit 25. Is read out and returned to the conversion requesting user PC1.

  On the other hand, if no keyword exists in any of the index information, the intermediate file generated by the intermediate file generation unit 23 based on the input text data and stored in the memory 22 is sent to the speech synthesis server 3, and the speech synthesis server The voice file converted (voice synthesized) by 3 is returned to the requesting user PC 1. Text data may be sent.

  The registration unit 29 registers the dictionary information (text data and accent symbols) that the user independently inputs or edits in the registration dictionary 42 for each user ID on the dictionary registration screen displayed by the GUI unit 21.

  As shown in FIG. 3, the index information storage unit 24 includes a plurality of pieces of index information (a registration dictionary 42 for each user ID, a use history 43 for each user ID, and the like) associated with each user ID in the user ID table 41. Stored is a usage history 44 (hereinafter referred to as “server usage history 44”) of all users of the application server 2).

  In the user ID table 41, a user ID that is identification information of a user who can log in to the application server 2 is set. The user identification information includes not only the user ID but also a password.

  The registration dictionary 42 for each user ID includes text data uniquely registered by the user (referred to as “text”), an intermediate file that is a set of text and its access symbol, and an audio file corresponding to these data. And a storage location index indicating the storage location of.

  The registration dictionary 42 for each user ID is set to the first and second ranks (first and second ranks) as the ranks for the search, and is referred to first when the search unit 26 searches the cache data. Is done. In the dictionary, the first rank is an intermediate file, and the second rank is text. The search order as index information is first.

  The usage history 43 for each user ID includes a text when this voice conversion function is used for each logged-in user ID, an intermediate file that is a set of the text and its access symbol, and a voice file corresponding to these data. And a storage location index indicating the storage location of. Text, intermediate files, save destination indexes, etc. are called usage history.

  The usage history 43 for each user ID is set to the third and fourth ranks (third and fourth ranks) as the ranks for retrieval, and the third and fourth ranks when the retrieval unit 26 retrieves cache data. Referenced to th. In the history, the third rank is an intermediate file, and the fourth rank is text. The search order as index information is second.

  The server usage history 44 stores usage histories of all users when the user logs in to the application server 2 and uses the voice conversion function. The server usage history 44 is set to the fifth and sixth ranks (fifth and sixth ranks) as the rank in the search, and is referred to the fifth and sixth when the search unit 26 searches the cache data. Is done. In the history, the fifth rank is an intermediate file, and the sixth rank is text. The search order as index information is the third.

  Next, the operation of the speech synthesis system of this embodiment will be described with reference to the flowcharts of FIGS. First, the operation of the entire system will be described with reference to the flowchart of FIG.

  In the case of the speech synthesis system of this embodiment, when the user inputs a predetermined URL from the user PC 1 and accesses the application server 2, the GUI unit 21 displays a login screen on the user PC 1, so that the user displays the displayed login screen. In the input field, login information such as a login ID is input (step S101 in FIG. 4). In addition, a password may be input as login information.

  Then, the GUI unit 21 determines whether the login information has been registered by checking the input login information with the user ID in the user ID table 41 of the memory 22 (step S102). In the case (Yes in step S102), login to the application server 2 is permitted and a voice conversion screen is displayed (step S103).

  When the user inputs a character to be converted (text data) by key input in the character input field of the voice conversion screen (step S104), the GUI unit 21 receives the input.

  When the button for voice file creation instruction displayed on the voice conversion screen is pressed (step S105), the intermediate file generation unit 23 refers to the voice conversion dictionary using the received text data to be converted as a key. Then, the corresponding accent symbol is read from the speech conversion dictionary, an intermediate file of a set of text data and an accent symbol is generated (step S106), and the generated intermediate file is displayed on the speech conversion screen (step S107).

  Subsequently, when the user presses a button for a search instruction displayed on the voice conversion screen (step S108), the search unit 26 uses the received text data as a key to display the index information in the set order. Search in order (step S109). Details of the search operation will be described with reference to FIG.

  As a result of the search, when the target text data does not exist in any of the plurality of index information in the index information storage unit 24 (No in step S110), the data processing unit 28 is generated by the intermediate file generation unit 23. The voice conversion request is transferred (transmitted) to the voice synthesis server 3 together with the intermediate file (step S111).

  The voice synthesis server 3 receives the intermediate file and voice conversion request, synthesizes voice based on the intermediate file, and returns the generated voice file to the application server 2 through the network (step S112).

  In the application server 2, when the voice file transmitted from the voice synthesis server 3 is received by the communication processing unit 27, the voice file is transferred to the data processing unit 28, and the data processing unit 28 acquires the voice file (step S113).

  The data processing unit 28 stores the acquired audio file in the cache data storage unit 25 and sends it to the user PC 1, and the audio is output from the speaker of the user PC 1 (step S 114).

  As a result of the search in the search step in S109, when the target text data is present in any one of the plurality of index information in the index information storage unit 24 (Yes in step S110), the data processing unit 28 is searched. The corresponding voice file is read out and acquired from the storage location (cache data storage unit 25) indicated by the index information (step S113), and is transferred (transmitted) to the user PC 1 without requesting the voice synthesis server 3 ( Step S114).

  Thereafter, if the processing end operation is not performed (No in step S115), the next text input is waited. When the processing end operation is performed (Yes in step S115), the data processing unit 28 ends the text / speech conversion processing.

Next, a description will be given of a cache process for an audio file once created with reference to FIG.
In the application server 2, the created intermediate file and audio file are temporarily cached in the memory 22.

  When the intermediate file from which the audio file is converted is designated by the user on the audio conversion screen (step S121), the registration unit 29 reads the intermediate file from the memory 22 and displays the contents (text data and text data) on the audio conversion screen. (Accent part number) is displayed (step S122).

  When the user edits the text data and the accent code that are the contents of the intermediate file displayed on the voice conversion screen (step S123), the user presses the voice file save instruction button provided on the voice conversion screen. (Step S124), a save destination designation dialog box is displayed.

  When the user designates a save destination from this screen (step S125), the registration unit 29 saves the audio file in the designated save destination folder (step S126). When the ending operation is performed (Yes in step S127), the registration unit 29 registers the index information of the stored audio file in the registration dictionary 42 for each user ID in the index information storage unit 24 (step S128).

Next, the details of the search process in step S109 will be described with reference to FIG.
In this case, when a search button is operated from the voice conversion screen displayed by the GUI unit 21, the search unit 26 stores the created intermediate file text data and accent symbol in the index information storage unit 24 using the search key. The intermediate file in the registered dictionary 42 for each user ID, which is the first rank, of the index information being searched is searched (step S201: first search).

  As a result of the first search, if the same data as the search key exists in the intermediate file in the registration dictionary 42 for each user ID (Yes in step S202), the search unit 26 adds the data to the data in the next step 110. The corresponding audio file is read from the cache data storage unit 25 and passed to the data processing unit 28. As a result, the audio file is acquired by the data processing unit 28.

  On the other hand, as a result of the first search, if the same data as the search key does not exist in the intermediate file in the registration dictionary 42 for each user ID (No in step S202), the search unit 26 next selects the text of the intermediate file. Using the data as a search key, the index information stored in the index information storage unit 24 is searched for text in the registration dictionary 42 for each user ID in the second order (step S203: second search). .

  As a result of the second search, if the same data as the search key exists in the text in the registration dictionary 42 for each user ID (Yes in step S204), the process proceeds to the next step 110.

  If the same data as the search key does not exist in the text in the registration dictionary 42 for each user ID as a result of the second search (No in step S204), the search unit 26 then selects the text data of the intermediate file. And the accent mark as a search key, the intermediate file in the usage history 43 for each user in the third rank is searched from the index information stored in the index information storage unit 24 (step S205: third). Search).

  If the same data as the search key exists in the intermediate file in the registration dictionary 42 for each user ID as a result of the third search (Yes in step S206), the process proceeds to the next step 110.

  On the other hand, as a result of the third search, when the same data as the search key does not exist in the intermediate file in the registration dictionary 42 for each user ID (No in step S206), the search unit 26 next selects the text of the intermediate file. Using the data as a search key, the index information stored in the index information storage unit 24 is searched for text in the registration dictionary 42 for each user ID in the fourth rank (step S207: fourth search). .

  As a result of the fourth search, if the same data as the search key exists in the text in the registration dictionary 42 for each user ID (Yes in step S208), the process proceeds to the next step 110.

  If the same data as the search key does not exist in the text in the registration dictionary 42 for each user ID as a result of the fourth search (No in step S208), the search unit 26 next selects the text data of the intermediate file. And the accent mark as a search key, the intermediate file in the server usage history 44 in the fifth rank of the index information stored in the index information storage unit 24 is searched (step S209: fifth search). .

  As a result of the fifth search, if the same data as the search key exists in the intermediate file in the server usage history 44 (Yes in step S210), the process proceeds to the next step 110.

  On the other hand, as a result of the fifth search, when the same data as the search key does not exist in the intermediate file in the server usage history 44 (No in step S210), the search unit 26 next searches the text data of the intermediate file. Using the key, the text in the server usage history 44 in the sixth rank in the index information stored in the index information storage unit 24 is searched (step S211: sixth search).

  As a result of the sixth search, if the same data as the search key exists in the text in the server usage history 44 (Yes in step S212), the process proceeds to the next step 110.

  If the same data as the search key does not exist in the text in the server usage history 44 as a result of the sixth search (No in step S212), the search unit 26 indicates that no data exists in the cache data storage unit 25. Is notified to the data processing unit 28. Upon receiving this notification, the data processing unit 28 transmits the voice conversion request and the intermediate file to the voice synthesis server 3.

  As described above, according to this embodiment, the function of converting text into speech in response to a request from the user is divided into the application server 2 which is a user interface part and the speech synthesis server 3 which is a speech synthesis engine part. Thus, load distribution can be achieved.

  If the character or sentence (text data, character string) input by the user to the application server 2 is the same as the previously used character or sentence, the user does not request the speech synthesis server 3 to perform the process. By reading and returning the voice file cached in the hard disk device, the processing load on the voice synthesis server 3 can be reduced.

  Further, by first searching the registration dictionary 42 for each user ID registered by the user, for example, the user's unique access voice can be provided to the user.

  As a result, it is possible to reduce costs while maintaining the responsiveness of character / speech conversion, and to realize a characteristic service.

The described embodiments are presented by way of example and are not intended to limit the scope of the invention. The novel embodiment can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. The above-described embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and equivalents thereof.
In the above-described embodiment, the intermediate file and the text data are alternately searched. However, ranks may be set for a plurality of index information using only the intermediate file or only the text data, and the index information may be searched in order of the ranks.

  Further, each component shown in the above embodiment may be realized by a program installed in a storage such as a hard disk device of a computer, and the program is stored in a computer-readable electronic medium: electronic media, The computer may realize the functions of the present invention by causing a computer to read a program from an electronic medium. Examples of the electronic medium include a recording medium such as a CD-ROM, flash memory, and removable media. Further, the configuration may be realized by distributing and storing components in different computers connected via a network, and communicating between computers in which the components are functioning.

  DESCRIPTION OF SYMBOLS 1 ... User PC, 2 ... Application server, 3 ... Speech synthesis server 31 ... Graphic user interface part (GUI part), 22 ... Memory, 23 ... Intermediate file generation part, 24 ... Index information storage part, 25 ... Cache data storage part , 25 ... cache data storage unit, 26 ... search unit, 27 ... communication processing unit, 28 ... data processing unit, 29 ... registration unit, 41 ... user ID table, 42 ... registration dictionary for each user ID, 43 ... for each user ID Usage history, 44... Server usage history.

Claims (7)

In a speech conversion support device connected via a network to a speech conversion device that converts text data into speech data,
A cache data storage unit in which the speech data previously converted by the speech conversion device and the text data from which the speech data is converted are stored as past data;
An index information storage unit in which a plurality of pieces of index information with different narrowing down ranges are stored in order for searching past data stored in the cache data storage unit in descending order of affinity with the user;
A search unit that searches a plurality of pieces of index information in the index information storage unit based on text data input from a request source in the order of ranking;
The search portion searches results by, while said text data returns the audio data used in the past as indicated by the presence index information to the previous reading from the cache data storage unit Kiyo Motomemoto, in any of the index information A data processing unit that sends the text data that does not exist to the voice conversion device and returns the voice data converted by the voice conversion device to the request source ;
The index information storage unit is a user registration dictionary uniquely registered by a user who has logged in to the device, a first usage history which is a usage history of the user, and a usage history of all users who have logged into the device. With 2 usage histories,
The search unit
First, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the second usage history is searched second, and the text data is stored in the first usage history. If not, thirdly, the speech conversion support device searches the second usage history . In a speech conversion support device connected via a network to a speech conversion device that converts text data into speech data,
The voice data previously converted by the voice converter, the text data from which the voice data is converted, and the accent information that specifies the strength of the sound when the text data is converted to voice are stored as past data. Cache data storage unit
An index information storage unit in which a plurality of pieces of index information with different narrowing down ranges are stored in order for searching past data stored in the cache data storage unit in descending order of affinity with the user;
A search unit that searches a plurality of pieces of index information in the index information storage unit based on text data input from a request source in the order of ranking;
As a result of the search by the search unit, the past voice data indicated by the index information in which the text data existed is read from the cache data storage unit and returned to the request source, while no text data exists in any index information. A data processing unit that sends the voice data to the voice conversion device as voice synthesis data together with the accent information corresponding to the text data, and returns the voice data converted by the voice conversion device to the request source ,
The index information storage unit is a user registration dictionary uniquely registered by a user who has logged in to the device, a first usage history which is a usage history of the user, and a usage history of all users who have logged into the device. With 2 usage histories,
The search unit
First, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the second usage history is searched second, and the text data is stored in the first usage history. If not, thirdly, the speech conversion support device searches the second usage history . In the user registration dictionary, the first usage history, and the second usage history, text data that is a conversion source of previously converted voice data, and intermediate data that includes text data and accent marks are stored. And
The search unit
The voice according to claim 1 or 2 , wherein first the intermediate data is searched, and if the input text data does not exist in the intermediate data, the second conversion source text data is searched second. Conversion support device. A speech conversion device that converts input text data into speech data, and a speech conversion device that is connected to the speech conversion device via a network, and sends a conversion request for converting the text data to speech data to the speech conversion device. In a speech synthesis system having a speech conversion support device that acquires speech data output from the speech conversion device in response to a conversion request and sends back to the input source of text data,
The voice conversion support device includes:
A cache data storage unit in which the speech data previously converted by the speech conversion device and the text data from which the speech data is converted are stored as past data;
An index information storage unit in which a plurality of pieces of index information with different narrowing down ranges are stored in order for searching past data stored in the cache data storage unit in descending order of affinity with the user;
A search unit that searches a plurality of pieces of index information in the index information storage unit based on text data input from a request source in the order of ranking;
As a result of the search by the search unit, the voice data used in the past indicated by the index information in which the text data existed is read from the cache data storage unit and returned to the request source, but is not present in any index information A data processing unit that sends the text data to the voice converter and returns the voice data converted by the voice converter to the request source;
The index information storage unit is a user registration dictionary uniquely registered by a user who has logged in to the device, a first usage history which is a usage history of the user, and a usage history of all users who have logged into the device. With 2 usage histories,
The search unit
First, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the second usage history is searched second, and the text data is stored in the first usage history. If not, thirdly search the second usage history
A voice conversion support system characterized by this. A speech conversion device that converts input text data into speech data, and a speech conversion device that is connected to the speech conversion device via a network and sends a conversion request for converting the text data to speech data to the speech conversion device, In a speech synthesis system having a speech conversion support device that acquires speech data output from the speech conversion device in response to a conversion request and sends back to the input source of text data,
The voice conversion support device includes:
The voice data previously converted by the voice converter, the text data from which the voice data is converted, and the accent information that specifies the strength of the sound when the text data is converted to voice are stored as past data. Cache data storage unit
An index information storage unit in which a plurality of pieces of index information with different narrowing down ranges are stored in order for searching past data stored in the cache data storage unit in descending order of affinity with the user;
A search unit that searches a plurality of pieces of index information in the index information storage unit based on text data input from a request source in the order of ranking;
As a result of the search by the search unit, the past voice data indicated by the index information in which the text data existed is read from the cache data storage unit and returned to the request source, while no text data exists in any index information. A data processing unit that sends the voice data to the voice conversion device as voice synthesis data together with the accent information corresponding to the text data, and returns the voice data converted by the voice conversion device to the request source ,
The index information storage unit
A user registration dictionary uniquely registered by a user who has logged into this device, a first usage history which is the user's usage history, and a second usage history which is the usage history of all users who have logged into the device,
The search unit
First, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the second usage history is searched second, and the text data is stored in the first usage history. If not, thirdly, the second usage history is searched . A speech synthesis system , wherein the second usage history is searched . In a program that causes a speech conversion support device connected via a network to a speech conversion device that converts text data to speech data,
The voice conversion support device,
The voice data previously converted by the voice converter, the text data from which the voice data is converted, and the accent information that specifies the strength of the sound when the text data is converted to voice are stored as past data. Cache data storage unit
An index information storage unit in which a plurality of pieces of index information with different narrowing down ranges are stored in order for searching past data stored in the cache data storage unit in descending order of affinity with the user;
A search unit that searches a plurality of pieces of index information in the index information storage unit based on text data input from a request source in the order of ranking;
As a result of the search by the search unit, the past voice data indicated by the index information in which the text data existed is read from the cache data storage unit and returned to the request source, while no text data exists in any index information. The text data and the accent information corresponding to the text data is sent to the speech converter as speech synthesis data, and the speech data converted by the speech converter is functioned as a data processor that returns to the request source .
The index information storage unit
A user registration dictionary uniquely registered by a user who has logged into this device, a first usage history which is the user's usage history, and a second usage history which is the usage history of all users who have logged into the device,
The search unit
First, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the second usage history is searched second, and the text data is stored in the first usage history. If not, thirdly, the second usage history is searched . Connected via a network to a voice conversion device that converts text data into voice data, converts voice data previously converted by the voice conversion device, text data from which the voice data is converted, and the text data into voice. The cache data storage unit stores the past data stored in the cache data storage unit in correspondence with the accent information that specifies the strength of the sound, and searches the past data stored in the cache data storage unit in descending order of affinity with the user. And a speech conversion support method in a speech conversion support device comprising an index information storage unit in which a plurality of index information having different narrowing ranges are ranked and stored,
A search step of searching a plurality of index information in the index information storage unit in the order of ranking based on text data input from a request source;
As a result of the search in the search step, the past audio data indicated by the index information in which the text data existed is read from the cache data storage unit and returned to the request source, while no text data exists in any index information. A data processing step of sending the voice data together with the accent information corresponding to the text data to the voice conversion device as data for voice synthesis and returning the voice data converted by the voice conversion device to the request source;
The index information storage unit
A user registration dictionary uniquely registered by a user who has logged into this device, a first usage history which is the user's usage history, and a second usage history which is the usage history of all users who have logged into the device,
In the search step, first, the user registration dictionary is searched, and when the input text data does not exist in the user registration dictionary, the first usage history is searched second, and the first usage history is stored in the first usage history. The voice conversion support method , wherein, when the text data does not exist, thirdly, the second usage history is searched .

Images