JP2001134418A - Method and device for relaying hyper-text - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a hyper-text relaying method and device for providing a voice controlling function based on the pronunciation of specific language even to a hyper-text described in language other than the specific language. SOLUTION: When a hyper-text is displayed at a hyper-text display device having a controlling function based on a voice, the target words of voice recognition are specified from the hyper-text, and pronunciation notation in specific language corresponding to the target words of the voice recognition is derived, and the pronunciation notation and a command for voice control are added to the hyper-text.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and a device for relaying hypertext from a network to a hypertext display device having a voice control function.

[0002]

2. Description of the Related Art In recent years, a computer network such as the Internet or an intranet has been used to create HTML or S
Receiving and displaying a hypertext described in GML, XML, or the like is an effective information collecting means.
As a means for displaying the hypertext, a hypertext display device generally called a browser is used, and in particular, a device which can be controlled by voice may be used.

A hypertext display device with a voice control function is disclosed, for example, in Japanese Patent Laid-Open No. 10-222342.
What is disclosed in the gazette is known. This hypertext display device has a function of extracting a set of phonetic notation and a command described in an HTML document, using the phonetic notation as a target word for speech recognition, and executing a command corresponding to the recognized phonetic notation. are doing.

[0004] As a method of enhancing the function of a browser, there is a method of using a relay device called a proxy.
This relay device is usually used to bypass a firewall that restricts external access to the computer network. A new service can be provided by adding various functions to this relay device. For example, as disclosed in Japanese Patent Application Laid-Open No. H10-154063, a method is known in which a speech recognition target word is extracted from a hypertext and a command set corresponding to the speech recognition target is added. .

[0005]

However, in the prior art, in the combination of the browser with the voice control function and the voice command assignment proxy, the hypertext described in a language other than the user's native language is controlled by voice. There is a problem that it is difficult to do. This is for the following reasons.

That is, the voice command assignment proxy uses morphological analysis to assign a phonetic notation to a speech recognition target word. Therefore, it is not possible to process a hypertext written in a language other than the native language. For example, Japanese morphological analysis cannot process hypertext written in English.

Therefore, a method of preparing a morphological analysis corresponding to a plurality of languages and processing a hypertext is also conceivable. However, since the phonetic notation of the speech recognition target word processed by this method is the phonetic notation of the national language of the speech recognition target word, it is difficult for a user whose native language is a language different from the national language to pronounce. For example, if the hypertext is written in English, the target words for speech recognition are extracted using morphological analysis in English, and even if a corresponding phonetic transcription is given, the phonetic transcription is the English phonetic transcription. Therefore, it is difficult for Japanese users to pronounce according to the phonetic transcription.

The present invention has been made in view of the above circumstances, and an object thereof is to provide a voice control function by uttering a specific language even for a hypertext described in a language other than a specific language. It is an object of the present invention to provide a hypertext relay method and apparatus capable of performing the above.

[0009]

In order to achieve the above object, according to the first aspect of the present invention, a hypertext designated by a hypertext display device having a voice control function is obtained from a network, and the hypertext display device is obtained. In the hypertext relay method of relaying to
When relaying the hypertext, specify a speech recognition target word from the hypertext, and when the specified speech recognition target word is described in a specific language, derive a phonetic notation corresponding to the speech recognition target word, If the specified speech recognition target word is described in a language other than the specific language, the phonetic transcription of the specific language corresponding to the speech recognition target word is derived, and the phonetic transcription and the hyper corresponding to the specified speech recognition target word are derived. We propose a feature in which a command for voice control in a text display device is added to hypertext.

The principle of the present invention will be described with reference to FIG. FIG. 1 is a diagram for explaining the operation principle of the present invention. As shown in FIG. 1, when a hypertext is designated from a hypertext display device having a voice control function (step S101), the designated hypertext is obtained from a computer network (step S1).
02). Then, when relaying the hypertext to the hypertext display device on a route in the network, the speech recognition target words included in the hypertext are specified,
The phonetic notation of the specific language corresponding to the speech recognition target word and the command for speech control in the hypertext display device are added to the hypertext (step S10).
3) The edited hypertext is relayed to the hypertext display device (step S104). Here, in the derivation of the phonetic notation in the step S103, when the speech recognition target word is described in a specific language, the phonetic notation in the language is derived, and the speech recognition target word is described in a language other than the specific language. In that case, the phonetic notation in the specific language is derived.

[0011] Thus, if the hypertext display device specifies the device according to the relay method of the present invention as the relay device, the hypertext display device will add the phonetic notation of the specific language corresponding to the speech recognition target to the requested hypertext. And a command to which a voice control command is added. Therefore, the user can perform voice control on the hypertext display device with the pronunciation of a specific language even if the hypertext is written in a language other than the specific language. The specific language is, for example, the user's native language.

According to a second aspect of the present invention, in the hypertext relay method according to the first aspect, when there are a plurality of specific languages, a phonetic transcription corresponding to each of the specific languages is derived. Suggest.

According to the present invention, when there are a plurality of specific languages, the phonetic notation for each specific language is derived for one speech recognition target word. Therefore, the user can perform voice control on the hypertext display device by pronouncing phonetic notation in one of a plurality of specific languages for one voice recognition target.

According to a third aspect of the present invention, in the hypertext relay method according to any one of the first to second aspects, the speech recognition target word is specified by a link from a hypertext. Is extracted, and the speech recognition target word is specified from the description of the extracted link.

Further, in the invention of claim 4, according to claims 1 to 3,
The hypertext relay method according to any one of the preceding claims,
The specification of the speech recognition target word proposes that a word determined as a noun by morphological analysis or pattern matching is specified as a speech recognition target word.

Further, according to the invention of claim 5, in claim 1 to claim 1
4. In the hypertext relay method according to any one of the above 4, the derivation of the phonetic transcription is performed by transliterating the speech recognition target word into a specific language when the speech recognition target word is described in a language other than the specific language. We propose a feature that derives the notation.

Further, according to the invention of claim 6, in a hypertext relay device for relaying hypertext to a hypertext display device having a voice control function via a route in a network, the hypertext relay device is requested by the hypertext display device. Reading means for reading hypertext via the network, and phonetic notation of a specific language corresponding to a speech recognition target word included in the hypertext read by the reading means and the hypertext display device And a transfer means for transferring the hypertext edited by the editing means to the hypertext display device.

The principle of the present invention will be described with reference to FIG. FIG. 2 is a configuration diagram of a hypertext relay device for explaining the principle of the present invention. Generally, a hypertext display device has a function of designating a relay device.
According to the present invention, the hypertext display device 200 uses the hypertext relay device 1 according to the present invention by using this function.
Specify 00. First, the hypertext display device 20
When 0 requests the hypertext relay device 100 for the hypertext, the reading unit 110 reads the hypertext from the network 300. The read hypertext is edited by the editing means 120 into the phonetic notation of a specific language corresponding to the speech recognition target word included in the hypertext and the hypertext display device 2.
A command for voice control at 00 is added. The edited hypertext is transferred to the hypertext display device 200 by the transfer unit 130.

Thus, the hypertext display device 2
00 can receive the requested hypertext with the phonetic notation of the specific language corresponding to the speech recognition target and the command for voice control added. Therefore,
The user, in the hypertext display device 200,
Even a hypertext described in a language other than a specific language can be voice-controlled by pronunciation in a specific language style. The specific language is, for example, the user's native language.

According to a seventh aspect of the present invention, in the hypertext relay device according to the sixth aspect, the editing means includes a link extracting means for extracting a description of the link from the read hypertext, and a description of the extracted link. A speech recognition target word specifying means for specifying a speech recognition target word from
If the specified speech recognition target word is described in a specific language, derive a phonetic transcription corresponding to the speech recognition target word,
When the specified speech recognition target word is described in a language other than the specific language, phonetic transcription deriving means for deriving a phonetic transcription of a specific language corresponding to the speech recognition target word, and the derived phonetic transcription and the hypertext display The present invention proposes an apparatus having an additional unit for adding a set of commands for voice control in the apparatus to hypertext.

Referring again to FIG. 2, the link extracting means 121 extracts the description of the link from the hypertext read by the reading means 110. Then, the target word for voice recognition specifying means 122 specifies the target word for voice recognition from the description of the link. If the phonetic recognition target word is described in a specific language, the phonetic notation deriving means 123 derives a phonetic notation corresponding to the voice recognition target word, and the specified voice recognition target word is other than the specific language. In the case of the description, the phonetic notation of the specific language corresponding to the speech recognition target word is derived. Then, the derived phonetic notation and command are added to the hypertext read by the reading means 110 by the adding means 124. By such an operation, the hypertext relay device according to claim 6 can be reliably implemented.

As a preferred embodiment of the present invention, in the invention of claim 8, in the hypertext relay device according to claim 7, the phonetic notation deriving means includes a speech specified by the speech recognition target word specifying means. For the target words,
In the case where the speech recognition target word is described in a language other than the specific word, a phonetic translation process is used to derive a phonetic transcription in the specific language.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A hypertext relaying method and apparatus according to an embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a network configuration diagram, and FIG. 4 is a functional block diagram of the hypertext relay device.

As shown in FIG. 3, a hypertext relay device (hereinafter referred to as a “relay device”) 10 is a WAN or L
It is interposed on a communication path connecting a computer network (hereinafter, referred to as “network”) 20 such as an AN and a hypertext display device (hereinafter, referred to as “display device”) 30 having a voice control function. Relay device 1
0 and the network 20 are connected via a communication line 41. The relay device 10 and the display device 30 are connected via a communication line 42. When the relay device 10 is requested to acquire the hypertext from the display device 30,
The server 2 connects to the hypertext server 21 connected to the network 20 as a proxy for the display device 30.
1 and obtains the hypertext from the display device 30.
It has a relay function to transfer to The present invention is characterized in that, when relaying a hypertext, predetermined editing is performed on the hypertext.

The relay unit 10, as shown in FIG.
1, a buffer 12 as storage means for temporarily storing the read hypertext, and a transfer unit 13 for transferring the hypertext stored in the buffer 12 to the display device 30.

The relay device 10 has the following components for editing hypertext stored in the buffer 12. That is, the relay device 10
Receives a description of the link included in the hypertext by referring to the hypertext stored in the buffer 12 and a description of the link extracted by the link extraction unit 14; A speech recognition target word specifying unit 15 for specifying a speech recognition target word from a description of a link
A phonetic notation deriving unit 16 that derives a phonetic notation for the speech recognition target word specified by the speech recognition target word specifying unit 15
And an adding unit 17 for adding the phonetic notation and voice control command derived by the phonetic notation deriving unit 16 to the hypertext stored in the buffer 12.

The link extracting section 14 extracts a description of a link included in the hypertext. For example, when the hypertext is HTML, a portion surrounded by <A> and </A> tags is extracted as a description of a link. Specific methods of extraction include, for example, a pattern matching method and a syntax analysis method.

The speech recognition target word specifying section 15 includes a morphological analysis section 18. Then, the morphological analysis unit 18 performs morphological analysis on the description of the target link, and specifies a speech recognition target word. Here, the morphological analysis unit 18 divides the target sentence into morpheme units, and extracts a morpheme having a specific property from the obtained morphemes. For example, nouns included in a sentence are extracted. This allows
The speech recognition target word specifying unit 15 specifies a noun included in the description of the link as a speech recognition target word. The morphological analysis unit 18 supports a plurality of languages.

The phonetic notation deriving unit 16 is a means for deriving a phonetic notation in a specific language for a word to be recognized. That is, if the target speech recognition word is described in a specific language, the phonetic transcription of the word is directly derived. If the target speech recognition target word is described in a specific language, the pronunciation in the specific language is performed by transliteration processing. Derive the notation. Here, the derived phonetic notation is described in a notation method that can be processed by the voice control function of the display device 30.

When the target word for speech recognition is described in a specific language, for example, when the specific language is Japanese, "Sumou" is derived as a phonetic notation for the target word for speech recognition "Sumo". I do. If the specific language is English, the phonetic notation “# an0Im0 @ l” is derived for the speech recognition target word “animal”.

As a means of the transliteration processing, there is a general means for registering a speech recognition target word and its phonetic notation in the dictionary 19. For example, when the specific language is Japanese, the phonetic notation “station” is derived for the English-language speech recognition target word “station”. If the specific language is English, the phonetic notation "s'um% oU" is derived for the Japanese word for speech recognition "Sumo".

When there are a plurality of specific languages, the phonetic notation deriving unit 16 derives the phonetic notation as described above for each language. For example, when the specific languages are Japanese and English, the phonetic notation “Sumou” and “s'um% oU” are derived for the speech recognition target word “Sumo”, and the speech recognition target word “a
nimal "for phonetic notation" animal "and"# an0Im0 @ l "
Is derived.

The adding unit 17 adds, to the hypertext stored in the buffer 12, a set of phonetic notation derived by the phonetic notation deriving unit 16 and a command for voice control in the display device 30. Here, the set of phonetic notation and command is added in a format that can be interpreted on the display device 30. Here, the command to be added includes the description of the link destination in the description of the link that included the speech recognition target word.

An example of a specific hardware configuration of the relay device 10 will be described with reference to FIG. FIG. 5 is a hardware configuration diagram of the hypertext relay device.

As shown in FIG. 5, the relay device 10 includes a main processing unit 51, a main storage device 52 composed of a volatile storage device such as a DRAM, and an auxiliary storage device composed of a nonvolatile storage device such as a fixed magnetic storage device. The device 53 and the network 20
A first network interface 54 that connects to
The main components are a second network interface 55 connected to the display device 30 and a bus 56 connecting them. The main storage device 52 or the auxiliary storage device 53 stores a hypertext relay program. The relay apparatus 10 operates by the main processing unit 51 executing the hypertext relay program, thereby realizing the relay of the hypertext.

Next, the display device 30 will be described with reference to FIG. FIG. 6 is a configuration diagram of the hypertext display device.

As shown in FIG. 6, the display device 30 includes a main device 31 connected to the relay device 10, a voice input device 32 composed of a device such as a microphone and a noise removing device for inputting a user's voice, and a hyper The display unit 33 that displays text to the user is a main component. The display device 30 can display a hypertext via the relay device, and the above-described relay device 10 is designated.

The main unit 31 includes a communication control unit 34 for controlling communication with the relay device 10, a request signal transmitting unit 35 for transmitting a request signal for acquiring a hypertext, and a document receiving unit for receiving the transmitted hypertext. Unit 36, a voice recognition processing unit 37 that performs voice recognition processing on a signal input from the voice input device 32, and a display unit 3 that displays the received hypertext.
3 for interpreting and processing so that it can be displayed, interpreting and processing commands included in the voice information processed by the voice recognition processing unit 37, and acquiring the hypertext to the request signal transmitting unit 35 according to the content of the interpretation. The main component is a document designating a request and a command interpretation processing unit 38.

Next, a method of relaying a hypertext will be described with reference to the flowchart of FIG. FIG. 7 is a flowchart illustrating a method for relaying a hypertext.

First, the display device 3 is operated by a user's instruction or the like.
0 designates a hypertext for the relay device 10 (step S1).

The reading unit 11 of the relay device 10 reads the designated hypertext from the network 20 via the communication line 41 and stores the hypertext in the buffer 1.
2 (step S2).

Next, the link extracting unit 14
Then, a syntactic analysis or a pattern matching process is performed on the hypertext stored in the file to extract the description of the link (step S3). Here, a link destination is specified in the description of the extracted link.

Next, the speech recognition target word identifying unit 15 extracts a noun by analyzing the description of the link extracted by the link extraction unit 14 using the morphological analysis unit 18 and extracts the noun. (Step S
4).

Next, the phonetic notation deriving unit 16 derives a phonetic notation for the speech recognition target word specified by the speech recognition target word specifying unit 15 (step S5). The derived phonetic notation is temporarily stored in the buffer 12.

Next, the adding unit 17 adds the stored phonetic notation and the command corresponding to the speech recognition target word to the hypertext stored in the buffer 12 (step S6).

Finally, the transfer unit 13 takes out the edited hypertext from the buffer 12 and transfers it to the display device 30 (step S7).

Through the above steps, the phonetic notation of the speech recognition target word and the command used in the speech recognition control of the display device 30 are added to the hypertext relayed to the display device 30. Thus, the display device 30 can perform voice control based on the added description.

[0048]

Next, a specific example of a hypertext relay method and apparatus according to the present invention will be described with reference to the drawings. FIG. 8 is a diagram illustrating an example of the hypertext.

Here, HTML will be described as an example of hypertext. As the hypertext display device 30 with the voice control function, Java Ap
Plet can be executed, and this JavaApplet
An example provided with a voice control function is illustrated by t. To perform voice control in the display device 30, a description related to JavaApplet may be added to the HTML document.

Hereinafter, description will be made with reference to the flowchart of FIG. Here, a case where the specific language is Japanese, which is the user's native language, will be described.

First, the display device 30 selects the relay device 10 using the function of designating the relay device, and selects the relay device 10.
Specify a hypertext for (Step S
1).

Thus, the reading unit 11 of the relay device 10
Reads the hypertext specified from the network 20 via the communication line 41 and stores the hypertext in the buffer 12 (step S2). The read hypertext is shown in List A of FIG.

Next, the link extracting unit 14
Then, a syntactic analysis or a pattern matching process is performed on the hypertext stored in the file to extract the description of the link (step S3). Here, in the description of the extracted link, as shown in FIG. 8, a link destination indicated by <AHREF =...> Is specified.

Next, the speech recognition target word specifying unit 15 extracts the noun by analyzing the description of the link extracted by the link extraction unit 14 using the morphological analysis unit 18 and extracts the noun. (Step S
4). In the example of FIG. 8, “sumo” is specified from “watch sumo”, “animal” is specified from “the big animal”,
Specify "iron" from "iron".

Next, the phonetic notation deriving unit 16 derives a phonetic notation for the speech recognition target word specified by the speech recognition target word specifying unit 15 (step S5). Here, since the specific language is Japanese, "Sumo" is derived for "Sumo", "Animal" is derived for "animal",
Specify "iron / iron" for "iron".
Here, “/” included in the phonetic notation is a symbol for writing a plurality of phonetic notations in parallel. The derived phonetic notation is temporarily stored in the buffer 12.

Next, the adding unit 17 adds the stored phonetic notation and the command corresponding to the speech recognition target word to the hypertext stored in the buffer 12 (step S6). As a result, the hypertext shown in the list B of FIG. 8 is obtained.

Finally, the transfer unit 13 takes out the edited hypertext from the buffer 12, and transfers it to the display device 30 (step S7).

Through the above steps, the phonetic notation of the speech recognition target word and the command used in the speech recognition control of the display device 30 are added to the hypertext relayed to the display device 30. Thus, the display device 30 can perform voice control based on the added description. Specifically, when the user utters the text corresponding to the link, the screen can be switched to the hypertext of the link destination.

When the user has a plurality of native languages, these native languages may be set as specific languages.
For example, if English and Japanese are the specific languages, the hypertext shown in List C in FIG. 8 can be obtained. This means that anyone who pronounces English in Japanese in their native language, who pronounces Japanese in English in their native language, and who pronounces Japanese and English like native speakers Reliable voice control can be performed for a person.

Although the method and apparatus for relaying hypertext according to the present invention have been described in detail, the present invention is not limited to the above embodiments and examples. For example,
In the above-described embodiment, the description has been made with reference to HTML as the hypertext. However, the present invention is not limited to this.
The present invention can also be implemented in L or XML. Also, the case where the language described in the hypertext and the specific language are Japanese and English has been described, but the present invention is not limited to this, and the present invention may be implemented in other languages. it can.

In the above-described embodiments and examples, the transliteration process using the dictionary is performed when the speech recognition target word is described in a language other than the specific language. However, the present invention is not limited to this. is not. For example, a transliteration model is created by statistically processing a large number of words other than the specific language and the corresponding phonetic transcription data of the specific language. And
A method of calculating the probability based on this transliteration model and searching for the most appropriate transliteration notation can be considered. in this way,
It is possible to eliminate the drawbacks of dictionary maintenance costs and the inability to deal with other than registered words, which are problems in transliteration processing using a dictionary.

In the above-described embodiments and examples, the speech recognition target word specifying unit extracts a noun from the hypertext and specifies the noun as the speech recognition target word.
The present invention is not limited to this. That is, another part of speech, for example, an adjective, an adjective verb, a verb, an adverb, or the like may be specified as a speech recognition target word.

In the above-described embodiments and examples, the speech recognition target word is extracted from the description of the link included in the hypertext, and the command relating to the link is exemplified as a command added to the hypertext. It is not limited to this. For example, a speech recognition target word may be extracted from a description in a specific format. As a command to be added to the hypertext, a command related to a format may be added. If this is used, it is also possible to perform voice control such as changing the color of the screen, the size of characters, and the like by the user uttering the target word for voice recognition.

[0064]

As described in detail above, according to the present invention,
If the hypertext display device specifies the device according to the relay method of the present invention as a relay device, the hypertext display device can, of course, not only perform hypertext written in a specific language such as a native language, but also use a language other than the specific language. A user who cannot accurately pronounce a hypertext written in a language other than that language in a language other than the specific language can also perform voice control using pronunciation in a specific language style such as a native language.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the operating principle of a hypertext relay method.

FIG. 2 is a configuration diagram of a hypertext relay device for explaining the principle of the present invention;

FIG. 3 is a network configuration diagram.

FIG. 4 is a functional block diagram of a hypertext relay device.

FIG. 5 is a hardware configuration diagram of the hypertext relay device.

FIG. 6 is a configuration diagram of a hypertext display device.

FIG. 7 is a flowchart illustrating a method for relaying a hypertext.

FIG. 8 is a diagram showing an example of a hypertext.

[Explanation of symbols]

10: hypertext relay device, 11: reading unit, 12
... Buffer, 13 ... Transfer unit, 14 ... Link extraction unit, 15
... Speech recognition target word specifying unit, 16 ... Pronunciation notation deriving unit, 17
... Addition unit, 18 ... Morphological analysis unit, 19 ... Dictionary, 20 ... Computer network, 30 ... Hypertext display device, 31 ... Main device, 32 ... Speech input device, 33 ... Display unit, 34 ... Communication control unit, 35 ... Request signal transmitter, 36 ...
Document receiving unit, 37 Voice recognition processing unit, 38 Document and command interpretation processing unit, 41, 42 Communication line, 100 Hypertext relay device, 110 Reading means, 120
Editing means, 121: Link extraction means, 122: Voice recognition target word identification means, 123: Phonetic notation derivation means, 124 ...
Phonetic notation and command addition means, 125 ... transfer means, 2
00: computer network, 300: hypertext display device

Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (reference) // G06F 17/30 G06F 15/419 320 F term (reference) 5B075 KK01 KK07 ND03 ND20 ND36 PQ02 PQ04 PQ20 PQ32 PQ42 UU01 UU06 5B091 AA01 BA02 CA02 CB24 CB32 CC02 CD03 CD15 DA06 5D015 GG01 GG03 KK01 9A001 GG03 HH17 HH33 JJ26

Claims (8)

[Claims] 1. A hypertext relay method for acquiring a hypertext specified by a hypertext display device having a voice control function from a network and relaying the hypertext to the hypertext display device. From the speech recognition target word, and if the specified speech recognition target word is described in a specific language, derive a phonetic notation corresponding to the speech recognition target word,
If the specified speech recognition target word is described in a language other than the specific language, a phonetic transcription of the specific language corresponding to the speech recognition target word is derived, and the phonetic transcription and the hyper corresponding to the specified speech recognition target word are derived. A hypertext relay method, wherein a command for voice control in a text display device is added to hypertext. 2. The hypertext relay method according to claim 1, wherein when there are a plurality of specific languages, phonetic notations corresponding to each of the specific languages are respectively derived. 3. The speech recognition target word is specified by extracting a description of a link from a hypertext and specifying the speech recognition target word from the extracted link description. 2. The hypertext relay method according to claim 1. 4. The speech recognition target word according to claim 1, wherein the speech recognition target word is specified by specifying a word determined as a noun by morphological analysis or pattern matching as a speech recognition target word. Hypertext relay method. 5. The phonetic notation derivation is characterized in that, when the speech recognition target word is described in a language other than a specific language, the phonetic notation is derived by transliterating the speech recognition target word into a specific language. The hypertext relay method according to claim 1. 6. A hypertext relay device for relaying a hypertext to a hypertext display device having a voice control function via a route in a network, wherein the hypertext requested from the hypertext display device is transmitted via the network. Reading means for reading and reading, for the hypertext read by the reading means, a phonetic notation of a specific language corresponding to a speech recognition target word included in the hypertext and a command for voice control in the hypertext display device. A hypertext relay device comprising: an editing unit for adding; and a transfer unit for transferring the hypertext edited by the editing unit to the hypertext display device. 7. The linking means for extracting a description of a link from the read hypertext; a speech recognition target word specifying means for specifying a speech recognition target word from the extracted link description; If the speech recognition target word is described in a specific language, derive a phonetic transcription corresponding to the speech recognition target word,
If the specified speech recognition target word is described in a language other than the specific language, phonetic transcription deriving means for deriving a phonetic transcription of a specific language corresponding to the speech recognition target word, and the derived phonetic transcription and the hypertext display 7. The hypertext relay device according to claim 6, further comprising an adding unit for adding a set of commands for voice control in the device to the hypertext. 8. The phonetic notation deriving means, if the speech recognition target word specified by the speech recognition target word specifying means is described in a language other than the specific word, transliteration processing. The hypertext relay device according to claim 7, wherein the phonetic transcription in the specific language is derived by the following.