JP4014361B2 - Speech synthesis apparatus, speech synthesis method, and computer-readable recording medium recording speech synthesis program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that a voice synthesizing which reads a hypertext aloud can not read aloud information at a link destination connected in a network state by a hyperlink. SOLUTION: This voice synthesizing device has a means which extracts link destination information according to property information showing a link destination in an input hypertext, a means which obtains the hypertext corresponding to the link destination information, a means which generates a read- aloud text at the link destination according to the obtained hypertext, and a means which synthesizes a voice of the generated read-aloud text at the link destination.

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a speech synthesizer, a speech synthesis method, and a computer-readable recording medium on which a speech synthesis program is recorded. In particular, the present invention relates to a speech synthesizer that reads out hypertext, a speech synthesis method, and a computer-readable recording medium that records a speech synthesis program.
[0002]
[Prior art]
Conventionally, a text-to-speech synthesis technique that reads a sentence in a kanji-kana mixed sentence is known ("Digital Speech Processing" by Furui (Tokai University Press).
[0003]
On the other hand, in recent years, with the development of the Internet, a hypertext reading device has been proposed that reads hypertext described in HTML (Hyper Text Markup Language) defined by ISO / IEC15445 as a description format of information in WWW (World WideWeb). (JP-A-8-83089). According to this hypertext reading device, a hyperlink portion is detected from the text, and the hyperlink portion is read out and the normal portion of the read-out speech is changed and output. Can be perceived by voice.
[0004]
[Problems to be solved by the invention]
However, although the above hypertext reading apparatus can make the user perceive the presence of a hyperlink in the text from the difference in the reading voice, since the contents of the link destination cannot be read out, the link destination It was impossible to inform the user what kind of information exists.
[0005]
The main object of the present invention is to provide a speech synthesizer that can read out information of a link destination connected in a network by a hyperlink in a speech synthesizer that reads out hypertext.
[0006]
[Means for Solving the Problems]
The present invention analyzes display means for displaying information indicated by hypertext, and attribute information indicating a link destination in the input hypertext, and from the input hypertext according to the analysis result. Represents the destination of hypertext on the Internet Link destination information extracting means for extracting link destination information indicating the link destination, and corresponding to the link destination information On the internet Hypertext Is obtained from the link destination on the Internet indicated by the link destination information, and then the hypertext And a link destination reading text extracting means for extracting the link destination reading hypertext from the hypertext based on the markup information that is included in the reading text and the extracted reading destination reading hypertext. A link-to-speech text generation unit that generates a link-to-speech text based on the input hypertext, and a display process by the display unit for the input hypertext, Extracted from hypertext obtained from the Internet and There is provided a speech synthesizer characterized by comprising speech synthesizer that synthesizes speech of a link destination read-out text generated.
[0007]
According to the present invention, the information of the link destination in the input hypertext is acquired and the link destination information is handled. On the internet Hypertext Is obtained from the link destination on the Internet indicated by the link destination information, and then the hypertext Can generate the text to be read out from the link based on the markup information. Reading text extracted and generated from hypertext Can be read aloud. At that time, since the display process by the display means for the input hypertext is maintained, it is possible to avoid a troublesome work of returning to the display of the original text after displaying the link destination. Furthermore, if markup information for designating text to be read out is inserted in advance in the hypertext at the link destination, it is possible to read out only the text of the designated part, and extra information There is no need to read it aloud.
[0008]
Further, the present invention provides a text-to-speech text generating means for generating text-to-speech text based on the input hypertext, and switching between the generated text-to-speech text and the generated text to be read at the link destination. There is provided a speech synthesizer characterized by further comprising a switched speech synthesizer for synthesizing.
[0009]
According to the present invention, it is possible to switch and output the text to be read out from the text and the text to be read out from the link destination. Therefore, the text text to be read out from the text and the text information from the link destination among the information indicated by the input hypertext. Can be synthesized in the order desired by the user.
[0010]
The present invention further includes an input means for receiving an input of a desired link destination, and when the input of the link destination is received by the input means, the read-out text of the link destination is voiced by the switching voice synthesizing means. On the other hand, a speech synthesizer is provided that displays the text to be read out by the display means on the display means.
[0011]
According to the present invention, when the user inputs a desired link destination, the link destination information can be synthesized, while the currently displayed text information can be displayed as it is. It is possible to avoid the troublesome work of returning to the display of the original text.
[0012]
Further, according to the present invention, the input of the link destination by the input means is performed by moving the pointer including the cursor to display information including a character or an image representing the link destination in the read text of the displayed text. A speech synthesizer is provided.
[0013]
The present invention further includes selection means for selecting whether or not to display information corresponding to the link-to-speech text when speech synthesis of the link-to-speech text is requested. A speech synthesizer is provided.
[0014]
According to the present invention, while viewing a currently displayed web page, the linked information can be acquired by voice synthesis without switching the display to another web page, and when the user desires It is possible to switch to the display corresponding to the information currently being synthesized.
[0015]
Further, the present invention analyzes the attribute information indicating the link destination in the input hypertext, the step of displaying the information indicated by the input hypertext, and from the input hypertext according to the analysis result. Represents the destination of hypertext on the Internet A step of extracting link destination information indicating a link destination, and corresponding to the link destination information On the internet Hypertext Is obtained from the link destination on the Internet indicated by the link destination information, and then the hypertext And a link destination based on the extracted hypertext of the link destination extracted from the hypertext based on markup information indicating that the text is a text to be read Generating the read-out text and maintaining the display processing for the input hypertext, Extracted from hypertext obtained from the Internet and And a step of synthesizing the generated text to be read at the link destination.
[0016]
According to the present invention, the information of the link destination in the input hypertext is acquired and the link destination information is handled. On the internet Hypertext Is obtained from the link destination on the Internet indicated by the link destination information, and then the hypertext Can generate the text to be read out from the link based on the markup information. Reading text extracted and generated from hypertext Can be read aloud. At that time, since the display process for the input hypertext is maintained, it is possible to avoid the troublesome work of returning to the display of the original text after displaying the link destination. Furthermore, if markup information for designating text to be read out is inserted in advance in the hypertext at the link destination, it is possible to read out only the text of the designated part, and extra information There is no need to read it aloud.
[0017]
The present invention also provides a computer-readable recording medium on which a speech synthesis program for causing a computer to execute each step of the speech synthesis method is recorded.
[0018]
According to the present invention, link destination information in the input hypertext From hypertext obtained from the link destination on the Internet Since the text to be read out can be generated, the link destination in the input hypertext Reading text extracted and generated from hypertext Can be read aloud.
[0019]
Further, the present invention has link destination reading text generation means for generating a reading text of the link destination text, and generates a reading text of the link destination by processing a part or all of the acquired hypertext. It is desirable to do so.
[0020]
According to the above-described configuration, since part or all of the acquired hypertext information is processed using text summarization technology, it is possible to generate text that summarizes link destination information, and the amount of link destination information. Even when there are many, it becomes possible to synthesize appropriate information.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
(Embodiment 1)
FIG. 1 is a functional block diagram showing an embodiment of a speech synthesizer according to the present invention. The speech synthesizer 100 includes a hypertext interpretation unit 101, a reading text generation unit 102, a language processing unit 103, a speech synthesis unit 104, and a speech output unit 105.
[0022]
In the present embodiment, the case where speech synthesis apparatus 100 is built in a general-purpose personal computer and hypertext received via a network is synthesized by speech and output will be described, but the present invention is limited to this. However, the present invention can be applied to all devices that perform speech synthesis on input hypertext.
[0023]
As shown in FIG. 1, hypertext is input to the speech synthesizer 100. This hypertext is described in a markup language such as HTML which is widely known as a language for describing a document on a WWW server. Further, in this hypertext, a URL (Uniform Resource Locator) that is information for indicating the location of a resource on the Internet, a sentence, and the like are embedded. In general, making it possible to connect to various servers using a browser on the Internet is called a link, and in particular, designating a connection destination using hypertext is called a hyperlink. In the following description, hyperlinks are simply called links unless otherwise specified. A connection destination specified by a URL embedded in hypertext is called a link destination.
[0024]
The hypertext interpretation unit 101 analyzes attribute information described in the input hypertext. As a result of the analysis, information indicating the content of the main body of the input hypertext (hereinafter referred to as text to be read out) is output to the language processing unit 103, while a URL indicating the link destination (hereinafter referred to as link destination information) is a character string. And is output to the text-to-speech generation unit 102.
[0025]
The text-to-speech generation unit 102 connects to a link destination indicated by the link destination information output from the hypertext interpretation unit 101 via a network, and displays hypertext indicating information recorded in the link destination (hereinafter referred to as link information). A text is acquired, and a text to be read out at the link destination is generated based on the attribute information in the hypertext and output to the language processing unit 103.
[0026]
The language processing unit 103 receives the text-to-speech text output from the hypertext interpretation unit 101 and the text-to-link text to be output from the text-to-speech generation unit 102, analyzes the text information indicated by the text, and reads the text information to be read out. A character string to be represented and an array each storing a pitch and a duration value corresponding to the character string are output to the speech synthesizer 104.
[0027]
The speech synthesizer 104 generates a time-series numerical array representing the speech waveform from the information output from the language processing unit 103 and outputs it to the speech output unit 105.
[0028]
The voice output unit 105 includes a voice output device such as a speaker, and outputs the synthesized waveform output from the voice synthesizer 104 after D / A conversion.
[0029]
FIG. 2 illustrates a specific example in which hypertext is input to the speech synthesizer 100 according to the present embodiment, and the hypertext interpretation unit 101 and the read-out text generation unit 102 generate text information to be read out from the hypertext. It is a conceptual diagram for doing.
[0030]
FIG. 2A shows an example of input hypertext. In FIG. 2 (a), <a href=...> to </a> are descriptions of links, and <ahref = ...> and </a> are attribute information indicating links. Further, “hokkaido.html” and “touhoku.html” are link destination information. In this example, it is shown that this hypertext is linked to Hokkaido weather information and Tohoku weather information.
[0031]
FIG. 2B is an example of a link destination read-out text generated by the read-out text generation unit 102. The text to be read at the link destination is information stored in the server at the link destination (hokaido.html). In the hypertext interpretation unit 101, attribute information (<a href=...> And </a>) indicating the link is displayed. ) Is extracted from the link destination information (hokaido.html), and the text-to-speech generation unit 102 connects to the link destination and acquires hypertext corresponding to the link destination information. Thereafter, the text to be read out at the link destination is generated based on the attribute information in the acquired hypertext. In this case, information corresponding to all attribute information in the hypertext may be generated as read-out text at the link destination, or information corresponding to attribute information indicated other than the attribute information indicating the link in the hypertext. May be generated as a text to be read out at the link destination.
[0032]
FIG. 2C is an example of text-to-speech text generated by the hypertext interpretation unit 101. This text-to-speech text is information indicated by attribute information other than attribute information indicating a link in the hypertext.
(Embodiment 2) FIG. 3 is a functional block diagram showing another embodiment of the speech synthesizer according to the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the same description is omitted.
[0033]
The reading text management unit 106 manages the text information input from the hypertext interpretation unit 101 and the reading text generation unit 102 so as to be output to the language processing unit 103 in an appropriate order.
[0034]
FIG. 4 is a flowchart showing an example of processing in the reading text management unit 106. The processing of the text-to-speech management unit 106 will be described with reference to FIG.
[0035]
When text information is input from the hypertext interpretation unit 101 and the read text generation unit 102 to the read text management unit 106, each character string constituting the text information is distinguished and stored (S101). That is, the character string of the text-to-speech text input from the hypertext interpretation unit 101 is stored in the array H, and the character string of the link-to-speech text input from the text generation unit 102 is stored in the array Y. At this time, assuming that the number of character strings input from the hypertext interpretation unit 101 is N and the number of character strings input from the read-out text generation unit 102 is M, the arrays H and Y have H [1 ], H [2],... H [N], Y [1], Y [2]..., Y [M] are stored.
[0036]
Next, 1 is assigned to the loop control variable i (S102), and H [i] is output to the language processing unit 103 (S103). Next, 1 is added to the control variable i (S104), and it is determined whether or not the condition of i <N is satisfied (S105). If the condition is satisfied, the process returns to S103, and the process is repeated until the character string input from the hypertext interpretation unit 101 is output. When the output is completed, the control variable i is reinitialized to 1 (S106), and Y [i] is output to the language processing unit 103 (S107). Next, 1 is added to the control variable i (S108), and it is determined whether or not the condition of i <M is satisfied (S109). If the condition is satisfied, the process returns to S107, and the process is repeated until the character string input from the text-to-speech generation unit 102 has been output.
[0037]
Through the above processing, it is possible to first synthesize text-to-speech text and then synthesize text-to-link text.
[0038]
In this embodiment, the text information is managed so that the text to be read out is output first, and then the text to be read out at the link destination is output. However, the text information is output in the reverse order or alternately. You may make it output. In other words, the read-out text management unit 106 sets whether to switch the output of the text-to-speech text and the output of the text-to-link text, or restricts whether to output the text-to-speech text or the text to be read out from the link destination. You may make it.
(Embodiment 3) FIG. 5 is a functional block diagram showing another embodiment of the speech synthesizer according to the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the same description is omitted.
[0039]
The text summarization unit 107 receives the read-out text to be linked from the read-out text generation unit 102, summarizes the text information, and returns the summary information to the read-out text generation unit 102. The text information may be summarized using an existing text summarization technique.
[0040]
With such a configuration, it is possible to generate text that summarizes the link destination information set in the input hypertext, so that even if there is a large amount of link destination information, speech synthesis of appropriate information Is possible.
(Embodiment 4) FIG. 6 is a functional block diagram showing another embodiment of the speech synthesizer according to the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the same description is omitted.
[0041]
The text translation unit 108 receives the read-out text to be linked from the text-to-speech generation unit 102, translates the text information into a preset language used by the user, and returns the translation information to the text-to-speech generation unit 102. . The translation of the text information may be performed by using an existing natural language automatic translation technique (supervised by Makoto Nagao, “Japanese Information Processing” The Institute of Electronics and Communication Engineers).
[0042]
With this configuration, the link destination information set in the input hypertext can be translated into the language used by the user in advance. It is possible to synthesize speech in an understandable language.
(Embodiment 5) FIG. 7 is a functional block diagram showing another embodiment of the speech synthesizer according to the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the same description is omitted.
[0043]
The text-to-speech extraction unit 109 receives link information from the text-to-speech generation unit 102, and when the link information includes attribute information for text-to-speech, extracts the part and reads the text-to-speech information to the text-to-speech generation unit 102. Return to.
[0044]
The attribute information for text-to-speech will be described with reference to FIG.
[0045]
FIG. 8 is a diagram for explaining an example of extracting a part from hypertext in which attribute information for text reading is set. In FIG. 8A, <speak> indicates text reading attribute information, and text information sandwiched between <speak> (text reading start tag) and </ speak> (text reading end tag) is included. As shown in FIG. 8B, it is extracted as read-out information.
[0046]
In this way, when information to be read out is designated in advance in the link information, only that portion can be read out, and unnecessary information need not be read out.
(Embodiment 6) FIG. 9 is a functional block diagram showing another embodiment of the speech synthesizer according to the present invention. The same components as those in FIG. 1 are denoted by the same reference numerals, and the same description is omitted.
[0047]
The display unit 112 includes a display device such as a liquid crystal display, and displays information indicated by hypertext. When displaying, software for displaying hypertext such as an existing WWW browser is used.
[0048]
The input unit 111 receives an input of a link destination desired by the user. The character string indicating the link destination is output to the reading text generation unit 102 and the hypertext input unit 110. The input unit 111 includes input devices such as a keyboard and a mouse, and a URL is typed with the keyboard via the WWW browser screen displayed on the display unit 112, or a character or image representing a link destination is clicked with the mouse. As a result, the corresponding character string is output to the reading text generation unit 102 and the hypertext management unit 110.
[0049]
The hypertext management unit 110 determines and outputs the hypertext that is the basis of the Web page displayed on the display unit 112 via the WWW browser. For example, in an initial state such as when a personal computer equipped with the speech synthesizer according to the present invention is activated, the display unit 112 is determined to display a start page set in advance by the WWW browser. Further, when a character string indicating a link destination is input from the input unit 111 or the read-out text generation unit 102, hypertext representing link destination information corresponding to the character string is acquired, and the hypertext is displayed on the display unit 112. And output to the hypertext interpretation unit 101.
[0050]
Further, when the user designates a link destination on the web page currently displayed on the display unit 112, the web page currently displayed is kept displayed as it is, while the information on the designated link destination is read out. Alternatively, it is possible to provide an interface that allows the user to select whether to switch the currently displayed web page to the linked web page and to read out the specified linked page information. In the latter case, for example, an image representing a button for switching pages is displayed on the display unit 112. When the button is selected by the input unit 111, the hypertext management unit 110 is currently reading it out. What is necessary is just to acquire the hypertext of the link destination corresponding to a sentence, and to output the information which the hypertext shows to the display part 112.
[0051]
With such a configuration, it is possible to obtain link destination information by voice synthesis without visually switching the currently displayed web page and switching the display to another web page, and when the user desires It is possible to switch to the display corresponding to the information currently being synthesized.
[0052]
In addition, when the input unit 111 moves the cursor to a character or image representing a link destination with the mouse, the link destination character string may be output to the read-out text generation unit 102. With such a configuration, it is possible to read out the information of the designated link destination by speech synthesis without changing the currently displayed Web page.
[0053]
Also, the character string corresponding to the link destination currently read out by speech synthesis is output from the reading text generation unit 102 to the display unit 112, and the character corresponding to the link destination is changed from other characters, for example, blinking. By displaying it, it is possible to notify the user of the link destination currently being read out by speech synthesis. As for how to change the display, in addition to blinking, for example, an image of a speaker may be displayed near the linked character or image.
[0054]
Also, as shown in FIG. 10, a link destination display unit 113 may be provided, and a character string representing a link destination corresponding to a text to be read out by voice may be displayed there.
[0055]
In the present embodiment, when the link destination is clicked with the mouse in the input unit 111, the link destination character string is output to the hypertext management unit 110, and when the cursor is moved to the link destination, the link destination Are output to the text-to-speech generator 102, but their output destinations may be reversed or output to both.
(Embodiment 7) The speech synthesizer according to the present embodiment has a configuration in which the speech synthesizer according to Embodiments 3 to 6 described above is combined, that is, each configuration of the speech synthesizer according to Embodiment 1. In addition, a speech synthesizer (not shown) configured to include a text summarization unit 107, a text translation unit 108, a reading text extraction unit 109, a hypertext management unit 110, an input unit 111, a display unit 112, and a link destination display unit. It is. In the present embodiment, processing in the text-to-speech generation unit 102 when a WWW browser is activated and a start page is displayed on a personal computer equipped with this speech synthesizer will be described.
[0056]
FIG. 11 is a flowchart illustrating an example of processing in the read-out text generation unit 102. The processing of the text-to-speech generation unit 102 will be described with reference to FIG.
[0057]
When the WWW browser is activated, hypertext corresponding to the start page is input to the hypertext management unit 110 and the start page is displayed on the display unit 112, and attribute information in the hypertext is analyzed by the hypertext interpretation unit 101. The link destination information is input to the read-out text generation unit 102.
[0058]
The link destination information is represented by an array A having a character string as an element. If the number of elements is N, the link destination information is represented as A [1], A [2],. When the link destination information is input to the read-out text generation unit 102, 1 is assigned to the loop control variable i (S201). Next, a text to be read out is generated from the link destination information corresponding to A [i], and stored in correspondence with a character string representing the link destination (S202). Detailed processing in S202 will be described later.
[0059]
Next, 1 is added to the control variable i (S203), and it is determined whether or not the condition of i <N is satisfied (S204). If the condition is satisfied, the process returns to S202, and the process is repeated until the end of the character string of the link destination information (A [n]).
[0060]
In S204, when the processing is completed up to the last character of the link destination information, data from the input unit 111 is received next. When a character string designating a link destination is input from the input unit 111, a character string representing the text to be read corresponding to the designated link destination is output to the language processing unit 103, and the character string representing the link destination is linked to the link destination. It is displayed on the display unit 113. When a display page switching instruction is input from the input unit 111, the character string displayed on the link destination display unit 113 is output from the reading text generation unit 102 to the hypertext management unit 110, and the link destination page is displayed. It is displayed on the display unit 112. In S205, if there is no input from the input unit 111, the process waits until there is an input.
[0061]
FIG. 12 is a detailed flowchart of the process in S202 described above.
[0062]
Link information is acquired from the link destination specified by the character string A [i] representing the link destination information and stored in the buffer. The link information can be acquired using a well-known HTTP protocol (hypertext transfer protocol) or the like.
[0063]
Next, it is determined whether or not the acquired link information includes attribute information for text reading indicated by, for example, <speak> (S302). If there is attribute information for text reading, the process proceeds to S306, where the text Read-out information is extracted, and the information is stored as read-out text information in association with a character string representing a link destination.
[0064]
If there is no text-to-speech attribute information in S302, the process proceeds to S303 to determine whether or not the link information needs to be translated. The language described in the header portion of the hypertext of the link information is compared with the language used by the user. If they are different, it is determined that translation is necessary, and the link information is translated (S304). If they match, it is determined that translation is unnecessary, and the process proceeds to S305.
[0065]
In S305, the link information is summarized and stored as read-out text information in association with a character string representing the link destination.
[0066]
It is also possible to provide a speech synthesizer that arbitrarily combines the configurations of the speech synthesizers in the first to seventh embodiments. Thus, it is obvious that even when the respective configurations are arbitrarily combined, the same operational effects described in the respective embodiments can be obtained.
[0067]
It is also possible to provide part or all of the processing in each of the above-described embodiments as a program (program) including an ordered sequence of instructions suitable for processing by a computer. It is also possible to provide the program as a computer-readable recording medium on which the program is recorded for installation, execution, and distribution of the program.
[0068]
【The invention's effect】
In the present invention, in a speech synthesizer that reads out a hypertext, a link destination connected in a network by a hyperlink Reading text extracted and generated from hypertext Thus, it is possible to provide a speech synthesizer that can read out the text in a state where the text of the text is displayed.
[Brief description of the drawings]
FIG. 1 is a functional block diagram of a speech synthesizer according to Embodiment 1 of the present invention.
FIG. 2 is a conceptual diagram of hypertext to be input to the speech synthesizer according to Embodiment 1 of the present invention and text information generated from the hypertext.
FIG. 3 is a functional block diagram of a speech synthesizer according to Embodiment 2 of the present invention.
FIG. 4 is a flowchart showing processing of a text management unit of the speech synthesizer according to Embodiment 2 of the present invention.
FIG. 5 is a functional block diagram of a speech synthesizer according to Embodiment 3 of the present invention.
FIG. 6 is a functional block diagram of a speech synthesizer according to Embodiment 4 of the present invention.
FIG. 7 is a functional block diagram of a speech synthesizer according to Embodiment 5 of the present invention.
FIG. 8 is a conceptual diagram showing the operation of the text-to-speech extraction unit of the speech synthesizer according to Embodiment 5 of the present invention.
FIG. 9 is a functional block diagram of a speech synthesizer according to Embodiment 6 of the present invention.
FIG. 10 is a functional block diagram of a speech synthesizer according to Embodiment 6 of the present invention.
FIG. 11 is a flowchart showing processing of a text-to-speech generation unit of the speech synthesizer according to Embodiment 7 of the present invention.
FIG. 12 is a flowchart showing processing of a text-to-speech generation unit of the speech synthesizer according to Embodiment 7 of the present invention.
[Explanation of symbols]
100 Speech synthesizer
101 Hypertext interpretation part
102 Reading text generator
103 Language processor
104 Speech synthesis unit
105 Audio output unit
106 Text-to-speech management department
107 Text summary section
108 Text Translation Department
109 Text-to-speech extraction unit
110 Hypertext Management Department
111 Input section
112 Display section
113 Link destination display

Claims (7)

Display means for displaying information indicated by hypertext;
Link destination information that analyzes the attribute information indicating the link destination in the input hypertext and extracts link destination information indicating the link destination that represents the hypertext connection destination on the Internet from the input hypertext according to the analysis result Extraction means;
After a hypertext on the Internet corresponding to the link information acquired from the link destination on the Internet indicated by the link destination information, along with contained in the hypertext, based on the markup information indicating that it is a text-to-speech , A linked text-to-speech extracting means for extracting the text to be read out from the hypertext,
Link-to-speech text generation means for generating link-to-speech text based on the extracted link-to-speech hypertext,
A speech synthesizer for synthesizing the read-out text of the link destination extracted and generated from the hypertext acquired from the Internet, while maintaining the display processing by the display means for the input hypertext;
A speech synthesizer characterized by comprising: Text-to-speech text generating means for generating text-to-speech text based on the input hypertext,
A switched speech synthesizer that synthesizes speech by switching between the generated text-to-speech text and the generated text-to-link text to be linked;
The speech synthesizer according to claim 1, further comprising: It further has an input means for receiving an input of a desired link destination,
When an input of a link destination is received by the input means, the read-out text of the link destination is voice-synthesized by the switching voice synthesizing means, while the text-to-speech text is displayed by the display means. The speech synthesizer according to claim 2.   The input of the link destination by the input means is performed by moving the pointer including the cursor to display information including a character or an image representing the link destination in the reading text of the displayed text. The speech synthesizer according to claim 3.   2. The apparatus according to claim 1, further comprising selection means for selecting whether or not to display information corresponding to the text to be read at the link destination when speech synthesis of the text to be read at the link destination is requested. The speech synthesizer according to claim 2 or claim 3. Displaying information indicated by input hypertext;
Analyzing attribute information indicating a link destination in the input hypertext, and extracting link destination information indicating a link destination representing a hypertext connection destination on the Internet from the input hypertext according to the analysis result;
After acquiring the hypertext on the Internet corresponding to the link destination information from the link destination on the Internet indicated by the link destination information, the hypertext is included in the hypertext and based on markup information indicating that it is a read-out text Extracting the hypertext to be read from the hypertext,
Generating link-to-speech text based on the extracted link-to-speech hypertext,
Maintaining the display processing for the input hypertext while synthesizing the read-out text of the link destination extracted and generated from the hypertext obtained from the Internet;
A speech synthesis method characterized by comprising:   A computer-readable recording medium recording a voice synthesis program for causing a computer to execute each step of the voice synthesis method according to claim 6.

Images