JP4514144B2 - Voice reading apparatus and program - Google Patents

Description

The present invention relates to a speech reading apparatus and a program for determining whether or not each part in a page has already been read out by a speech reading apparatus, that is, whether or not the part has been read, and reading out only the read part or only the unread part. About.

  With the spread of the Internet, it is possible to obtain various information other than broadcasting such as television and radio.

  One of the methods of using the Internet is browsing a website. The feature of the website is that various photos and texts can be arranged in two dimensions as provided by the web content creator and provided to the user.

  On the other hand, it is known that a voice browser has been developed, and the contents of a web page can be converted into voice and provided to a user (for example, see Patent Document 1). Not only is it convenient for visually impaired users, but it is also convenient for a healthy person to be able to listen to the contents of the converted web page by voice in situations where the display cannot be viewed.

As described above, the voice browser technology is convenient not only for visually impaired users but also for all users.
JP 11-249867 A

  When using a voice reading device that reads out a content by voice, in order to know the content, it is necessary to listen to the read voice one by one.

  For example, when considering a voice browser, when listening to the voice of a voice browser reading out the web content, it must be relied upon whether or not the part has been heard before. Therefore, if the storage is inaccurate, the part that has been read once (read part) is read again, and there is a problem that it takes time to reach the part that has not been read yet (unread part).

  On the other hand, there is also a request to refer again to a part that has been read in a page that has been read once. In this case, since all the parts are read out without distinguishing between the unread part and the read part, there is a problem that it takes time to reach the read part.

  An object of the present invention is to provide a speech reading apparatus and a program that do not take a long time to reach a portion that has not yet been read (unread portion).

It is another object of the present invention to provide a speech reading apparatus and a program that can quickly reach an already read portion when it is desired to refer to a read portion of a page that has been read once. It is.

According to the present invention, in a speech reading apparatus that converts a content into a sound and reads out the content, a reading attribute indicating a distinction between an unread part and a read part is held for each reading unit divided by a predetermined break for the content. Data holding means, and if the held reading attribute indicates a read part, a reading skip means for skipping the reading target part indicated by the corresponding reading unit, and the held reading attribute indicates an unread part. If so, it has reading means for reading the reading target part indicated by the corresponding reading unit, and setting means for setting the reading attribute of the reading unit of the reading target part read by the reading means to the read part. This is a featured speech reading apparatus.

  According to the first, second, fourth, fifth, sixth and seventh aspects of the present invention, when inputting a URL or jumping to a link destination, information written on a new web page is received by voice. The user can skip the part that has been heard once, which is convenient without having to hear it twice. When reading a web page from the beginning, it takes time to reach the part that has not yet been read (unread part). There is an effect that it does not take.

According to the first, third, fourth, fifth, sixth, and eighth aspects of the present invention, when it is desired to refer to a part that has been read once on a page that has been read once, the user reaches the already-read part. Has the effect of fast.

  The best mode for carrying out the invention is the following examples.

  FIG. 1 is a block diagram showing a speech reading apparatus 10 that is Embodiment 1 of the present invention.

  The speech reading apparatus 10 is specifically a web browser and is connected to the web server 20 via the network NW. Note that the voice reading device 10 may be configured by hardware.

  The voice reading device 10 communicates with the web server 20 via the network NW, receives and displays web content, the information input unit 11, the control unit 12, the transmission / reception unit 13, the information analysis unit 14, and the information The storage unit 15, the speech synthesis unit 16, and the information output unit 17 are included.

  The information input unit 11 includes an input device such as a keyboard and a mouse.

  The control unit 12, the transmission / reception unit 13, the information analysis unit 14, the information storage unit 15, and the voice synthesis unit 16 are configured by a PC (personal computer).

  The information output unit 17 includes a monitor and a speaker.

  Example 1 is an example in which unread portions are skipped.

  In order to do this, it is necessary to distinguish the unread portion and the read portion and store the data of the reading target portion. When storing the already read part, all the sentences to be read out are stored. The unit for determining whether it is a read part or an unread part is a character, a section delimited by punctuation marks, a sentence, a paragraph, an entire web page, and the like.

  In general sentences, there are units that can be clearly recognized by anyone, such as punctuation marks and paragraphs. Therefore, it is only necessary to determine whether each of the units is an already read part or an unread part. Further, if there is a delimiter or the like peculiar to a predetermined web page, the delimiter may be used as the determination unit. In the following description, an unread part and an already read part are discriminated in units of sentences.

  Next, the operation of the first embodiment will be described.

  FIG. 2 is a flowchart illustrating the operation in the first embodiment.

  In the operation shown in FIG. 2, the reading unit is a sentence. However, when the sentence is not a reading unit but a punctuation mark is used as a reading unit, “sentence” in the explanation shown in FIG. It should be read as “punctuation”.

  In S <b> 1, a URL is input via the information input unit 11. In S2, the transmission / reception unit 13 communicates with the web server 20 of the input URL via the network NW via the control unit 12, and the transmission / reception unit 13 receives the web content.

  In S3, the URL of the received web content and the web content. It is stored in the information storage unit 15. Subsequently, in S <b> 4, the information analysis unit 14 analyzes the stored web content, extracts the text to be read out, and stores the extraction result in the information storage unit 15.

  A sentence to be read out is stored in a sentence unit which is a discrimination unit.

  FIG. 3 is a diagram illustrating an example of data of a reading target portion stored in the information storage unit 15.

  Here, if there is a punctuation mark or a line break, or if it is delimited by a specific HTML tag, it is determined that there is a break in the sentence, and as shown in FIG. 3, an element number that identifies each sentence for each sentence Is stored in the information storage unit 15. A delimiter HTML tag is stored in advance in the information storage unit 15 and used for analysis of web content. Here, the HTML tag used as a delimiter is not an HTML tag used in a sentence, such as <a>, but a tag used to delimit a sentence, such as <H1>, etc. Can be separated.

  The data of the reading target part shown in FIG. 3 is the data of the reading target part described in the web page visited for the first time, and all the data of the reading target part is before the reading, and the unread part and the read part “0” indicating an unread portion is assigned to the “reading attribute” column indicating the distinction.

  In S5, the reading attribute of the first sentence “my food diary” of the reading target sentences is read. As shown in FIG. 3, “0” indicating an unread portion is given. In S6, this reading attribute is evaluated. If it is “1”, it is a read part. Therefore, in S7, it is determined whether or not it is the last sentence. If it is the last sentence, the process ends. Otherwise, in S8, the read-out attribute of the next sentence is read out, and the process returns to S6.

  If the reading attribute is determined to be “0” (unread portion) as a result of determining the reading attribute in S6, the text to be read is read in S9. Specifically, the content of the data of the reading target portion shown in FIG. Read the data described in the column. In the example shown in FIG. 3, since the reading attribute of the first sentence is “0”, the process proceeds from S6 to S9 without going through S7 and S8, and the first sentence “My food diary” to be read is read out. .

  In S10, the speech synthesis unit 16 synthesizes the speech corresponding to the read sentence, and in S11, the information output unit 17 outputs the synthesized speech. In S12, when setting the unread part or the read part for the text to be read out, “1” indicating the read part is overwritten. In this way, it is possible to determine that the determination unit is the read portion when reading of the text that is the unread / read determination unit is started.

  Now, in a voice browser, it is common to jump to a link destination when the Enter key is pressed while reading a link portion of web content. In S13, when a command to jump to the link destination is input via the information input unit 11, the process returns to S1, and the control unit 12 interprets and acquires the URL of the link destination embedded in the link portion. Then, the process after S2 is started about the jump destination web page.

  If a command to jump to the link destination is not input in S13, it is determined in S14 whether or not a command to stop reading is input. A reading stop command is issued when an operation that should stop reading is performed via the information input unit 11, that is, in the case of a PC browser, through a keyboard operation or a mouse operation, such as a reading stop or a pause operation. It is interpreted that a reading stop command is input.

  FIG. 4 is a diagram illustrating an example of the data of the reading target portion stored in the information storage unit 15 and is a diagram illustrating that the data of the reading target portion is partially a read portion.

  For example, it is assumed that a reading stop command is input after reading the reading target portion of element number 6 shown in FIG. At this time, as shown in FIG. 4, the reading attributes of the element numbers 1 to 6 are “1” indicating that the part is already read, and the others remain “0”.

  As described above, as shown in FIG. xxx. yyy. In the contents of jp, information indicating that the data of the part to be read out with the element numbers 1 to 6 is a read part and the data of the part to be read out with the element numbers 7 to 10 is an unread part. It remains accumulated. Each time several web pages are browsed, data in the form as shown in FIG. 4 is accumulated and increased. When the already-read portion exists in a skipped manner, for example, as shown in FIG. 15, the reading-out attribute is stored.

  When an instruction to stop reading is input (S14), reading is terminated halfway. If an instruction to stop reading is not input in S14, it is determined in S15 whether it is the last sentence. If it is not the last, it returns to S8 and reads the reading-out attribute of the next sentence. If it is the last sentence in S15, the process ends.

  Here, the case where the unread / read discrimination unit is a sentence has been described. However, as described above, the unread / read discrimination unit is not limited to a sentence, but is a section, sentence, paragraph, or web page separated by characters and punctuation marks. Arbitrary units, such as the whole, may be sufficient.

  By the way, blog sites are popular in recent years. Blog sites are often used for diaries, and the content increases each day. As described above, content may be added even if the URL is unchanged. In this case, if unread / read is determined based on only the element number of the content, the determination of unread / read may be erroneous. is there. A blog web page is composed of a plurality of articles, each of which is a diary for a different day. New articles are often displayed at the top, so when comparing element numbers, the determination of unread and read is wrong. For such sites, it is effective to accumulate all the texts already read. Furthermore, when it is determined whether or not the text of the data to be read (read data) stored in the information storage unit 15 is present in the text to be read, the text from the top of the text to be read It may be judged by sequentially comparing and pattern matching.

  FIG. 5 is an explanatory diagram for checking whether or not the already-read portion is included in the reading target in the first embodiment.

  For example, when checking whether or not the read portion “big hit!” Is a reading target, first, as shown in FIG. 5, the characters are compared character by character from the beginning of the sentence that is the data of the reading target portion.

  FIG. 6 is a diagram illustrating an example in which the read portion “big hit!” Is shifted by one character from the head and compared with the data of the reading target portion in the first embodiment.

  In the example shown in FIG. 5, since they do not match, as shown in FIG. 6, the characters are shifted from the beginning and compared. As shown in FIG. 6, after comparing the read portion “big hit!” By shifting one character, the read portion “big hit!” Is compared by shifting two characters, and the read portion “big hit!” Is shifted by three characters. Compare, ……, to the end, compare the read part “big hit!”, And if it matches in the process, it can be determined that the read part “big hit!” Is the read part, If they do not match up to the end, it can be determined that the read portion “big hit!” Does not exist in the data of the portion to be read out. As a result, it is possible to create read-out attribute data as shown in FIGS.

  Also, since the blog site is created according to a predetermined format, the structure of the page is always constant, which part of the page is the title of the article, which part is the article, and which part is the comment You can find out if there is an HTML tag analysis. Therefore, if an unread / read determination unit is specified as an article title and an article, it is possible to distinguish an article that has been read from an article that has not been read. For example, for a person who writes a blog as a diary, the article title and the article mean a diary for a certain day. Therefore, if the article title and the article are unread / read determination units, the day of reading and the day of not reading can be found in the diary. Thus, based on the configuration of the web page, a specific part is cut out and used as an unread / read determination unit, which can be used more conveniently.

  In the first embodiment, the read part is stored in the information storage unit 15, but conversely, the unread part is stored in the information storage unit 15 and the read part is deleted from the storage target. May be.

  FIG. 7 is a diagram illustrating an example in which the start number and the end number of the element number of the read part are stored in the information storage unit 15 in the first embodiment.

  FIG. 8 is a diagram illustrating an example in which two sets of the start number and the end number of the element number of the read part are stored in the information storage unit 15 in the first embodiment.

  In the first embodiment, in S12, as shown in FIG. 4, the reading attributes are stored for each element number of the reading target portion, but instead of doing this, as shown in FIG. You may make it accumulate | store in the information storage part 15 the start number and end number of the element number of a reading part. Further, as shown in FIG. 15, when a plurality of already-read portions are scattered, a row may be added and accumulated as shown in FIG.

  Further, the first embodiment is an example in which the already read portion is skipped and only the unread portion is read out. On the contrary, the unread portion may be skipped and only the already read portion may be read out. Is feasible.

  In FIG. 9, in the first embodiment, among the data of the reading target portion, data of the reading target portion to be skipped is set, and the set data of the reading target portion to be skipped is stored in the information storage unit 15. It is a figure which shows an example.

  As shown in FIG. 9, the data of the part to be read out to be skipped is stored in the information storage unit 15 in advance. In the example shown in FIG. 9, “1” described in the column of the skipped part is skipped. "0" means not skipping.

  The example shown in FIG. 9 indicates that the read portion is skipped and the unread portion is not skipped (normally read out). In the first embodiment, an operation according to the setting shown in FIG. 9 is executed.

  The above setting is performed, and skipping may be performed according to the flowchart shown in FIG. If it is determined in S6 that the part has been read, the process proceeds to the skipping operation (S7, S8), and the skipping may be executed according to the setting data shown in FIG. As shown in FIG. 9, when the already read portion is skipped, the operation is the same as that shown in FIG.

  FIG. 16 is a diagram illustrating an example of skip setting data stored in the information storage unit 15 in the first embodiment.

In order to skip the unread portion, when the unread portion is determined in S6 shown in FIG. 2, since the skipped portion is “1” as shown in FIG. 16, the process proceeds to S7 in the flowchart shown in FIG. If the already read portion is determined in S6 shown in FIG. 2, the skipped portion is “0” as shown in FIG. 16, so that the flowchart proceeds to S9 in the flowchart shown in FIG. Can be changed.

  The second embodiment is an embodiment in which, in the first embodiment, instead of not reading the already read portion, the unread portion is not read.

  The first embodiment is an embodiment in which information indicating a read portion is accumulated at the time when reading starts in a unit for determining read / unread, but the second embodiment determines read / unread. This is an embodiment in which information indicating that it is a read part is accumulated when reading is completed in units.

  In the first embodiment, since the discrimination unit is already read when it starts to be read out, it is unclear whether the discrimination unit has been read up to the last part, especially when the discrimination unit is long, such as a paragraph. In the case of “read”, it means “I have started reading”. In some content, just listening to the beginning of the content reveals that you are not interested in it, and you may read other parts. In this case, Example 1 is more effective.

  On the other hand, in the second embodiment, since the reading of the discrimination unit is finished, it is already read. Therefore, even when the discrimination unit is long, such as a paragraph, it is read after being read to the last part of the discrimination unit. That is, the already read in this case means that “the discrimination unit has been read”, which is evidence that the discrimination unit has been read to the end, and the second embodiment is convenient depending on how it is used.

  The configuration of the second embodiment is basically the same as that of the first embodiment.

  FIG. 10 is a flowchart showing the operation of the second embodiment of the present invention.

  Since the process of S21-S30 is the same as the process of S1-10 in Example 1, description is abbreviate | omitted.

  In S31, the information output unit 17 outputs the synthesized voice. In S32, when a command to jump to the link destination is input via the information input unit 11, the process returns to S21, and the control unit 12 interprets and acquires the URL of the link destination embedded in the link portion. Then, the process of S22-S31 is performed about the web page of a jump destination.

  If a command to jump to the link destination is not input in S32, it is determined in S33 whether or not a command to stop reading is input. If a command to stop reading is input, reading is terminated halfway. If it is determined in S33 that a reading stop command is not input, it is determined in S34 whether the output of the synthesized speech has been completed. If the reading is not completed, the process returns to S31 to continue the output. When the reading is completed (S34), the setting of the reading voice of the sentence is overwritten with "1" meaning that it is the already read part in S35. In this manner, when the reading of the sentence that is the unread / read determination unit is completed, it can be determined that the determination unit is the read portion.

  Next, it is determined whether it is the last sentence by S36. If it is not the last, it returns to S28 and reads the reading-out attribute of the next sentence. If it is the last sentence in S36, the process ends.

  In the second embodiment, the unit of unread / read determination is a sentence. However, it is not limited to a sentence, and an arbitrary unit such as a character, a section delimited by punctuation, a sentence, a paragraph, or an entire web page is determined as unread / read. It may be used as a unit.

Further, the second embodiment is an embodiment in which the read portion is skipped and only the unread portion is read out. Conversely, in the second embodiment, the unread portion may be skipped and only the already read portion may be read out. .

  The third embodiment of the present invention is an embodiment for notifying that when the already read portion is skipped.

  The configuration of the third embodiment is basically the same as the configuration of the first embodiment.

  The third embodiment is an embodiment in which it is determined that the read portion is read when reading is started, as in the first embodiment. However, the read portion is read when the reading is finished as in the second embodiment. You may make it judge.

  Next, the operation of the third embodiment will be described.

  FIG. 11 is a flowchart showing the operation of the third embodiment of the present invention.

  Since the process of S1-S6 is the same as the process of S1-S6 in Example 1, description is abbreviate | omitted.

  FIG. 12 is a diagram illustrating an example of already-read portion presence / absence data indicating whether or not there is a read portion among the data of the read target portion stored in the information storage unit 15 in the third embodiment. It is an example which shows that the reading part does not exist.

  FIG. 13 is a diagram illustrating an example of read portion presence / absence data indicating whether or not there is a read portion among the data of the read target portion stored in the information storage unit 15 in the third embodiment. It is an example which shows that a reading part exists.

  That is, if there is no read portion, the read portion presence / absence data (read skipping presence / absence data) is “0” as shown in FIG. 12, and if there is a read portion, the read portion is read as shown in FIG. The partial presence / absence data (read skipping presence / absence data) is “1”.

  In S6, the read-out attribute is determined, and if the read-out attribute is already read, the presence of a skipped portion is stored in the information storage unit 16 in S41. That is, if it is a read portion, the read portion presence / absence data (read skipping presence / absence data) “1” is overwritten on the data shown in FIG.

  If it is determined in S42 that the sentence is the last sentence, the process ends. If not, the reading attribute of the next sentence is read in S43, and the process returns to S6.

  On the other hand, if it is determined in S6 that the reading attribute is unread, it is determined whether or not there is a skipped portion in the past. If there is a skipped portion in the past, the fact that the read portion is skipped in S45. The voice synthesizer 16 synthesizes the sound file of the guidance sentence (notification sentence), and outputs the sound file from the information output part 17 in S46. That is, as shown in FIG. 13, if the already-read portion presence / absence data (read skipping presence / absence data) is “1”, the read portion has been skipped, so that the fact that the read portion has been skipped is communicated by voice ( S45, S46).

  If there is no skipped portion in the past in S44, the text is read as it is (S50).

  FIG. 14 is a diagram illustrating an example of a guidance sentence (guidance voice) indicating that there is a read portion in the third embodiment.

  In S50, the read text (data described in the data contents column of the reading target portion in FIG. 3) and the guidance text indicating that there is a read portion are displayed in S51. Is synthesized. The guidance text indicating that the read part is present is stored in advance in the information storage unit 15 in the form shown in FIG.

  In S51, a voice message is synthesized by combining the guidance sentence shown in FIG. 14 "The read portion has been skipped. The unread portion is read from here" and the text read in S50.

  Alternatively, the guidance text indicating that there is a read portion is stored in advance in the information storage unit 15 as a sound file, and is output together with the guidance text in S51 without being synthesized in S51. It may be.

  The subsequent processes of S54 to S56 are the same as the processes of S13 to S15 in the first embodiment, and thus the description thereof is omitted.

  As described above, since the fact that the read portion has been skipped is transmitted by voice, the user listening to the read-out voice can know that the read portion has been skipped.

Further, the third embodiment is an embodiment in which the already read portion is skipped and only the unread portion is read out. Conversely, in the third embodiment, the unread portion may be skipped and only the already read portion may be read out. .

  The fourth embodiment of the present invention is an embodiment in which an already read portion is skipped and a visual indication that the unread portion has begun to be read is displayed.

  Even in a conventional voice browser, the screen is displayed like a normal browser in order to show where to read. Which part is being read out is indicated by highlighting. The highlight display in this conventional example is a display at the moment of reading, and the previously read portion cannot be known. Even voice browsers can be used by healthy people who can see if their voice quality and ease of use are good. In other words, even if you can see, you may not be able to see the screen while you are doing something else. In this case, a voice browser is useful. However, considering the use by such a healthy person, it is also important to visually convey the information of the read part and the unread part.

  For this purpose, as in the case of following the intra-page link, the already read portion may be scrolled and displayed from the top of the unread portion based on the data of the reading target portion shown in FIG. That is, the top of the unread portion is displayed at the top of the browser. By doing so, it is visually understood that the already read portion is skipped.

  FIG. 15 is a diagram illustrating an example of data of the reading target portion stored in the information storage unit 15 in the fourth embodiment.

  In the case of the data shown in FIG. 15, among the plurality of unread portions, the top unread portion may be displayed at the top of the browser of the fourth embodiment. Alternatively, a portion that is not read out may not be displayed. If not displayed, “Reading omitted” or the like may be displayed in a portion that is not read out.

  Further, the fourth embodiment is an embodiment in which the read portion is skipped and only the unread portion is displayed. Conversely, in the fourth embodiment, the unread portion is skipped and only the read portion is displayed. Also good.

  Each of the above embodiments is a web browser that receives content from the network NW and reads it aloud, but instead of doing so, a memory card in which the content is stored, a CD drive that drives a medium such as a CD or DVD, etc. The transmission / reception unit 13 may receive the content and read out the received content by voice.

  Further, the content may be downloaded via the network NW, stored in the information storage unit 15, and the stored content may be read out by voice. In this way, the content can be read aloud even when the network NW is not connected, which is convenient. Furthermore, the reading target need not be limited to web content.

  Further, the voice reading device 10 may be connected to a plurality of terminal devices via a network, that is, the function of the voice reading device 10 may be taken into a server. In this way, you can place the read and unread information on the network, so even if you use different voice reading devices, you can operate them as if you were using the same voice reading device. It is.

  The above embodiment can be grasped as a program invention. That is, the above embodiment is a read / unread determination procedure for determining whether or not a part to be read out in the content has been read out in the past, Of the parts to be read out, the part that cannot be read out is set as an unread part, or the part that is read out is set as a read part, and the setting procedure for storing in the storage device and the above setting A read-out procedure for reading out the read-out target portion corresponding to the read-out unread portion or the read-out portion, and a read-out read-out target portion storing procedure for storing the read-out read-out target portion when the read-out target portion is read aloud. It is an example of the program which makes a computer run.

  Further, in the above-described embodiment, in the speech reading program that reads out the content by voice conversion, the read / unread determination procedure for determining whether or not the portion to be read out in the content has been read out in the past, Among the parts to be read out, it is set that the part that cannot be read out is an unread part, a setting procedure for storing in the storage device, and a reading out procedure for reading out the part to be read out corresponding only to the set unread part, It is an example of a program that causes a computer to execute a stored read target portion storing procedure for storing a read target portion to be read when the read target portion is read aloud.

  Further, in the above-described embodiment, in the speech reading program that converts the content into speech and reads it out, the read / unread determination procedure for determining whether or not the portion to be read out in the content has been read out in the past; Of the reading target parts, set that the read-out part is a read part and store it in the storage device, and read out the reading target part corresponding to only the read part that is set above And an example of a program that causes a computer to execute a stored read target portion storing procedure for storing a read target portion when the read target portion is read aloud.

The program may be stored in a storage medium such as a CD, MD, HD, or semiconductor memory.

1 is a block diagram showing a speech reading apparatus 10 that is Embodiment 1 of the present invention. FIG. 3 is a flowchart illustrating an operation in the first embodiment. 4 is a diagram illustrating an example of data of a reading target portion stored in an information storage unit 15. FIG. It is a figure which shows the example of the data of the reading-out object part accumulate | stored in the information storage part 15, and is a figure which shows that the data of the reading-out object part are a part already read. In Example 1, it is explanatory drawing in the case of investigating whether the already read part is contained in the reading-out object. In Example 1, it is a figure which shows the example in the case of shifting the already read part "big hit!" By 1 character from the head, and comparing with the data of the read-out object part. In Example 1, it is a figure which shows the example which has accumulate | stored in the information storage part 15 the start number and end number of the element number of a read part. In Example 1, it is a figure which shows the example which is accumulate | stored in the information storage part 15 2 sets of the start number of the element number of an already-read part, and an end number. FIG. 6 is a diagram illustrating an example in which data of a reading target portion to be skipped among data of a reading target portion is set in the first embodiment, and the set data of the reading target portion to be skipped is stored in the information storage unit 15. It is. It is a flowchart which shows operation | movement of Example 2 of this invention. It is a flowchart which shows operation | movement of Example 3 of this invention. In Example 3, it is a figure which shows the example of the read part presence / absence data which shows whether there exists a read part among the data of the reading object part accumulate | stored in the information storage part 15, and a read part exists It is an example showing that it is not. In Example 3, it is a figure which shows the example of the read part presence / absence data which shows whether there exists a read part among the data of the reading object part accumulate | stored in the information storage part 15, and a read part exists It is an example which shows having done. In Example 3, it is a figure which shows the example of the guidance sentence (guidance voice | voice) which shows that there exists a read part. In Example 4, it is a figure which shows the example of the data of the reading-out object part accumulate | stored in the information storage part 15. FIG. 6 is a diagram illustrating an example of skip setting data stored in an information storage unit 15 in Embodiment 1. FIG.

Explanation of symbols

10 ... Voice reading device (web browser),
11 ... Information input part,
12 ... control unit,
13: Transmitter / receiver,
14 ... Information analysis department,
15 ... Information storage unit,
16 ... speech synthesis unit,
17 ... Information output unit,
20: Web server.

Claims (4)

In a speech-to-speech device that reads out the content by voice conversion,
Data holding means for holding a reading attribute indicating a distinction between an unread portion and a read portion for each reading unit divided by a predetermined break for the content;
A reading skip means for skipping the reading target portion indicated by the corresponding reading unit if the held reading attribute indicates a read portion ;
If the held reading attribute indicates an unread portion, reading means for reading the reading target portion indicated by the corresponding reading unit ;
The reading attributes of reading units reading was reading target portion by the upper Symbol reading means and setting means for setting the read unit amount;
A speech-to-speech device comprising: In claim 1,
The above-mentioned reading skip means has a read skip guidance means for guiding at least one of voice and vision when skipping reading. In claim 1 or claim 2,
A speech reading apparatus , further comprising: a read-to-read target portion storing means for storing the character read by the reading means . A speech read-out program for causing a computer to function as each means constituting the speech read-out device according to claim 1.

Images