JPH05197384A - Voice reading out device - Google Patents

Abstract

PURPOSE:To efficiently repeat when repetition is instructed. CONSTITUTION:When the reading-out of a message is started and a message (the study of a new product concept...) is vocalized (42), and a user issues (43) an instruction of repetition at a time when the middle way (...following) of the message is listened, a numerical character in the message is searched as an important point, and a part including the numerical character is extracted as the important point, and the important part (1991... conference room) from the first important part to an repetition-instructed part is set as a repetition range starting from the beginning of the message, and only the important part in such repetition range i.e., a date and a place in this example is repeated (44) as a message for repetition, and after that, the reading of the message is performed (45) from a position where the reading is performed when the repetition is instructed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice reading device capable of reading a message represented by a character code by voice.

[0002]

2. Description of the Related Art A conventional voice reading device also has a function of repeatedly reading a message. For the range of reciting, a method of reciting from the beginning of the message being read, a method of reciting from the beginning of the sentence being read,
Etc.

[0003]

DISCLOSURE OF THE INVENTION In the above-mentioned conventional voice reading apparatus, when a recitation is instructed, a message is read back from the beginning of the message being read, even if the immediately preceding portion cannot be heard. Since I read it from the beginning, I have to listen to the less important parts, which is a waste of time and annoying.

On the other hand, in a device that repeats a sentence just before when a repeat is instructed, the sentence just before may not contain an important part to be heard again. However, there is a drawback that operations such as repeating key operations are necessary. Also, in both cases, the expression of the message to be repeated is the same as in the case of the first reading aloud, so that already heard parts and less important parts are repeatedly uttered, and the listening time is wasted and bothersome. There was a drawback that you could feel it.

An object of the present invention is to provide a voice read-aloud device which can repeat a voice so as to efficiently convey it to the user when the user gives an instruction to repeat the reading.

[0006]

According to the present invention, the range of recitation is determined based on the important part extraction unit for extracting the important part of the message and the important points extracted by the important part extraction unit. A recitation range determination unit is added, and in the invention of claim 2, a recitation message creation unit that creates a message emphasizing the essential points of the replay range is further centered on important points included in the replay range determined by the replay range determination unit. It is provided.

[0007]

When the reading control unit instructs the voice synthesis unit to read the voice, the voice synthesis unit converts the character code string of the message stored in the message storage unit into voice, as in the conventional case. .. When the signal receiving unit receives the signal for instructing recitation while the message is being read aloud, the important portion extracting unit first extracts the important points of the message. Next, based on the extracted important points, the recitation range determination unit determines the range to be repeated, and instructs the speech synthesis unit to read the sentence in the determined recitation range.

Alternatively, the repeat message creating unit creates a message that emphasizes the main points of the repeat range, centering on the important points included in the determined repeat range.
The reading control unit instructs the speech synthesis unit to read the emphasized recitation message. In any case, when the repeat is finished, the reading control unit
The speech synthesizer is instructed to resume reading the part of the message that has not been read yet. From the above, if the user cannot hear the message and instructed to repeat,
Since the message is repeated around the important points you want to hear, it is possible to convey the message efficiently.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1A shows the configuration of a system for reading aloud an electronic mail by using the voice reading device according to the present invention. The message from the electronic mail system 11 is supplied to the voice reading device 12 according to the present invention, is subjected to voice synthesis, and is output from the voice output device 13. A signal input device 14 is also connected to the voice reading device 12. The voice output device 13 may be considered as a telephone receiver of the telephone, and the signal input device 14 may be considered as a push button of the telephone.

FIG. 1B shows the configuration of the voice reading device 12 according to the present invention. A message storage unit 21 for storing a message to be read aloud as in the conventional device, a voice synthesizing unit 22 for converting a character code string into a voice, a signal receiving unit 23 for receiving a signal instructing repeat reading, and a message aloud. When the signal receiving unit 23 receives a signal instructing recitation during the course of replaying, the signal receiving unit 23 is provided with a rereading control unit 24 for instructing the rereading of a portion that is not yet read out when the replaying is completed. There is.

In the present invention, an important part extracting section 25 for extracting an important point of a message, a recitation range determining section 26 for deciding a range of recitation based on the important points extracted by the important part extracting section 25, and A repeat message creating unit 27 is provided that creates a repeat message centering on the important points included in the determined repeat range and emphasizes the main points of the repeat range.

FIG. 2A shows an example of the message to be read out, and FIG. 3 shows the message of FIG.
An example of the interaction between the voice reading device 12 according to the present invention and the user when the user gives an instruction to listen again will be shown. Hereinafter, based on the above example, an operation example of the voice reading device 12 according to the present invention will be described.

When an e-mail is input to the voice reading device 12 as an example of a message, the message is stored in the message storage unit 21, and then the reading control unit 24 reads "read the message" in the voice synthesis unit 22. More output (41). Next, the message storage 2
The voice synthesis unit 22 is instructed to convert the character code string of the message stored in 1 into voice. I heard the message aloud halfway ("... future plans") (4
2) After that, it is assumed that the user gives an instruction to listen again with the signal input device 14 (43). Then, the signal receiving unit 23 receives the signal, and the important portion extracting unit 25 searches for a portion that becomes an important point in the messages read up to that point. In the message example of FIG. 2A, based on the idea that numbers are important, the part including numbers (range separated by punctuation marks, spaces, etc.) is the important part 5.
FIG. 2B shows an example of the result of extraction as 0, 51 and 52.

In addition to this, as a method of extracting this important part, the user on the sending side adds a special character code to an important part when the message is created, and the important part extraction part 25 uses this special character. A method of extracting the important part by searching the code, performing morphological analysis of the message, inferring the importance from the part-of-speech classification, noun attributes, etc. (for example, judging that numbers, place names, etc. are important), There are methods such as cutting out.

Next, the recitation range determination unit 26 determines the recitation range based on the extracted important part. In this example, the position of the important point that appears first from the beginning of the message is the start point of the recitation range, and the position read aloud when the signal instructing the recitation is the end point of the recitation range. Here is an example. In addition to this, as a method for determining the recitation range, there is a method in which the important part immediately before the position read aloud when the signal instructing the recitation is received is used as the starting point. When the recitation range determination unit 26 determines the recitation range, the reading control unit 24 instructs the speech synthesis unit 22 to read the sentence in the determined range.

Alternatively, by creating the repeat message 27,
A message for recitation that emphasizes the main points is created centering on the important part of the message included in the recitation range determined by the recitation range determination unit 26. In this embodiment, the important part 5 is simply
This is an example in which a message emphasizing the main points is created by listing only 0, 51, and 52. Besides this,
There is a method of creating a message emphasizing an important point by changing the reading control information from the first reading, such as slowing down the speed of reading the important point or increasing the volume. When the repeat message is created by the repeat message creation 27, the reading control unit 24 sends the read message to the voice synthesizing unit 22 to instruct the read-out message.

In any case, the voice output device 13 generates the message converted into voice by the voice synthesizer 22. As a result, in this example, a repeat message indicating the date and time and the place is uttered (44). Next, the reading control unit 24 starts reading from the position that was being read when the signal instructing the recitation is received (45). When listening to the reading of this message halfway ("Please ask ..."), it is assumed that the user gives an instruction from the signal input device 14 to listen again (40). Then, similarly to the above, the repeat message is output as voice. As a result, in this example, the telephone number of the secretariat is uttered as a repeat message (47). When the recitation is completed, there is no unread part, so that "the message is over" is uttered (48), and the message reading is completed.

[0018]

As described above, according to the present invention, when the user cannot hear the message and instructs the repeater to repeat the message, the message is repeated centering on the important point that the user wants to hear. It becomes possible.

[Brief description of drawings]

FIG. 1A is a block diagram showing a configuration of a system for reading an e-mail by using a voice reading device according to the present invention, and B is a block diagram showing a configuration of a voice reading device according to the present invention.

2A is a diagram showing an example of a message to be read out, and FIG. 2B is a diagram showing an example of an important part to be extracted.

FIG. 3 is a diagram showing the interaction between the voice reading device according to the present invention and the user when the user gives an instruction to listen again while reading the message of FIG. 2A.

Claims (2)

[Claims] 1. In a voice reading device comprising a voice synthesizing unit for converting a character code string into voice, and reading out a message represented by the character code string by voice, a message storing unit for storing a message to be read, and repeated reading ( (Repeat) A signal receiving unit that receives a signal instructing (Repeat), an important part extraction unit that extracts the important points of the message, and a recitation that determines the range to be repeated based on the important points extracted by the important part extraction unit. Range determination unit, in the middle of reading a message, the signal receiving unit, when receiving a signal for instructing recitation, instruct the speech synthesis unit to read the sentence in the range determined by the recitation range determination unit, When the recitation is completed, a reading control unit for instructing the voice synthesis unit to read a portion not yet read is provided. A voice reading device. 2. The voice reading device according to claim 1, wherein a message is created for emphasizing and transmitting the main points of the recitation range, centering on important points included in the recitation range determined by the recitation range determination mechanism. A read-back message creation unit is added, and the reading control unit instructs the voice synthesis unit to read the message created by the read-back message creation unit at the time of reading back.