JPS6073724A - Document reading up device - Google Patents

Abstract

PURPOSE:To discriminate the voice for reading up warning messages from that for reading up a document by outputting different sound which is different from the voice for reading up a warning message produced during reading up of a document. CONSTITUTION:A statement read out from a document file 8 is fed to a rule synthesis section 12 through a statement analysis/read converting section 9. Then the rule synthesis section 12 outputs the statement through a loudspeaker 14 in, for example, the male voice of a predetermined tone quality. When the output request of a warning message, etc., occurs while the statement is read up, a main control device 3 controls a switching section 11 and feeds the message to another rule synthesis section 13 after reading out the message from a message file 5. Then the synthesis section 13 outputs the message through the loudspeaker 14 in, for example, the female voice of a predetermined tone quality. Therefore, the voice for reading up warning messages can be discriminated from that for reading up a statement.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides operation commands, warning messages, etc. generated during text reading by making them into different sounds so that they can be distinguished from the text reading sound. This invention relates to a document reading device with improved performance.

[Prior art and problems]

Japanese document processing device (Japanese word processor, etc.)
There is a document reading device that reads a document created by a document from a document file, converts it into audio, and outputs it. This document reading device issues a warning message if the reading cannot be understood because there is a homograph or allophone, such as ``devise'' or the dictionary does not have the reading. .

Furthermore, when reading out an instruction manual or an instruction manual, the content that is simply for the operator to listen and understand may sometimes contain content that prompts some kind of operation.

However, since conventional document reading devices can only output one tone color, there is a problem in that unless the operator is constantly paying attention, it is difficult for the operator to distinguish between the text being read and the operation command message. In particular, warning messages that occur while reading a sentence are difficult to distinguish because they suddenly appear in the middle of the sentence in the same tone (timbre), making it difficult to hear them. In addition, when proofreading a created document, it is necessary to distinguish between kanji, kana, kana, alphabetic characters, etc., but with conventional document reading devices,
No such distinction could be made.

[Purpose of the invention]

The present invention is based on the above consideration, and provides a document reading device that can distinguish between reading sentences and various messages such as operation command messages, and distinguishing between character types in reading sentences. The purpose is to

[Structure of the invention]

To this end, the document reading device of the present invention includes a dictionary that stores the pronunciations of words, clauses, phrases, and other words, and a text analyzer that analyzes sentences that match the reading target and converts the words, etc. into pronunciations by referring to the dictionary. & pronunciation converter, a message file storing the message, a message output control unit that reads and outputs the message from the message file, and the pronunciation converted by the text analysis & pronunciation conversion unit and the message output control unit. an audio output unit that outputs the message to be output as voice; and a main control that instructs the message output control unit to output the message when a message output request is issued from the text analysis and reading conversion unit, and also performs overall control. In the text reading device, the audio output unit has a function of outputting audio in a plurality of tones, and is configured to be able to output audio by changing the audio of the reading and the message. It is something.

[Embodiments of the invention]

Embodiments of the present invention will be described below with reference to the drawings.

The figure is a diagram showing the configuration of one embodiment of the present invention. In the figure, 1 is a display device, 2 is a keyboard, 3 is a main control unit, 4 is a message output control unit, 5 is a message file, 6 is a dictionary, 7 is a document input/output control unit, 8 is a document file, and 9 is a Sentence analysis/reading conversion unit, 10 is character type detection unit,
11 is a switching section, 12, 13, and 15 are rule combining sections, and 14 is a speaker. When creating a document, the main control section 3 controls the document input/output control section 7 to store the document created by manually inputting keys from the keyboard 2 while displaying the text on the display device 1 in the document file 8. , when reading out a document, controls the document input/output control unit 7 to read the document from the document file 8 and sends it to the text analysis/reading conversion unit 9;
When a message output notification is received from the text analysis/reading conversion section 9, the switching section 11 is set to the rule synthesis section having the tone to be outputted, and the message output control section 4 is controlled to read the message from the message file 5 and switch. The section 11 performs processing to send the rule to the selected rule synthesis section.

The sentence analysis/reading conversion section 9 analyzes the sentence read from the document file 8 by the document input/output control section 7 and converts it into a reading by referring to the dictionary 6. When a word consisting of a kanji code is found to have two readings such as ``devise'' as a result of referring to the dictionary 6, or when the reading as a word is incorrect, or in other specific cases, the main control unit 3 Notify of message output.

The dictionary 6 stores readings of words, clauses, phrases, and other words. The character type detection unit 10 detects the character type (kanji, kana, kana,
alphabetic characters, etc.), and based on the detection results, the switching unit 11
control. The rule synthesis units 12 and 13 are configured to synthesize voices of different tones, and for example, the rule synthesis unit 12 outputs a high-pitched female voice, and the rule synthesis unit 13 outputs a low-pitched male voice. .

Next, an example will be described in which the text reading voice is outputted from the rule synthesis unit 13 in a low male voice, and the voice such as an operation command is outputted from the rule synthesis unit 12 in a high voice female voice. The main control section 3 switches the switching section 11 to the rule synthesis section 13. The sentences read from document file 8 are processed by sentence analysis/
The pronunciation conversion unit 9 analyzes and converts the pronunciation, and the rule synthesis unit 13 outputs it through the speaker 14, where it is read out in a low male voice. During the reading process, the sentence analysis/pronunciation conversion unit 9 detects homographs and allophones as mentioned above.
The main control unit 3 is notified of this.

As a result, the main control section 3 changes the switching section 11 to the rule synthesis section 1.
2, the message output control section 4 is controlled to read the message from the message file 5, and the rule synthesis section 12 outputs a message in a female high-pitched voice through the speaker 14, for example, ``There are homographs and allophones''. Output. When the output from the message file 5 is completed, the main control section 3 sets the switching section 11 to the rule synthesis section 13. Similarly, when it is necessary to read aloud to the end of the recording medium and replace it with the next recording medium, when the text analysis/reading conversion section 9 notifies the main control section 3 of this, the main control section 3 controls the message output control section 4 to output a message such as "Please replace the recording medium with the next recording medium" using a female high-pitched voice from the rule synthesis section 12 through the speaker 14. In this way, the low voice of a man who was reading out a text suddenly changes to a high voice of a woman and the message is played from the speaker, so the operator can immediately distinguish that it is a message.

In addition, if the text reading voice is a male voice, the operation command voice is a low female voice, and the warning message is a high voice female voice, the 3S can be immediately distinguished. In this case, as shown by the broken line frame in the figure, the rule synthesis unit 1
5 is added to prepare a switching section 11 capable of controlling switching of three rule synthesis sections.

The rule synthesis section 13 is used for male voices, the rule synthesis section 12 is used for female low voice voices, and the rule synthesis section 15 is used for female high voice voices. The message output notification from the document analysis/reading conversion unit 9 to the main control unit 3 is sent by the main control unit 3 as a warning message (“There is a homograph and allophone”, “A word not found in the dictionary was detected”, etc.). Classified as an operation command (such as "Please replace the recording medium with the next one"). The main control unit 3 sends messages and files 5
When outputting the message inside, if it is an operation command, switch section 1
1 is set in the rule synthesis section 12, the warning message switching section 11 is set in the rule synthesis section 15, and the message output control section 4 is controlled to output a message. When the output of the message is finished, the switching section 11 is set to the rule synthesis section 13 to continue reading out the text. By controlling the three types of sounds in this manner, the types of output sounds can be distinguished by ear, and work efficiency can be improved.

Furthermore, by making the sounds of kanji a high female voice, the voice of kana a low voice of a woman, the voice of kana a high voice of a male voice, and the voice of English characters a low voice of a male voice, the timbre of the character type can be changed even though the reading is the same. It is possible to distinguish them by ear, and when proofreading a created document, it is possible to improve work efficiency. In addition, when using three or more types of tones as mentioned above, the rule synthesis section is not two as shown in the figure, but three or corresponding to the number of types, and the switching section is also increased according to the number. It goes without saying that a device capable of corresponding switching control is used.

〔Effect of the invention〕

As is clear from the above description, according to the present invention, it is possible to easily distinguish between operation commands, warning messages, etc. that are inserted during text reading by making the tones of the output sounds different. In addition, by making the tone of the output voice different for each type of character in the text being read out, various effects such as being able to easily proofread the created text can be obtained.

[Brief explanation of drawings]

The figure is a diagram showing the configuration of one embodiment of the present invention. 1...Display device, 2...Keyboard, 3.
...Main control section, 4...Message output control section, 5...
- Message file, 6... Dictionary, 7... Document input/output control section, 8... Document file, 9... Sentence analysis/reading conversion section, 10... Character type detection section, 11. ...Switching section, 12, 13, and 15... Rule synthesis section, 14 is a speaker. Patent Applicant: Fujitsu Limited Representative Patent Attorney Yotsube Kyotani

Claims (3)

[Claims] (1) Stores a dictionary that stores the readings of words, clauses, phrases, and other words, a text analysis and reading conversion unit that analyzes the text to be read aloud, and converts the words, etc. into readings by referring to the dictionary, and messages. A message e-file, a message output control unit that reads and outputs the message from the message file, and a message output control unit that reads the message from the message file and outputs it, reads the reading converted by the text analysis and reading conversion unit, and reads the message output by the message output control unit. A text reading device comprising a voice output unit that outputs an output, and a main control unit that instructs the message output control unit to output the message when a message output request is issued from the text analysis and reading conversion unit and performs overall control. The document reading device characterized in that the audio output unit has a function of outputting audio using a plurality of tones, and is configured to output audio by changing the audio of the reading and the message. (2) Claims characterized in that the audio output unit is configured such that the tone of the output voice of the reading is different from the tone of the output voice of the message such as an operation command or a warning message. The document reading device according to paragraph (1). (3) The text analysis and reading conversion unit has a character type detection unit that detects the character type of the sentence, and the audio output unit outputs the reading sound based on the character type detected by the character type detection unit. The document reading device according to claim 1, wherein the document reading device is configured to have different tones.