JPS60173663A - Word processor - Google Patents

Abstract

PURPOSE:To simplify the operation and to improve the working efficiency by generating automatically a Kanji (Chinese character) converting signal when the input is switched from a single syllable unit into a single word unit in case of voice input and converting automatically a sentence data inputted at that point of time into a Kanji. CONSTITUTION:When a word voice input is used while ''I'', ''Chi'' or ''Go'' is inputted with the state of single syllable input, word/single syllable changeover switch 7 is depressed, then a normally open contact 7a is closed and a gate 6a is opened. A Kanji converting signal is givaen from a Kanji converting signal generating circuit 8 to a sentence forming section 9 when the normally open contact 7a is closed, and the ''I'', ''Chi'' or ''Go'' is converted into corresponding Kanji. Since the input by the word unit is continued afterward until the depression of the word/single syllable changeover switch 7 is released, and convertion of ''strawberry field'' is conducted when ''field'' is uttered for example succeedingly.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a voice input type word processor, and more specifically, the present invention relates to a voice input type word processor. In this case, a kanji conversion signal is automatically generated, and the meaning data input at that time is automatically converted to kanji, simplifying the operation.
This paper proposes a word processor that improves work efficiency.

[Prior art]

Recently, word processors have been developed that allow voice input without manual input, but these word processors are capable of inputting single syllables such as "i", "chi", "go", etc. (-), (city), (position), (position) J, "go (five), (go), (back), (base), (word)・
...'' Furthermore, it is common to be able to identify and input word units, such as [Strawberry (strawberry), (-period)], and the switching method is switch operation or discrete input. A system has also been developed in which the syllable is automatically switched depending on whether the syllable in the voice uttered is plural or singular. )J, r5 (go)'' It was difficult to distinguish whether the two numbers were uttered.

In addition, in the case of monosyllable input, to convert the text data input by voice into kanji, operate the kanji conversion key or
Alternatively, it is necessary to input the word to be converted into kanji by identifying it in monosyllabic units, then switch to identifying by word, and then utter the word "henkan (conversion)" to convert it into kanji.

In this way, conventional voice input word processors operate a switch to switch between monosyllable input and word input and a kanji conversion key, or input monosyllables and then switch to word input. It is necessary to convert the meaning of the inputted sentence into kanji by saying ``Henkan (conversion)'', and such double operations are complicated and cause a decline in work efficiency.

[Purpose of the invention]

The present invention has been made in view of the above circumstances, and
When switching from monosyllable input to word input, a kanji conversion signal is automatically generated, kanji conversion is performed, and then the word input state is entered. The purpose of the present invention is to provide a voice recognition word processor which eliminates the trouble of operating both a changeover switch for word-by-word input and a kanji conversion key, and improves work efficiency.

[Structure of the invention]

The word processor according to the present invention includes an identification section that identifies text data input by voice in units of monosyllables or words, and converts the text data identified by the identification section into sentences containing kana and kanji. A writing department that creates the text,
The word processor is equipped with a kanji conversion signal generation circuit that provides a conversion signal for converting the text data into kanji to the text creation section, and the text data inputted by the voice is identified in units of single pre-clauses or in units of words. A kanji conversion signal is provided from the kanji conversion signal generation circuit to the sentence creation section when the kanji conversion switch is operated to identify sounds from monosyllable units to word units. It is characterized by something that has been done to the best of its ability.

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below based on drawings showing embodiments thereof.

The drawing is a block diagram showing the configuration of main parts of a word processor according to the present invention.

The voice input to the microphone 1 is given as an electrical signal to the preprocessing unit 2, where it is subjected to preprocessing for extracting its features, such as consonant emphasis, and then to a word voice feature extraction unit 3a and a monosyllabic voice feature extraction unit. 3b and its features are extracted.

The features of the extracted speech are identified by the RA language speech identification section 5a and the monosyllabic speech identification section 5b and specified as one syllable or -word, but the standard feature set necessary for this identification is determined in advance from the word speech standard. Feature parameter storage unit 4a
and is registered in the monosyllabic speech standard feature parameter storage section 4b.

The sounds identified by the word sound identification unit 5a and the monosyllabic sound identification unit 5b are provided to the sentence creation unit 9 via the gates 6a or 6b, respectively, and are converted into sentences.
Only one of a and 6b is opened by operating the word/single syllable changeover switch 7. That is, the push button type word/single syllable changeover switch 7 has both contacts 7a and 7b.
Each end is connected to a predetermined potential Vcc through an appropriate resistor.
The other end of the normally open contact 7a is connected to the ground potential, one end of the normally open contact 7a is connected to the low active terminal of the gate 6a and the kanji conversion signal output circuit 8, and one end of the normally open contact 7b is connected to the gate 6b. is connected to the low active terminal of

Therefore, in the illustrated state where the word/single syllable selector switch 7 is not operated, the normally open contact 7a is open, and the potential Vcc (high level) is applied to the low active terminal of the gate 6a, and the gate t-6a is Since the gate is closed, no signal is sent from the word voice recognition section 5a to the sentence creation section 9, and the - lightning close contact 7b is closed, and the low active terminal of the gate 6b is at ground potential (low level), and the gate 6b is closed. ) 6b is open and a signal is sent from the monosyllabic speech recognition section 5b to the sentence creation section 9.

On the other hand, when the word/single syllable switch 7 is pressed, the normally open contact 7a is closed and the gate 6
The low active terminal of a becomes ground potential and the gate 6a
The gate is open, and a signal is sent from the word voice recognition unit 5a to the sentence creation unit 9, and the -marine close contact 7b is opened and the gate 6 is opened.
A predetermined potential Vcc is applied to the low active terminal of 6b, and the gate 6b is closed, and no signal is sent from the monosyllabic speech recognition section 5b to the sentence creation section 9.

Therefore, when the word/II syllable switch 7 is not operated, the voice signal identified and specified in monosyllable units is given to the sentence creation section 9 via the gate 6b, and the word/II syllable changeover switch 7 is not operated.
When the monosyllabic changeover switch 7 is pressed, a voice signal identified and specified on a word-by-word basis is provided to the sentence creation section 9 via the gate 6a.

Further, when the word/single syllable changeover switch 7 is pressed, the potential Vcc applied to the kanji conversion signal generation circuit 8 also becomes the ground potential due to the closing of the normally open contact 7a, so that the kanji conversion signal generation circuit 8 generates a kanji conversion signal and supplies it to the sentence creation section 9.

When the text creation section 9 is given the above-mentioned kanji conversion No. 48,
The text data given from the word speech identification section 5a or the monosyllabic speech identification section 5b after the last time when a Kanji conversion signal was given or the non-conversion key (not shown) was operated is converted into Kanji.

Words regarding the operation of the device of the present invention configured as described above:
The operation related to the operation of the monosyllabic changeover switch 7 will be mainly described below.

For example, when inputting "i", "chi", or "go" in a monosyllable input state, if you want to input a word voice input, the word/
When the monosyllabic changeover switch 7 is pressed, the open contact 7a is closed and the gate 6a is opened, and at the same time the normally open contact 7a is closed, the kanji conversion signal generation circuit 8 sends a message to the sentence creation section 9.
A kanji conversion signal is given to ``i'', ``chi'' or ``go''.
is converted into the kanji character "-", "ichi", etc., or "go", "ago", etc. From then on, input in i-word units continues until the word/single syllable changeover switch 7 is released from the pressed state, so that if the user continues to say, for example, "hataga", it will be converted to "strawberry field".

In addition, if you only want to convert text data entered in monosyllable units into kanji, for example, input "i" or "chi" and then press the word/monosyllable switch 7 to convert the kanji "-".
, and if another kanji "ichi" is needed, the word /
If the monosyllabic changeover switch 7 is pressed once, the words are converted into "city", "location", and "location" in sequence.

[Effects] As detailed above, in the word processor according to the present invention, when converting input in units of monosyllables to input in units of words, the operation for converting kanji, which was necessary in the past, is no longer necessary, and the process is automatically performed. Since the input state is in units of words that have been converted into kanji, the operation becomes simple and easy, and the efficiency of text input is improved.

[Brief explanation of the drawing]

The drawing is a block diagram of main parts of a word processor according to the present invention. 4a... Word sound identification unit 4b... Monosyllabic sound identification unit 7... Word/monosyllabic changeover switch 8... Kanji conversion signal generation circuit 9... Sentence creation unit Patent Applicant: Sanyo Electric Co., Ltd. Company agent Patent attorney Noboru Kono

Claims (1)

[Claims] 1. An identification unit that identifies text data input by voice in units of monosyllables or words, and converts the text data identified by the identification unit into a text containing kana and kanji. a word processor that includes a sentence creation section that creates a sentence, and a kanji conversion signal generation circuit that provides a conversion signal for converting the sentence data into kanji to the sentence creation section, A changeover switch is provided that causes the identification section to identify sounds in units of pre-clauses or in units of words, and when the changeover switch is operated to identify sounds from monosyllabic units to word units, the kanji conversion signal generating circuit A word processor characterized in that a kanji conversion signal is given to a text creation section.