JPH0833737B2 - Audio output device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial field of application) The present invention is a voice output device for converting a Japanese sentence written in a character code into a voice and outputting the voice.
Or, it relates to a device for giving correct reading and accent to the whole sentence including a Japanese honorific expression formed by prefixing "go" to a word.

(Prior Art) FIG. 2A shows the configuration of an example of a conventional audio output device. In FIG. 2, 11 is an input terminal, 12 is a language processing unit,
13 is a dictionary unit, 14 is a voice synthesis unit, and 15 is a voice output unit.

The code data of the Japanese sentence is input to the input terminal 11,
The Japanese sentence input by the language processing unit 12 is divided into words using information such as change points of the character type, and is read for each word by referring to the dictionary of the dictionary unit 13 to add an accent, and punctuation marks and The pause (voice pause section) is set using the information of the particle. In the dictionary section 13, headings, readings, accents, parts of speech, inflections, etc. of each word are registered.
The voice synthesis unit 14 combines voice parameters by combining voice parameters based on the information such as reading, accent, and pause provided by the language processing unit 12, and outputs the synthesized voice from the voice output unit 15. .

FIG. 2b is an illustration of an example of the output of the conventional device. That is, FIG. 2b shows Example 1, Example 2 and Example 3 in which an input sentence including a honorific expression and an output are compared, and the output for a Japanese sentence illustrates word boundaries, pauses, etc. is there.

Although it is not the device itself as a publicly known document related to the above-mentioned conventional device, it is a report for practical use of research, Vol. 32 , [
No. 11], [1983], The Telecommunications Research Institute, Nippon Telegraph and Telephone Public Corporation,
Kiyoshi Kadani et al., Sentence-sound conversion method applied to data home devices, pages 25-36. This known document discloses a simple and efficient sentence / speech conversion system.

(Problems to be Solved by the Invention) However, as described above, the conventional device of this type is composed of the language processing unit, the dictionary unit, the voice synthesis unit, and the voice output unit. Since no special processing was performed on the honorific expressions prefixed with "o" or "go", correct reading and accent could not be given to the sentences containing those honorific expressions, and the intelligibility of the output voice There was a drawback that it decreased.

Further, as shown in FIG. 2B, conventionally, the word "o" or "go" that is prepended to the word and the word are treated as different words, so that an appropriate accent cannot be given, and Or, there was a drawback that the phonetic transformation for the particles "ha" and "he" before "go" was not performed, and the correct reading was not given.

In order to solve these drawbacks, the present invention is a voice output device capable of giving correct reading and accent to a honorific expression in which "o" or "go" is prefixed. Is to obtain a voice output device capable of improving the correct reading rate and the accent imparting rate of an input sentence to improve the intelligibility of an output voice.

(Means for Solving Problems) In order to solve the problems, the present invention converts a Japanese sentence including character-coded kanji, hiragana, katakana, alphabets, numbers, symbols, etc. into speech and outputs it. In the voice output device that divides the input character string into words and searches the word dictionary to read each word, accent,
A language processing unit having means for giving pause information, a word dictionary unit having means for storing reading, accent, part-of-speech, inflectional form, etc. of each word, and hiragana "o" or "go"
It has a honorific processing unit that has a means for giving correct reading and accent to Japanese honorific expressions formed by prefixing a word with a word, and a means for synthesizing a voice signal from phonetic, accent, and pause information. A voice output device is constituted by a voice synthesizing unit and a voice output unit having means for outputting a synthesized voice.

(Operation) Since the present invention is configured as described above, a correct reading and accent is given to a sentence including a honorific expression, and an output voice with high intelligibility is obtained. Further, before "o" or "go" The phonological transformation of the particles "ha" and "he" can now be correctly performed.

Embodiment An embodiment of the present invention will be described with reference to the drawings. First
FIG. 1A is a configuration diagram of an embodiment of the present invention, FIG. 1B is an explanatory diagram of output of an embodiment of the present invention, and FIG. 3 is a flowchart of an embodiment of the present invention.

In FIG. 1a, 11 is an input terminal, 12 is a language processing unit,
Reference numeral 13 is a dictionary unit, 14 is a voice synthesis unit, 15 is a voice output unit, and 16 is a honorific processing unit. Since the respective means in each of the above-mentioned units are based on a known technique, their detailed description will be omitted. The operation of one embodiment of the present invention will be described below. The code-written Japanese sentence input from the input terminal 11 is divided into words in the language processing unit 12, and the dictionary unit 13 is referred to to add information such as word number, reading, and accent form to each word. Of the output of the language processing unit 12, the word number, the reading for each word of the input sentence stored in the reading buffer, the accent type for each word of the input sentence stored in the accent type buffer, and the input sentence buffer are stored. The input character string is sent to the honorific processing unit 16. In the honorific processing unit 16, the character string in the input sentence buffer is searched for a place where a kanji string or katakana string is connected immediately after the hiragana “o” or “go”, and the hiragana “o” or "1" is set in the honorific flag buffer of the word consisting of the kanji and katakana strings following "go". The honorific process shown in FIG. 1B below is applied to the words for which the honorific flag is set.

FIG. 1b is an illustration of the output of one embodiment of the present invention. FIG. 1b shows the honorific processing for the input sentences of Example 1, Example 2 and Example 3. In FIG. 1B, if the word number for which the honorific flag is set is N, the last reading of the word with the word number N-1 in the reading buffer is "o" or "go".
Is. Therefore, "o" or "go" at the end of the reading of the word number N-1 is replaced with "no" representing a word boundary between the reading of the word number N. As a result, the last “o” or “go” is deleted from the reading of the word number N−1, and the reading of the word of the word number N is “o” or “go”.
Is added to the beginning. At this time, if the accent form of the word of the word number N in the accent form buffer is given as form 1, it is set to form 2, and if an accent form other than form 1 is provided, it is set to form 0 (flat plate). form)
Change to Here, the n-type accent form is the n-th word
If there is no accent nucleus in the beat, it is type 0. Further, as a result of the above processing, the word number N- in the reading buffer
If the last reading of the word 1 is "ha" or "he", or if the last one is "he", these "ha",
Assuming that "H" is a particle, "H" is changed to "W"
Changes the reading to "d".

Exceptionally, even if the honorific flag is set for the word with the word number N, the word number N in the reading buffer is
When the last reading of the word -1 is "o" and the part immediately before "o" is "na", the above honorific processing is not performed.

In this way, by changing the contents of the reading buffer and the accent type buffer of the portion including the honorific expression, the appropriate reading and accent type are given to the input sentence and sent to the voice synthesizer 14.

In the speech synthesis unit 14, the language processing unit 12 and the honorific processing unit 16
The voice signal is synthesized based on the information such as reading, accent shape, and pose given in step 1, and sent to the voice output unit 15.

 The voice output unit 15 outputs the synthesized voice.

FIG. 3 is a flow chart showing the operation of the embodiment of the present invention, but the detailed description thereof will be omitted.

(Effects of the Invention) As described above, the present invention provides a proper reading for a Japanese sentence including a honorific expression formed by prefixing a word with a hiragana "o" or "go", This is a voice output device capable of giving an accent, and has an advantage over the conventional device in that the correct reading rate and the accent imparting rate with respect to the input sentence are improved, and thus the intelligibility of the output voice is improved.

In particular, when applied to a voice output device in a field where many honorifics and polite words are used for input sentences, such as information guidance service and telephone shopping, the effect of the present invention is greatly exhibited.

[Brief description of drawings]

FIG. 1a is a configuration diagram of an embodiment of the present invention, FIG. 1b is an explanatory diagram of output of an embodiment of the present invention, and FIG. 2a is a configuration diagram of an example of a conventional audio output device. FIG. B is an explanatory view of an example of the output of the conventional device, and FIG. 3 is a flow chart of an embodiment of the present invention. 11 …… input terminal, 12 …… language processing section, 13 …… dictionary section, 14
...... Speech synthesizer, 15 …… Speech output, 16 …… Honorifics processor.

Claims (2)

[Claims] 1. A voice output device for converting a Japanese sentence including a kanji, hiragana, katakana, alphabet, number, symbol, etc., which is written in a character code into a voice and outputting the voice, while dividing an input character string into words. A language processing unit having means for searching a dictionary and reading each word to give accent and pose information, a word dictionary unit having means for storing each word's reading, accent, part of speech, inflection, etc. For Japanese honorific expressions formed by prefixing "o" or "go" in a word, a kanji string or katakana string is connected immediately after the hiragana "o" or "go". Search for the part where there is a word, flag the word immediately after that part with the honorific flag, and the word "O" or "go" that is the word immediately before the word with the honorific flag attached is flagged as the word immediately after single When the accented form of the combined word is 1, it is changed to 2, and when it is other than 1, it is changed to 0, and the last word of the word immediately before the combined word is changed. If the reading of "ha" or "he" or the reading just before the last one is "he", change the reading of "ha" to "wa" and "he" to "e" respectively. By using the honorific word processing unit having a means for giving correct reading and accent, a voice synthesizing unit having means for synthesizing a voice signal from reading, accent, and pose information, and a voice synthesized by the voice synthesizing unit. An audio output device comprising: an audio output unit having a means for outputting. 2. The honorific processing unit attaches a honorific flag to a word, and then adds the honorific flag to the end of the word immediately preceding the word to which the honorific flag has been applied, which is "o" or "go". If the reading of "o" is "o" and the reading immediately before "o" is "na", then each processing of combining the words, changing the accent form, and changing the reading of the word is performed. The audio output device according to claim 1, wherein the audio output device is absent.