JP2003316388A - Voice input device, method for the same, and voice input program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize correct input processing of ordered items without being bound by the order of voice input. <P>SOLUTION: A phrase registered beforehand is extracted one by one out of a sentence inputted by voice. For each of the extracted phrases, an item containing the meaning of the phrase is determined. Whether an item combination pattern of each of the extracted phrases is the pattern satisfying the prescribed fixed conditions or not is judged. When the item combination pattern of each of the extracted phrases is judged to be the pattern satisfying the fixed conditions, each phrase is fixed as voice input data. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information processing apparatus that receives and processes order contents, for example, a ticket issuing device for a gas station refueling system, a ticket issuing device for airline tickets, boarding tickets, etc., in restaurants and fast food restaurants. The present invention relates to a voice input device used as an input device of order contents for an order registration device and the like, a method thereof, and a voice input program.

[0002]

2. Description of the Related Art For example, at a gas station, when a customer visits a store, a store clerk first refuels gasoline (regular gasoline, high-octane gasoline, light oil, etc.), refueling amount (quantity unit or monetary unit), and payment method ( Cash payment,
Ask customers for credit payments, etc.). On the other hand, for example, when the customer replies, “Regular, full, cash.” The clerk says that the type of gasoline to be refueled is “regular gasoline,” and the amount of refueling is “full.” The fact that the payment method is "cash payment" is input to the slip issuing device, and then refueling is started. By doing this, when refueling is completed, the slip issuing device automatically prints and issues a settlement slip according to the actual refueling amount, so that the clerk receives the refueling fee from the customer according to the settlement slip. .

In such a gas station refueling system, conventionally, as a method of inputting the customer's order contents into the slip issuing device, a method using an input device such as a keyboard or a touch panel has been generally used. However, with this method, the store clerk had to manually input the content heard from the customer, which was troublesome twice and input errors were easy to occur. In addition, the customer has to wait for the clerk to announce the order contents until approaching, and the clerk who heard the order goes to the place where the slip issuing device is placed and inputs the customer's order contents. After that, I had to go back to the place where the car was stopped to start refueling, and work efficiency was poor.

Therefore, in recent years, it has been considered that a customer, who came to refuel, inputs the order contents to the slip issuing device directly from the inside of the vehicle by connecting the voice input device to the slip issuing device as an input device. By doing this, the labor of the store clerk can be reduced, input mistakes can be eliminated, and the waste of the store clerk manually reciprocating between the car and the slip issuing device to manually enter the customer's order details can be eliminated, improving work efficiency. Can be expected.

However, the conventional voice input device has a format in which words recognized from voice data input through a microphone are fetched one by one in an input order and transferred to an information processing device such as a computer in that order. On the other hand, in an information processing device such as a slip issuing device that receives and processes order contents, the order of receiving order contents items is predetermined, and the transferred words and phrases are processed as data indicating the order contents of the corresponding items in that order. Will be done.

Therefore, for example, if the slip issuing device of the refueling system is programmed to first receive the type of gasoline, then the refueling amount, and finally the settlement method, the slip issuing device. Processes the first word transferred from the voice input device as data indicating the type of gasoline, the second word transferred as data indicating the refueling amount, and the last transferred data is settled. It will be processed as data indicating the method. Therefore, the person who performs voice input needs to say the order details in a predetermined order, such as first ordering the type of gasoline, then the refueling amount, and finally the payment method. It was

[0007]

As described above, when the voice input device is used as an input device for order contents, the order in which the order contents are input is fixed in the related art. Even if a customer constructs a refueling system that uses a voice input device to input refueling order details into a voucher issuing device, the order of notifying all customers of the order details at a gas station used by an unspecified number of customers It is difficult to thoroughly enforce the order, and there is a high possibility that the slip issuing device cannot correctly recognize the order contents and does not function properly.

The present invention has been made under such circumstances, and an object of the present invention is to provide a voice input device and method and a voice input program capable of correctly processing order contents without being caught in the order of voice input. Is to provide.

[0009]

The voice input device of the present invention is a phrase extraction unit for extracting one by one a pre-registered phrase from a sentence input by voice, and each phrase extracted by this phrase extraction unit. A word / phrase item judging means for judging an item including the meaning of the word / phrase, a word / phrase storing means for storing each word / phrase for which an item is judged by the word / phrase item judging means, and a word / phrase storing means for storing each word / phrase Determining condition judging means for judging whether or not the item combination pattern of the words and phrases satisfies a predetermined definite condition, and this definite condition judging means judges that the item combination pattern of each word and phrase is a pattern satisfying the definite condition. Then, an input confirming means for confirming each word / phrase stored by the word / phrase storage means as voice input data is provided.

Therefore, when the combination pattern of the items including the meanings of the respective words and phrases extracted from the voice input sentence matches the pattern satisfying the predetermined fixing condition, the respective words and phrases are confirmed as the voice input data. Therefore, the order of the words and phrases forming the sentence may be different.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings. In addition, this embodiment is a case where the present invention is applied to a voice input device used as an order content input device for a slip issuing device of a gas station refueling system.

FIG. 1 is a block diagram showing the configuration of a voice input device 1 according to this embodiment. This voice input device 1
Is composed of a microphone 11 as a voice input section, a speaker 12 as a voice output section, a voice engine section 13, a voice input confirming section 14, and a voice input switch 15. The voice input confirming unit 14 is connected to an information processing device that receives and processes order contents, that is, a slip issuing device 2 of a gas station refueling system in this embodiment.

The voice input switch 15 is a switch that is turned on when the user inputs voice from the microphone 11,
A microphone 11 and a speaker 12 are provided in the vicinity thereof. In this embodiment, a voice input switch 15, a microphone 11 and a speaker 12 are provided at a position where a customer stopped at a fueling place of a gas station can operate from inside the vehicle.

The voice engine section 13 includes a microphone 11
A / D (analog / digital) converter 16 for converting an analog voice signal input from the device into a digital voice signal, a voice recognition unit 17 for recognizing voices one by one syllable from the digital voice signal to create voice data, voice input A voice generation unit 18 that generates a digital voice signal from the guidance data supplied from the determination unit 14, and the speaker 12 by converting the digital voice data into analog voice data.
It is constituted by a D / A (digital / analog) converter 19 for outputting to.

The voice input confirming section 14 takes in the voice data generated by the voice recognizing section 17 while the voice input switch 15 is on, and the voice to be processed by the slip issuing device 2 from this voice data. It has a function of fixing input data and transferring it to the slip issuing device 2.

In this embodiment, the slip issuing device 2 receives orders for gasoline type, refueling amount, and settlement method in this order, and after these order contents are input,
When refueling is completed, a settlement slip according to the actual refueling amount is printed and issued. Therefore, the voice input determination unit 14 determines the voice input data from these words and phrases when the words and phrases indicating the type of gasoline, the phrases indicating the amount of refueling, and the phrases indicating the settlement method are obtained from the voice data. , To the slip issuing device 2. On the other hand, when there is a shortage of a word or phrase indicating the content of an order, guidance data for prompting voice re-input of the word or phrase indicating the order content is output to the voice generation unit 18, and the guidance is output by voice. Further, when the word / phrase includes, for example, a word / phrase for which the order content cannot be determined because it represents a generic term for the item, guidance data that prompts re-input of a voice for fixing the word / phrase is output to the voice generation unit 18,
The guidance is output by voice.

The specific configuration of the voice input confirming section 14 will be described below.

FIG. 2 is a block diagram showing the configuration of the voice input confirming section 14. The voice input device 1 includes a CPU (Central Processing Uni
t) 141, a ROM (Read Only Me) in which fixed data such as programs executed by the CPU 141 are stored in advance.
mory) 142, an RA in which various memory areas used by the CPU 141 as a temporary storage area for data are formed
An M / R (Random Access Memory) 143, an I / O (Input / Output) port 144 from which voice data is input from the voice recognition unit 17 of the voice engine unit 13 and guidance data is output to the voice generation unit 18. HDD (Hard Disk Drive) that stores various data files
A device 145, a communication interface 146 that is connected to the slip issuing device 2 and transfers voice input data, and an input port 1 to which a signal from the voice input switch 15 is input.
47, etc., CPU 141, ROM 142, R
AM 143, I / O port 144, HDD device 145
With the communication interface 146, the input port 147, etc.,
It is connected by the system bus 148. Here, the voice input determination unit 14 includes the CPU 141, the ROM 142, and the RA.
The main component is a microcomputer including M143.

As shown in FIG. 3, the HDD device 145 stores a vocabulary file 31, a dictionary file 32, a guidance file 33, and a confirmation condition table 34.

As shown in FIG. 4A, the vocabulary file 31 includes phrase voice data in which the feature amount of voice when a predetermined phrase is generated is quantified, and a preset phrase for the phrase. A record including a word code different for each word is stored, and in particular, a large number of words that are considered to be used when inputting order contents by voice to the slip issuing device 2 are selected, and the record of each word is respectively selected. Remembered For example, “regular”, “regular gasoline”, “high octane”, “high octane gasoline”, “light oil”, “gasoline”, etc. are selected as the terms used when ordering the type of gasoline, and the terms of these terms are selected. The voice data and the phrase code are stored in the vocabulary file 31. Also, "full tank", "10 liters", "20 liters", "1000 yen", "2000 yen", etc. are selected as words and phrases used when ordering the amount of refueling,
The phrase voice data and the phrase code of these phrases are stored in the vocabulary file 31. The terms used when ordering the payment method are "cash", "cash payment", "card", "credit", "debit",
"Credit card", "debit card", etc. are selected, and the phrase voice data and the phrase code of these phrases are stored in the vocabulary file 31.

As shown in FIG. 4B, the dictionary file 32 stores a record including a word / phrase code, an item code, an indeterminate flag, a guidance number, and a plurality of link word / phrase codes. There are stored a plurality of records each linked to the phrase code of each phrase stored in.

In the record item of the dictionary file 32, the item code is a code for identifying an item set according to the meaning of various words and phrases. In this embodiment, the item "item" (meaning that the type of gasoline is indicated) Item code = 1), item "quantity" (item code = 2) indicating the amount of refueling, item "amount" (item code = 3) indicating the amount equivalent to the amount of refueling, and the settlement method The meaning item “payment method” (item code = 4) is set.

The indetermination flag is a flag for identifying whether or not the corresponding word / phrase is a word / phrase for which the order content can be fixed. As described above, "1" is set for an indefinite word. Then, the guidance number and the link word code are set only for the uncertain words. The guidance number is a number that is set for the guidance data that prompts the user to re-enter a voice that confirms the uncertain phrase.
The link phrase code is a phrase code of some or all of the phrases whose order details can be confirmed by the phrases included in the item into which the undetermined phrase is classified.

For example, there are "regular", "high-octane", "gasoline", and the like as words and phrases that are input by voice when ordering the type of gasoline. The former two can specify the type of gasoline, but "gasoline" It is not possible to determine whether it is regular gasoline or high-octane gasoline by itself.
Therefore, the uncertain flag is set to "1" for the phrase code of the phrase "gasoline", and guidance data that prompts voice re-input to confirm the type of gasoline as the guidance number (for example, "is gasoline .........? )) And link word codes “regular” and “high-octane”. Similarly, there are “credit”, “debit”, “card”, etc. as words that are input by voice when ordering a payment method. The former two can specify the payment method, but the “card” alone is a credit. It is not possible to identify whether it is a card payment or a debit card payment. Therefore, the uncertain flag is set to “1” for the phrase code of the phrase “card”, and guidance data that prompts voice re-input to confirm the card type as the guidance number (for example, “Is the card ...? )), And the word codes of the words "credit" and "debit" are set as the link word codes.

As shown in FIG. 4 (c), the guidance file 33 stores a record consisting of a guidance number and guidance data. Various types of voice output are required until the voice input data is determined. Guidance data are stored together with different guidance numbers.

As shown in FIG. 5, the fixed condition table 34 includes items set according to the meaning of various words and phrases (in this embodiment, four types of "item", "quantity", "amount", and "payment method"). Data indicating whether or not the item combination pattern includes a corresponding item for each condition number for identifying the combination pattern (1: included, 0: not included).
And a guidance number of guidance data that prompts voice re-input of a word having a meaning of an item that is insufficient to satisfy the final condition when the item combination pattern does not satisfy the final condition. is there.

That is, the item combination pattern identified by the condition number P1 is a pattern including the items "item", "quantity" and "payment method" of the data "1", and the item combination identified by the condition number P2. The pattern is a pattern including the items “item”, “price” and “payment method” of the data “1”, and since all satisfy the confirmation condition as the voice input data to the slip issuing device 2, the guidance number is It is not set (= 0).

On the other hand, none of the item combination patterns identified by the condition numbers P3 to P12 satisfy the confirmation condition as the voice input data for the slip issuing device 2, so that the guidance numbers of the corresponding guidance data are set. ing. For example, condition number P
In the patterns identified by 3 and P4, since the word having the meaning of the item "payment method" has not been entered, guidance data prompting voice re-input of the payment method (for example, "What is the payment method?") Guidance number [100] is set. Further, in the pattern identified by the condition number P5, since the word having the meaning of the item "quantity" or "amount" has not been input, the guidance number [101] of the guidance data for prompting voice re-input of the refueling amount is set. Has been done. Similarly, in the patterns identified by the condition numbers P6 and P7, since the word having the meaning of the item “item” has not been input, the guidance number [102] of the guidance data for prompting voice input of gasoline type is the same. In the pattern that is set and identified by the condition number P8, since the words having the meanings of “quantity” or “amount of money” and “payment method” are not input, the voice re-input of the refueling amount and the payment method is input. The guidance number [103] of the guidance data for prompting the user is set, and in the pattern identified by the condition numbers P9 and P10, since the terms having the meanings of the items “item” and “payment method” have not been input, The guidance number [104] of the guidance data that prompts voice re-input of the gasoline type and payment method is set, and the condition number P1 In the pattern identified by, since the terms "item" and the terms "quantity" or "amount" respectively have not been entered, the guidance number of the guidance data that prompts voice re-entry of the gasoline type and refueling amount. [105] is set.

Now, the CPU 141 forms the memory areas 61 and 62 shown in FIG. 6 in the RAM 143 by the voice input program stored in the ROM 142, and repeatedly executes the processing shown in the flowchart of FIG. 7.

The memory area 61 is an area for storing the phrase voice data and the phrase code extracted from the vocabulary file 31, in the order of extraction, and is hereinafter referred to as the phrase memory 61. The memory area 62 stores the words and phrases having the meaning of each item, which is set according to the meaning of each word (four types of “item”, “quantity”, “amount”, and “payment method” in this embodiment). A region (phrase storage means) for storing the phrase voice data, the phrase code, and the uncertain flag one by one.
It is referred to as an itemized memory 62.

With the word memory 61 and the itemized memory 62 formed in the RAM 143, the CPU 141
As (step) 1, it is waiting for the voice input switch 15 to be turned on. In the initial state, the phrase voice data area and the phrase code area of the phrase memory 61 and the phrase voice data area, the phrase code area and the indetermination flag area of the itemized memory 62 are all cleared.

When it is determined that the voice input switch 15 is turned on by the signal input to the input port 147, the CPU 141 fetches the voice data recognized by the voice recognition unit 17 as ST2. The audio data is ST
It is continuously taken in as 3 until the voice input switch 15 is turned off.

When it is determined that the voice input switch 15 is turned off by the signal input to the input port 147, the CPU 141 executes the word / phrase extraction process at ST4. That is, the voice data captured from the voice recognition unit 17 is compared and collated with the phrase voice data of various phrases registered in the vocabulary file 31 in the order of input, and the phrase that matches the phrase voice data registered in the vocabulary file 31 is matched. Are extracted one by one. Every time a phrase is extracted, the phrase voice data of the phrase and the corresponding phrase code are stored in the phrase memory 61 in the order of extraction (phrase extraction means). It should be noted that from the voice data 1
If neither of them can be extracted, an error is set, and the operation of waiting for the voice input switch 15 to be turned on again is waited.

When the phrase voice data is extracted from the voice data, the CPU 141 executes the phrase item determination process. That is, the phrase voice data and the phrase code stored in the phrase memory 61 are arranged in ascending order of memory number,
That is, the words and phrases extracted from the voice data are read in order. Then, the dictionary file 32 is searched for the word code to obtain the corresponding item code and the uncertain flag, and the word voice data, the word code, and the uncertain flag are stored in the corresponding item area of the item-based memory 62. . At this time, if data is already stored in the corresponding item area, the existing data is overwritten and stored. Therefore, if a plurality of phrase voice data and phrase codes of a phrase having the same item meaning are stored in the phrase memory 61, the phrase extracted most later, that is, the phrase input by voice most later is used. When the word is valid, the word voice data, the word code, and the indetermination flag of the word are stored in the itemized memory 62 (word item determination means).

When the item of each phrase voice data stored in the phrase memory 61 has been determined, the CPU 141 causes ST6
The presence or absence of an indeterminate phrase is determined as. That is, the uncertain flag stored in the itemized memory 62 is checked to determine whether or not the uncertain flag “1” is stored.
Then, when the uncertain flag “1” is stored, it is determined that there is an uncertain phrase, and when the uncertain flag “1” is not stored, it is determined that there is no uncertain phrase.

When it is judged that there is an indeterminate phrase, CP
In ST7, the U141 searches the dictionary file 32 again for the word / phrase code corresponding to the uncertain flag “1” in ST7, and reads the corresponding guidance number and link word / phrase code. Further, the guidance data corresponding to the guidance number is read from the guidance file 33, and the phrase voice data corresponding to each link phrase code is read from the vocabulary file 31. Then, based on the guidance data and each word / phrase voice data, guidance data for prompting re-input of a voice for fixing an uncertain word, that is, so-called fixed word inquiry guidance data is created. After that, ST
The determined phrase inquiry guidance data is output to the voice generation unit 18 as 8 and is output as voice from the speaker 12 (determined phrase prompting means). After that, the CPU 141 proceeds to ST
It returns to 1 and waits for the voice input switch 15 to be turned on again.

When it is determined that there is no uncertain phrase, CP
The U 141 executes the definite condition determination processing in ST9. That is, the item-by-item memory 62 is checked to distinguish between items in which the word / phrase voice data, the word / phrase code, and the data of the indetermination flag “0” are stored, and items that are not stored. Then, the confirmation condition table 34 is searched for the combination pattern of the items in which each data is stored, that is, the item combination pattern of each word or phrase extracted from the voice data, and the guidance number of the same pattern is acquired.

Here, when the guidance number of the same pattern is other than "0", that is, when the guidance number is set, the CPU 141 determines the item combination pattern of each word / phrase extracted from the voice data in ST10. Input is not confirmed because the pattern does not satisfy the conditions. On the other hand, when the guidance number of the same pattern is “0”, the item combination pattern of each word / phrase extracted from the voice data is a pattern that satisfies the confirmation condition, and therefore the input is confirmed (the confirmation condition determination means). ).

If the input is not confirmed, the guidance data corresponding to the guidance number is read from the guidance file 33 in ST11, and the voice input of the word having the meaning of the item insufficient to satisfy the confirmation condition is performed again. Create guidance data for prompting, so-called insufficient item word inquiry guidance data. Then, in ST12, the insufficient item word / phrase inquiry guidance data is output to the voice generation unit 18 and output by voice from the speaker 12 (insufficient word / phrase prompting means). After that, the process returns to ST1 and waits for the voice input switch 15 to be turned on again.

When the input is confirmed, the word code of each item stored in the itemized memory 62 in ST13 is read in the order of item "item", "quantity" or "amount", and "payment method". When the voice input confirmation data is created by combining these word codes in order, the voice input confirmation data is transferred to the slip issuing device 2 via the interface 146 (input confirmation means).

After that, if the CPU 141 clears the word / phrase memory 61 and the item-based memory 62 in ST14 and returns to the initial state, the processing of this time is ended. And
The voice input switch 15 waits for the next ON operation, and if ON, the processing from ST1 is repeated.

As described above, in the present embodiment, the voice input device 1 is used as an input device of the order contents for the slip issuing device 2 of the gasoline refueling system. Then, the customer who has come to refuel with gasoline turns on the voice input switch 15 from inside the vehicle after stopping at a predetermined position, and speaks into the microphone 11 the order contents, that is, the type of gasoline, the amount of refueling, and the settlement method. At this time, the order of notifying the order details is not particularly limited.

Now, for example, it is assumed that the customer utters the order contents "gasoline for 2000 yen in cash". Then, as shown in FIG. 8A, the voice data “gasoline for 2000 yen for cash” is recognized by the voice recognition unit 17 and input to the voice input determination unit 14.

In the voice input confirming section 14, a word / phrase extraction process is executed on the voice data "gasoline for 2000 yen for cash". As a result, the word “gasoline” indicating the type of gasoline, the word “cash” indicating the payment method, and the word “2000 yen” indicating the amount of money corresponding to the amount of refueling are extracted from the voice data. Then, as shown in FIG. 8B, the phrase voice data and the phrase code of each phrase are stored in the phrase memory 61 in the order of extraction.

Next, in the voice input confirming section 14, word / phrase item determination processing is executed. As a result, the item-based memory 62
8C, the phrase voice data of the phrase “gasoline”, the phrase code, and the indeterminate flag are stored for the item “item”, and the phrase “2000” is stored for the item “amount”. The phrase voice data of “Yen”, the phrase code and the indetermination flag are stored respectively, and the phrase voice data of the phrase “cash”, the phrase code and the indetermination flag are stored respectively for the item “payment method”.

In this case, since the word "gasoline" is an uncertain word (uncertain flag = 1), the guidance data "Is gasoline ...? ...?" From the phrases “regular” and “high octave” corresponding to the link phrase code of the phrase “gasoline”, the fixed phrase inquiry guidance data “is gasoline regular or high octave” shown in FIG. 8D is created. Then, the determined phrase inquiry guidance data is output to the voice generation unit 18, converted into voice, and output from the speaker 12.

It is assumed that a customer who hears this guidance similarly utters, for example, "regular". Then,
As shown in FIG. 8 (e), the voice data “regular” is recognized by the voice recognition unit 17, and the voice input determination unit 14 is recognized.
Entered in. In the voice input confirming unit 14, word extraction processing is executed on the voice data “regular”. As a result, the word “regular” representing the type of gasoline is extracted from this voice data. And FIG.
As shown in (f), the phrase voice data of the phrase “regular” and the phrase code are added to the phrase memory 61.

Next, in the voice input confirming section 14, the word / item determination processing is executed again. As a result, as shown in FIG. 8G, the item-specific memory 62 is overwritten with the phrase audio data of the phrase “regular”, the phrase code, and the indetermination flag with respect to the item “item”.

As a result, since the uncertain words do not exist any more, the voice input confirming section 14 executes the confirming condition judging process. In this case, the item-based memory 62 stores data of words and phrases that belong to the items “item”, “amount of money”, and “payment method”.
Is recognized as a pattern that matches. This pattern is a pattern in which the guidance number is not set, that is, the confirmation condition is satisfied, so the words "regular" and "2000
Voice input confirmation data is created from the word codes of “yen” and “cash” and transferred to the slip issuing device 2.

In the slip issuing device 2, the first word code of the voice input confirmation data is recognized as the data indicating the type of gasoline, the next word code is recognized as the data indicating the amount of refueling, and the last word code is settled. Recognize and process as data indicating the method. Therefore, the issuance of the slip is processed normally.

It is also assumed that another customer utters the order contents as "Regular gasoline with a card". Then, as shown in FIG. 9A, the voice data “Regular gasoline with a card” is recognized by the voice recognition unit 17 and input to the voice input determination unit 14.

In the voice input confirming unit 14, a word / phrase extraction process is executed on the voice data "Regular gasoline with a card". As a result, the phrase “regular gasoline” indicating the type of gasoline and the phrase “card” indicating the payment method are extracted from this voice data. And
As shown in FIG. 9B, the phrase voice data and the phrase code of each phrase are stored in the phrase memory 61 in the order of extraction.

Next, in the voice input confirming section 14, word / phrase item determination processing is executed. As a result, the item-based memory 62
As shown in FIG. 9 (c), the phrase voice data of the phrase "regular gasoline", the phrase code, and the indeterminate flag are stored for the item "item", and the phrase is set for the item "payment method". The word voice data of the "card", the word code, and the uncertain flag are stored.

In this case, since the word "card" is an uncertain word (uncertain flag = 1), the guidance data "card is ...
?? ”and the words“ credit ”and“ debit ”corresponding to the link word code of the word“ card ”, the fixed word inquiry guidance data“ card is credit or debit ”shown in FIG. 9D. Is it created? Then, the determined phrase inquiry guidance data is output to the voice generation unit 18, converted into voice, and output from the speaker 12.

It is assumed that a customer who hears this guidance similarly utters, for example, "It is a credit card". Then, as shown in FIG. 9E, the voice data "Credit card" is recognized by the voice recognition unit 17 and input to the voice input determination unit 14. The voice input confirmation unit 14 executes a phrase extraction process on the voice data “is a credit card”. As a result, the phrase "credit card" representing the payment method is extracted from this voice data. Then, as shown in FIG.
The phrase voice data of the phrase “credit card” and the phrase code are added to the phrase memory 61.

Next, in the voice input confirming section 14, the word / item determining process is executed again. As a result, in the item-based memory 62, as shown in FIG. 9G, the phrase voice data, the phrase code, and the indetermination flag of the phrase “credit card” are overwritten with respect to the item “payment method”.

As a result, since the uncertain words do not exist any more, the voice input finalizing section 14 executes the finalizing condition judging process. In this case, since the item-based memory 62 stores data of words and phrases that belong to the items “item” and “payment method”, respectively, it is recognized that the pattern matches the condition number P5. Since this pattern is a pattern in which the guidance number [101] is set, that is, a pattern that does not satisfy the confirmation condition, the guidance data of the guidance number [101] shown in FIG. Insufficient item phrase Inquiry guidance data “Please enter the planned amount or amount of refueling” is created. Then, the insufficient item word / phrase inquiry guidance data is output to the voice generation unit 18, converted into voice, and output from the speaker 12.

It is assumed that a customer who hears this guidance utters "full tank" in the same manner. Then, as shown in FIG.
As shown in (i), the voice data “full tank” is recognized by the voice recognition unit 17 and input to the voice input determination unit 14. The voice input confirming unit 14 executes a word / phrase extraction process on the voice data “full”. as a result,
From this voice data, the phrase "full tank" representing the amount of refueling is extracted. Then, as shown in FIG. 9J, the phrase voice data and the phrase code of the phrase “mantan” are stored in the phrase memory 6
Added to 1.

Next, in the voice input confirming section 14, the word / item determination processing is executed again. As a result, in the item-based memory 62, as shown in FIG. 9K, the phrase voice data, the phrase code, and the indetermination flag of the phrase “full” are stored for the item “quantity”. Next, since there is no uncertain phrase, the voice input confirmation unit 14 executes the confirmation condition determination process. In this case, the itemized memory 6
Items 2 are "Item", "Quantity" and "Payment method"
Since the data of the words and phrases that respectively belong to are stored,
It is recognized that the pattern matches the condition number P1. In this pattern, since the guidance number is not set, that is, the confirmation condition is satisfied, the voice input confirmation data is created from the phrase codes "regular gasoline", "full tank", and "credit card", and the sequential slip issuing device is created. 2 is transferred.

In this case as well, in the slip issuing device, the first phrase code of the voice input confirmation data is recognized as the data indicating the type of gasoline, the next phrase code is recognized as the data indicating the refueling amount, and the last phrase is recognized. Since the code is recognized and processed as the data indicating the payment method, the issuance of the slip is processed normally.

As described above, according to the present embodiment, the order of the input items when voice inputting the refueling order contents to the slip issuing device 2 is not limited, so that it is easy to use, and an unspecified number of customers can use it. It becomes possible to input the order contents into the slip issuing device 2 by voice. As a result, the labor of the store clerk can be reduced, input errors can be eliminated, and the waste of the store clerk manually reciprocating between the car and the slip issuing device to manually input the order details of the customer can be omitted, so that the work efficiency can be improved.

Further, in the present embodiment, when the confirmation condition of the order content is not satisfied only by the first phrase input by voice, a word having a meaning of an item which is insufficient to satisfy the confirmation condition is ordered. Guidance for prompting voice re-input is automatically generated and voice output to the user. Then, if the determination condition is satisfied from the phrase input by voice after that, the voice input data is determined. Therefore, even if there is an input item that is omitted when inputting the order contents by voice, only the order contents of the item need to be input again, so that anyone can easily and correctly input the order contents by voice. .

When an uncertain phrase whose order content cannot be confirmed is included in the phrase input by voice, a guidance for prompting the user to re-enter the voice for confirming the phrase is automatically generated, and the user is prompted. To output audio. Then, when an uncertain phrase is determined from the phrase that is input by voice after that, the phrase that is input subsequently is validated and the voice input data is determined. Therefore, even if the customer can't specify whether gasoline is regular gasoline or high-octane gasoline because he says only "gasoline" regarding the type of gasoline, the type of gasoline can be confirmed by subsequent voice input. Anyone can easily and correctly enter the order details by voice.

By the way, in this embodiment, when there are a plurality of words having the same meaning of the words in the words extracted from the voice data and stored in the word memory 61, the word inputted by the earliest voice is used. Is enabled and stored in the item-based memory 62. Therefore, for example, when the type of gasoline is mistakenly corrected and later corrected, the word type of the corrected type becomes effective, which is also convenient from this point of view.

In this embodiment, the present invention is applied as the order content input device for the slip issuing device of the gas station refueling system, but the information processing device to which the present invention can be applied is not limited to this. It can also be used as a ticket issuing device for airline tickets, passenger tickets, etc., or as an order content input device for order registration devices at restaurants and fast food restaurants. For example, when used as an input device for a ticket issuing device for airline tickets, the items including the meanings of the words are set such as departure date, departure place, destination, airline, seat (non-smoking seat, window side, aisle side, etc.). To be done.

When used as an input device for the order registration device at a fast food restaurant, the item name, quantity, settlement method, etc. are set as items including the meaning of the word.

Further, in this embodiment, the fixed phrase inquiry guidance data and the insufficient item phrase inquiry guidance data are output by voice, but the display unit is connected to the voice input determination unit 14 to display and output the guidance data on the screen. Even if comprised in this way, the same effect can be produced.

Further, in this embodiment, the ROM 142 is used.
The function as the voice input device is realized by the CPU 141 executing the voice input program stored in the HDD.
Download the voice input program to D device 145,
The function as the voice input device may be realized by the CPU 141 executing this program.

[0069]

As described above in detail, with the voice input device, voice input method, and voice input program of the present invention, the order contents can be correctly input and processed without being caught in the order of voice input. In addition to improving the usability of the input device, anyone can easily and correctly input the order contents by voice.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a voice input device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a voice input confirming unit in FIG.

FIG. 3 is a view showing main data files and a data table stored in the HDD device of the voice input confirming unit.

FIG. 4 is a diagram showing a record configuration of each data file shown in FIG.

5 is a diagram showing a data configuration of the data table shown in FIG.

FIG. 6 is a diagram showing a main memory area formed in a RAM of the voice input determination unit.

FIG. 7 is a flowchart showing a procedure of main program processing executed by a CPU of the voice input confirming unit.

FIG. 8 is a diagram showing an example of data in an example of the operation of the voice input confirming unit.

FIG. 9 is a diagram showing an example of data in another example of the operation of the voice input confirming unit.

[Explanation of symbols]

1 ... Voice input device 2. Voucher issuing device 11 ... Microphone 12 ... Speaker 13 ... Voice engine section 14 ... Voice input confirming section 15 ... Voice input switch 31 ... Vocabulary file 32 ... dictionary file 33 ... Guidance file 34 ... Confirmation condition table

Front page continuation (51) Int.Cl. ⁷ Identification code FI theme code (reference) G10L 15/10 G10L 3/00 571K 15/24 571R (72) Inventor Nobuo Ishijima 3-21, Nihonbashihamacho, Chuo-ku, Tokyo No. 1 Nihonbashi F Tower Toshiba Tech Co., Ltd. F term (reference) 5B091 AA11 CA12 CB12 5D015 KK02 KK04 LL06 5D045 AB30

Claims (13)

[Claims] 1. A phrase extracting means for extracting pre-registered phrases one by one from a sentence input by voice, and an item including a meaning of the phrase is determined for each phrase extracted by the phrase extracting means. The phrase item determining means, the phrase storing means for storing each phrase for which an item is determined by this phrase item determining means, and the item combination pattern of each phrase stored by this phrase storing means satisfy a predetermined confirmation condition. Determining condition determining means for determining whether or not a pattern to be performed, and when the determining condition determining means determines that the item combination pattern of each word is a pattern satisfying the above-mentioned setting condition, it is stored by the word storing means. A voice input device, comprising: input confirming means for confirming each word or phrase as voice input data. 2. The definite condition determining means extracts a matching pattern from a plurality of item combination patterns preset based on the item of each word stored by the word storing means, and the extracted pattern 2. The voice input device according to claim 1, wherein the voice input device is means for determining whether or not the pattern satisfies a predetermined confirmation condition. 3. The term having the meaning of an item that is insufficient to satisfy the finalizing condition when the finalizing condition determining unit determines that the item combination pattern of each word is not a pattern that satisfies the finalizing condition. Further comprising a deficient word urging means for urging the re-input of the voice, the input confirming means, the item including the meaning of the word extracted by the word extracting means after urging the voice re-input by the insufficient word reminder means 3. The voice according to claim 1, wherein when the item is insufficient to satisfy the confirmation condition, this phrase and the phrase stored by the phrase storage unit are confirmed as voice input data. Input device. 4. When the word extracted by the word extracting means includes an uncertain word, the word storing means further comprises a fixed word urging means for urging the user to re-input a voice for fixing the word. 4. The method according to any one of claims 1 to 3, wherein the fixed phrase prompting means stores the phrase extracted by the phrase extracting means after prompting a voice re-input in place of the indeterminate phrase. Voice input device. 5. When a plurality of words having the same item meaning are included in the words extracted by the word extracting means, the word most recently input by voice is stored in the word storing means. The voice input device according to claim 1. 6. A step of extracting pre-registered words one by one from a voice input sentence, a step of determining an item including a meaning of the word for each extracted word, and the extracted word A step of determining whether or not the item combination pattern of each word and phrase satisfies a predetermined definite condition; and the item combination pattern of each extracted word and phrase is determined to be a pattern that satisfies the definite condition. And a step of determining each word as voice input data. 7. The voice input method according to claim 6, wherein the determining step extracts a matching pattern from a plurality of item combination patterns preset based on the items of the extracted words and phrases. A voice input method characterized in that it is judged whether or not the extracted pattern satisfies a predetermined confirmation condition. 8. The voice input method according to claim 6 or 7, wherein when it is determined that the extracted item combination pattern of each word and phrase does not satisfy the confirmation condition, the determination condition is satisfied. A voice input method characterized by adding a step of prompting for voice re-input of a word having a meaning of a missing item. 9. The voice input method according to claim 6, wherein when an uncertain phrase is included in the extracted phrases, a step of prompting re-input of a voice for fixing the phrase. A voice input method characterized by the addition of. 10. A function having a computer having a voice input function for extracting pre-registered phrases from a voice-input sentence one by one, and determining for each extracted phrase an item including the meaning of the phrase. A function to determine whether the extracted item combination pattern of each word or phrase is a pattern that satisfies a predetermined confirmation condition, and a pattern that the extracted item combination pattern of each word or phrase satisfies the determination condition A voice input program that realizes a function of determining each of the phrases as voice input data when it is determined that 11. The voice input program according to claim 10, wherein the determination function extracts a matching pattern from a plurality of item combination patterns preset based on the items of the extracted words and phrases. A voice input program having a function of determining whether or not the extracted pattern is a pattern that satisfies a predetermined confirmation condition. 12. The voice input program according to claim 10, wherein when the computer determines that the extracted item combination pattern of each word or phrase is not a pattern that satisfies the determination condition, the determination condition is set. A voice input program that further realizes a function of prompting voice re-input of a phrase having a meaning of an item that is insufficient to satisfy. 13. The voice input program according to claim 10, wherein when the computer includes an uncertain word or phrase in the extracted word or phrase, a voice is re-established to fix the word or phrase. A voice input program that further realizes the function of prompting for input.