JPS60246468A - Information input/output device - Google Patents

Abstract

PURPOSE:To expand a leading-in range of an information input/output device for executing services of information without lowering the services to a user by constituting said device so that the user can select an interaction to a machine or an interaction to an operator. CONSTITUTION:An input/output terminal equipment 1 can input information by various means such as a voice input, etc., and it is digitized and sent out to a circuit 2. An interface 3 unifies and processes information such as a voice, a character picture, etc. from the circuit 2, and converts the information so that it is suitable to each interaction of a user and a machine, and the user and an operator. When the other party of an interaction is designated by a means such as a voice, a key input or a character recognition, etc. from the user, a computer/operator switching device 4 is controlled through the circuit 2 and the interface 3, and one of a computer 5 and an input/output terminal equipment 6 for operator is connected to the circuit 2 through the interface 3.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to an information input/output device that inputs and outputs information using voices, characters, gaps, etc. as media, and allows a human and a machine (computer) to interact. Regarding.

[Technical background of the invention and its problems]

In the New Tade 4F era, information services such as CATV and Videotex (Captain System) are becoming increasingly important. Also, LSI technology, high performance, low cost, advances in semiconductor technology, lower price of computers, and higher performance optical communication technology.

Humans and machines (computers) can interact using artificial satellites, pattern recognition technology, voice recognition technology, information retrieval technology, advances in knowledge engineering, large-scale databases, text, voice, and other input/output media. There is a growing need for dialogue systems (man-machine systems) that provide information services. for example.

In order to reduce labor costs, increase profits and sales, improve services, and improve economic efficiency, we offer 0 telephone services, 1 telephone services, JNR telephone services, airline tickets, movie seat reservations, restaurant reservations, play guides, etc.

However, compared to natural conversations between humans and human voices, interactions between humans and machines (computers) are currently unnatural and cumbersome, and the methods of using information input/output devices (such as key operations)
It is difficult to remember.

However, the scope of application of man-machine systems between humans and machines is expanding in order to simplify, reduce costs, improve economic efficiency, and increase profits and sales.

If the user's input method or input information is incorrect, the computer in the information input/output device may not be able to fully understand and respond.

There are many problems, especially when the user is inexperienced or when the information input/output device is first introduced.

In addition, owners of information input/output devices that provide information services,
In other words, travel agencies, play guides, information service companies, and telephone shopping companies want to improve their sales and services.

That is, there is a strong demand to provide services through natural human-to-human voice conversations, secure users and customers, and increase profits.

If the interaction between humans and machines (computers) is not operated properly, users may use the device incorrectly or do not know how to use it, and in the case of voice input or online character input, voice recognition devices and There has been a problem in that the recognition rate of the character recognition device is poor, resulting in incorrect information input.

(W1 essential point of the invention) As mentioned above, if the interaction between humans and machines (computers) does not go smoothly, and when the user provides the information service, the user can use the operator's real voice to communicate with the user and provide the service. If the user wishes to receive such information, please indicate that the user wishes to interact with the operator,
An object of the present invention is to provide an information input/output device having means for notifying an operator.

〔Effect of the invention〕

The user can choose to interact with the machine or with the operator, so if the user is inexperienced or the information input/output is not going smoothly, the user can choose to interact with the operator.
Information can be input and output smoothly. For this reason, the range of introduction of information input/output devices that provide information services can be expanded without causing a decline in service to users. and,
Users can select the machine or operator as appropriate, so they can use the information input/output device with peace of mind and receive information services.

[Embodiments of the invention]

Embodiment 0 FIG. 1 shows a block diagram of an information input/output device between humans and machines according to the present invention.

(Speech, images, and characters are used as output when Sample A/
The signal is D-converted, 8-bit logarithmically compressed and quantized, and converted into a digital signal.

Such an LSI for CODEC has already been commercialized, and the device of this embodiment is used for audio encoding.

0 keyboard (allows you to enter alphabets, numbers, kana, and symbols.

) The encoded 8-bit digital signal of θ audio is converted to D/
A conversion and convert to analog signal.

Output audio from the speaker.

(Some of the LSIs in ■ already include D/A functions.) O character 2 image display display (CRT display, liquid crystal display, etc.) When it comes to voice recognition response systems.

Line 2 A digital line that transmits digital signals of text, images, and audio information.

Interface 30 Integrates and handles information from the user such as voice, text and drawings, and interacts with the user and the machine (calculator) and between the user and the operator. Perform the conversion.

When the output of 0 people uses voice as a medium and the dialogue with a machine (computer) is performed using a voice recognition device and a voice rule synthesis device, the contents of the interface in step 3 will be as shown in Figure 2.

Speech recognition device 7 (Word speech recognition device for unspecified speakers that can recognize speech signals such as keywords and numbers for information retrieval input from the line) Speech rule synthesis bag rt8 (Character information from a computer ( language information) into audio signals.

For example, this can be done in units of syllables using LSP parameters and cepstrums.) The interface 3 is for combining the information input and output from the line 2 from the user side with the information from the computer 5 and the operator side. .

When there is a request from the user side to switch to dialogue with the operator using keystrokes or voice input, switching information is extracted by the interface 3, and the dialogue from the computer 5 is switched to the dialogue with the operator. In the case of a system that does not have a voice recognition function, the switching is performed by the user's key power. - Knowledge is stored in the database, and the system performs processing based on the input information from the user and outputs appropriate information. Output.

For example, an information service system that provides play guides uses input information from users and desired search fields (movies, concerts, plays, etc.).

Sports...) Desired event (baseball, soccer, rugby.

Pro-wrestling, etc.) Desired date: May 15th, etc. The user inputs information using a non-voice input such as a key or by voice, and the computer processes it and sends the appropriate information to the user. Output to.

The output information at this time may be configured to include non-audio information such as characters and images, or audio.

Operator input/output terminal device 6 Allows the operator to communicate with the user using voice.
It is equipped with a microphone and a speaker, and can monitor the interaction between the user and the computer as text or image information on a display such as a CRT.

In the device of this embodiment having the above configuration, when switching to the following, the following system operation mode can be considered.

(a) What 11 → Guema I7 When a user and an operator simply have a conversation, the operator interacts with the e-shiba computer, acts as an intermediary with the user, and notifies the user of information by voice or the like.

of users and operators Switch to interactive mode Calculator? The output of computer terminal, or Output to user's terminal It will be done.

☆The following variations exist depending on the input/output means (media).

δIf all input and output are audio The user input/output terminal 1 in Figure 1 is a telephone. Line 2 is a public phone line Interface 3 is the same as Hemp 2 diagram. Speech recognition device and speech rule synthesis device The switching device 4 obtains information based on the voice recognition result. This is the information provided by the operator, etc. Detect the word and use the calculator and Select one of the following.

In this case, since the input is based on speech pattern recognition, normal data input/output between the user and the computer may not be carried out smoothly due to the user's personality, speaking style, noise, or poor line quality. In such cases, words that are easy to recognize, such as "
Since the user can switch to a conversation with the operator by saying something like "Operator," the drawbacks of the current telephone service, which is mainly used for balance inquiries, etc., can be overcome. In other words, voice recognition technology is insufficient,
When a user is unable to input information, it can be backed up by a phone call with an operator, which has the advantage of being able to introduce such an automated system in many fields, reducing personnel costs and improving economic efficiency. , compared to manual services.

[Phase] When voice is not included in the interaction between humans and machines, the input/output terminal 1 in Figure 1 is a key input device and a microphone, and the output is a CRT or facsimile speaker line 2 is a text input device and a microphone. A digital line that can transmit a single screen and voice in a unified manner.

Interface 3 is a voice recognition device and does not include synthesis equipment. stomach. Operator and user Audio is only used during conversations. be exposed.

If the user does not know how to use keys, figures, strokes, or text information even after inputting information, or wants to obtain another type of information, the user can press a special key such as "alternate". This allows you to request interaction with the operator. For this reason, a large number of operators are required at the time of introduction, but this can be reduced over time.

■It is also applicable to input using character recognition (online recognition).

There are two types of modes in Honkomei.

Among these, ROBU i-m performs time-sharing processing, and can target many lines and interact with many users using one processing device. However, Country 11 ++ Figure=7=7
Since this is a conversation between people, it is impossible to have many conversations at the same time.

Therefore, if there are too many requests from users to perform conversation processing with the operator (conversation is not possible), the following response may be considered.

(a) A message is output to the user saying, "The operator is currently unavailable, so please communicate with the computer." (From the computer) (b) [To all of you, the operator will be in touch with you next time." I send out a message. (from a computer) tc) Like modern telephones, the - operator interrupts the conversation with the user and tells the user that they cannot keep their hands off them. Operators treat when hands are on.

1(1) The system device according to the present invention records the user number, and the operator will contact you later.

The above method can be considered as a countermeasure when the operator is busy.

☆As described above, the present invention alleviates the deficiencies in conventional man-machine systems and is based on the user's perspective.While promoting interaction between humans and machines, the present invention You can switch to dialogue with
It can be applied to information input/output systems that provide various information services.

(Can back up voice recognition equipment (can compensate for incomplete natural language processing technology)

[Brief explanation of drawings]

FIG. 1 is a diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing an example of the configuration of an interface 3 in an embodiment of the present invention, and FIGS. 3 and 4 explain the dialogue system of the present invention. FIG. 5 is a diagram showing a processing procedure of an embodiment of the present invention. 1... Input/output terminal device, 2... Line, 3... Interface, 4... Computer/operator switching device, 5
... Total of X machines, 6 ... Input/output terminal equipment for operators. Representative Patent Attorney: Nori Chika, Right (and 1 other person) Figure 1: Operation Figure 2 Figure 8 Figure 4 Operation Figure 5

Claims (3)

[Claims] (1) In an information input/output device having means for a user to input information, and means for processing the inputted information and outputting information to the user, the user and the information input/output device Information characterized by comprising means for interacting with an operator, and means for requesting and notifying the operator that the user interacts with the operator while the user is using the information input/output device. Input/output device. (2) The information input/output device according to claim 1, wherein the dialogue between the user and the operator is performed by voice. (3) The information input/output device according to claim 2, wherein information input/output between the user and the computer of the device is performed using a speech recognition device and a speech synthesis device.