JP3247943U - Ticket machine - Google Patents

Abstract

[Problem] To provide a ticket vending machine that allows the user to easily confirm that information entered by voice input has been accepted, and allows the user to purchase a ticket in a manner similar to purchasing a ticket while talking to a station attendant. [Solution] In a ticket vending machine (1), when an extracted word, which is information entered by a user through speech operation, is accepted by a reception unit (19), a word indicating that the reception unit (19) has accepted the extracted word, for example, a sentence such as "We have received your request for XX", is output as a voice from a speaker (20) as a response phrase. Therefore, the user can easily confirm that the information entered by voice input (extracted word) has been accepted by the response phrase. In addition, the user's speech and the response phrase output as a voice from the speaker (20) in accordance with the content of the speech allow the user to purchase a ticket in a manner similar to purchasing a ticket while talking to a station attendant. [Selected Figure] Figure 2

Description

This invention relates to a ticket vending machine.

In recent years, there have been proposals for automatic ticket vending machines that sell tickets such as passenger tickets and commuter passes, which allow users to purchase tickets by inputting the necessary information (such as information about the name of the destination station) by voice input, in addition to operating mechanical or touch panel buttons (see, for example, Patent Document 1).

JP 2008-204351 A

When a voice including the name of the destination station is input into the ticket vending machine described in Patent Document 1, text information is extracted from the voice, and then the name of the destination station is extracted from the extracted text information. After that, ticket issuance information is read out based on the extracted station name, and the read out ticket issuance information is given to a screen control unit that controls the screen display, and this screen control unit changes the screen display from the standard screen to a ticket issuance information confirmation screen. Therefore, the user can confirm the ticket issuance information input by voice input on the confirmation screen.

However, with the ticket vending machine described in Patent Document 1, although the user can check on the confirmation screen the updates to the ticket issuing information based on the user's voice input (speech operation), if the change in the displayed content before and after the screen transition is small, it can be difficult to tell by visual confirmation alone whether the spoken content has been correctly recognized. In addition, when the user makes voice input (speech operation), the user speaks unilaterally to the ticket vending machine, which makes it difficult to feel like they are having a conversation compared to purchasing a ticket while talking to a station attendant at the station counter, and the ticket vending machine's service can seem cold. This may be a factor that makes users hesitant to purchase tickets by voice input.

Therefore, there is a demand for a ticket vending machine that allows users to easily confirm that information entered by voice has been accepted by means other than visual inspection, and that allows users to purchase tickets in a manner similar to purchasing a ticket while speaking with an attendant.

The purpose of this invention is to provide a ticket vending machine that allows users to easily confirm that information entered by voice has been accepted, and allows them to purchase tickets in a manner similar to purchasing a ticket while speaking with an attendant.

The ticket vending machine of the present invention is a ticket vending machine capable of issuing tickets in response to the content of a user's speech, and is equipped with a voice data generation unit that generates voice data from the speech collected by a sound collection means, a conversion unit that converts the voice data into text data, an extraction unit that extracts words that match predetermined conditions from the text data, a reception unit that receives extracted words that are the words extracted by the extraction unit, and a voice generation output unit that voice-synthesizes a response phrase that is a word indicating that the reception unit has received the extracted word, and outputs it as voice from a voice output means, and is characterized in that the response phrase that is voice-output from the voice output means allows the reception unit to confirm that the extracted word has been received.

The ticket vending machine of the present invention is characterized in that the voice generation output unit has a sentence generation means for generating a sentence including the extracted word, and the sentence generated by the sentence generation means is used as the response phrase.

The ticket vending machine of the present invention is characterized in that when multiple extracted words are received by the reception unit, the voice generation output unit uses only one generic phrase as the response phrase.

The ticket vending machine of the present invention is capable of issuing tickets according to the contents of the user's speech, and is equipped with a voice data generation unit that generates voice data from the speech collected by the sound collection means, a conversion unit that converts the voice data into text data, an extraction unit that extracts words that match a predetermined condition from the text data, a reception unit that accepts the extracted words that are the words extracted by the extraction unit, and a voice generation output unit that synthesizes a response phrase that is a word that indicates that the reception unit has accepted the extracted word and outputs it as a voice from the voice output means, and the response phrase that is output as a voice from the voice output means can be used to confirm that the extracted word has been accepted by the reception unit. Therefore, it is easy to confirm that the information (extracted word) input by voice input has been accepted by the response phrase. In addition, the user's speech and the response phrase that is output as a voice from the voice output means allow the user to purchase a ticket with a feeling similar to that when purchasing a ticket while talking to an attendant. Furthermore, the response phrase allows the user to easily grasp the update status of the information, so there is no need to worry about whether the information has been accepted properly, and voice input (speech operation) can be performed with confidence.

In the ticket vending machine of the present invention, the voice generation output unit has a sentence generation means for generating a sentence including the extracted word, and the sentence generated by the sentence generation means is used as a response phrase. In this way, since the sentence including the extracted word is output as a response phrase, the accepted extracted word can be more clearly communicated to the user than when a response phrase not including the extracted word is output as a response phrase.

In the ticket vending machine of the present invention, when multiple extracted words are received by the reception unit, the voice generation and output unit outputs only one generic phrase as a response phrase. In this way, even if multiple extracted words are received by the reception unit, multiple response phrases are not output in voice corresponding to the multiple extracted words, so the user can smoothly purchase a ticket without having to wait for the end of the voice output.

FIG. 2 is a front view of a ticket vending machine according to an embodiment of the present invention. 1 is a block diagram showing an outline of the internal configuration of a ticket vending machine according to an embodiment of the present invention. FIG. 1 is an explanatory diagram for explaining an outline of the screen transitions and voice output of the display means in a ticket vending machine according to an embodiment of the present invention, where (a) is a diagram showing the screen before voice input, (b) is a diagram showing an image of the screen and voice output after voice input, and (c) is a diagram showing a confirmation screen for confirming ticket issuance information including information entered by voice input. 4 is a flow chart showing the flow of a recitation process in a ticket vending machine according to an embodiment of the present invention. FIG. FIG. 11 is an explanatory diagram showing the relationship between the display means and the recitation process in a ticket vending machine according to an embodiment of the present invention, and is an explanatory diagram showing the case where the AI automatic response unit asks a question to update each item of the ticket issuing information. FIG. 10 is an explanatory diagram showing the relationship between the display means and the recitation process in a ticket vending machine according to an embodiment of the present invention, and is an explanatory diagram for the case where the AI automatic response unit asks a question to confirm certain information.

The ticket vending machine 1 according to the embodiment of the present invention will be described below with reference to Figs. 1 to 6. In Fig. 1, the width of the ticket vending machine 1 is the left-right direction X, and the height of the ticket vending machine 1 is the up-down direction Y. In the following description, the ticket vending machine 1 will be used to issue train tickets, among various types of tickets. However, the ticket vending machine 1 may also issue tickets for transportation other than trains, such as buses and ships, or tickets for entertainment facilities other than transportation, such as movie theaters and theme parks, or meal tickets at restaurants.

As shown in FIG. 1, the ticket vending machine 1 is equipped with a touch panel display 2 that displays the ticket issuing information required to purchase a ticket and allows the user to input information by touching the display. This display 2 corresponds to a display means. Although not shown in detail, the standby screen of the display 2 displays multiple menu buttons 2a such as "Reserved seat", "Unreserved seat", "Pick up reserved ticket", "Purchase ticket by time search", "Commuter pass" and "Call call center". When the user touches the menu button 2a corresponding to the desired operation, the operation for purchasing a ticket can be started.

At the top of the ticket vending machine 1, above the display 2, there are provided a monitor 3 that can display the operator 30 who will respond when a call is made to the call center, a camera 4 that can capture video so that the operator 30 can check the state of the customer, a microphone 5 that is a sound collection means for collecting speech from the customer, and a speaker 20 that is a sound output means that outputs speech from the operator 30 and sound via a sound generation and output unit 21, which will be described later. Another speaker 20 that is a sound output means is provided to the right of the display 2. A handset 6 is provided to the left of the display 2, and this handset 6 can also be used to talk to the operator 30.

Provided at the bottom of the ticket vending machine 1, below the display 2, are a coin slot 7 for inserting coins, a bill insertion slot 8 for inserting bills, a bill return slot 9 for returning bill change, and a coin return slot 10 for returning coin change. Also, provided to the left of the bill insertion slot 8 is a card insertion slot 11 for inserting cards such as credit cards and IC commuter passes, and provided to the left of the bill return slot 9 is a ticket discharge slot 12 for discharging tickets.
In addition, below the coin slot 7, there is provided a reading device 13 capable of reading two-dimensional codes displayed on a mobile device such as a smartphone owned by the user, or two-dimensional codes printed on documents such as paper, and a numeric keypad 14 capable of inputting numbers such as a credit card PIN number.

As shown in FIG. 2, the ticket vending machine 1 is provided with, inside its housing, an AI automatic response unit 15 capable of automatically responding to operations performed by a user, a voice data generation unit 16 that generates voice data from speech collected by a microphone 5, which is a sound collection means, a conversion unit 17 that converts the voice data into text data, an extraction unit 18 that extracts words that match predetermined conditions from the text data, a reception unit 19 that accepts extracted words that are words extracted by the extraction unit 18, a voice generation output unit 21 that voice-synthesizes a response phrase, which is a word indicating that the reception unit 19 has accepted the extracted word, and outputs it as voice from a speaker 20, which is a voice output means, a display unit 22 that controls the display content of the display 2, which is a display means, a communication unit 23 that is capable of communicating with external devices such as a server (not shown) installed outside the ticket vending machine 1, and a control unit 24 that controls the functions and various devices installed in the ticket vending machine 1. Inside the housing of the ticket vending machine 1, various devices and means necessary for selling tickets are also provided, such as a currency identification device 25 that identifies the authenticity and type of inserted coins and inserted bills, a fare calculation unit 26 that calculates the fare for the travel section, and a ticket issuing device 27 that issues tickets on which necessary information is recorded in response to the payment of the fare. Detailed explanations of the configurations of the currency identification device 25, the fare calculation unit 26, the ticket issuing device 27, etc. will be omitted.

In the ticket vending machine 1 with the above configuration, a user can purchase a ticket by inputting the necessary information by operating the display 2 or by voice input from the microphone 5, and paying the specified fee using a payment method such as cash, credit card, or code payment via smartphone. The ticket vending machine 1 also has an AI automatic response unit 15 that enables automatic response by AI. Therefore, the user can purchase the desired ticket by performing various operations following the guidance (AI automatic response) from the AI automatic response unit 15. Note that if the user is still unsure how to operate the ticket vending machine 1 even after following the guidance from the AI automatic response, the user can call a call center and purchase the ticket while inquiring about the operation method from an operator 30.

When information is input by touch operation on the display 2 or voice input from the microphone 5, the AI automatic response unit 15 guides the user on the next operation or confirmation of the input information (AI automatic response) in accordance with the input information. The AI automatic response unit 15 guides the user on the general operations required when purchasing a ticket. Specifically, for example, the display content of the display 2 is changed via the display unit 22, or voice is output from the speaker 20 via the voice generation output unit 21, to prompt the user to input necessary or missing information, or to ask questions to confirm certain information. Therefore, in the ticket vending machine 1, the user can purchase the desired ticket by performing the operations guided by the AI automatic response while checking the display content displayed on the display 2 and the voice output from the speaker 20.

For example, when a user touches the "unreserved seat" button among the menu buttons 2a displayed on the standby display 2 of the ticket vending machine 1 installed at Osaka Station, the AI automatic response unit 15 transitions the screen displayed on the display 2 via the display unit 22 to display the main ticket issuing information as shown in FIG. 3(a). The main ticket issuing information items include the "departure station" and "arrival station" indicating the boarding section, the "date and time" indicating the departure date and time, and the "number of passengers" indicating the number of passengers. Note that, as initial settings for each item of the main ticket issuing information, for example, the "departure station" may display "Osaka", which is the name of the station where the ticket vending machine 1 is installed, the "date and time" may display the departure date and time which is the date and time a predetermined time after the current date and time (e.g., 15 minutes later), and the "number of passengers" may display one adult.

Here, since the "arrival station" information is missing from the main ticket issuing information required when a user purchases a ticket, the AI automatic response unit 15 displays the question "Welcome. Where are you going?" in the eye-catching display section 2b of the display 2. Then, after checking the guidance in the eye-catching display section 2b (display of "Welcome. Where are you going?"), the user can input information about the "arrival station" by touching the display 2 or by voice input (speech operation) through speech. Note that the content of the guidance provided by the AI automatic response unit 15 may be output as voice from the speaker 20 by the voice generation output unit 21 in addition to being displayed in the eye-catching display section 2b.

In order to respond to information input by the user, the AI automatic response unit 15 may be provided with a database in which response contents corresponding to the input information are pre-recorded, and when information is input by the user, the AI automatic response unit 15 may be configured to refer to the response contents recorded in the database and change the display contents on the display 2 or output voice according to the response contents, or may be configured to accumulate and record the input information and the guidance by the AI automatic response unit 15, and optimize the response contents through repeated learning based on the accumulated records.

The voice data generating unit 16 generates voice data by collecting speech by a user with the microphone 5. For example, assume that the "unreserved seat" button is touched on the ticket vending machine 1 installed at Osaka Station, and the AI automatic response unit 15 displays the question "Welcome, where are you going?" on the eye-catching display unit 2b (FIG. 3(a)), and the user utters "To Tokyo." Then, the speech of "To Tokyo" is collected by the microphone 5, and voice data is generated by the voice data generating unit 16 from the collected speech.
In addition, the voice data generation unit 16 may be configured to start collecting sound when the menu button 2a displayed on the display 2 is operated, or may be configured to include a detection device that detects speech by the user and to start collecting sound when this detection device detects speech.

The conversion unit 17 performs voice recognition based on the voice data generated by the voice data generation unit 16, and converts the voice data into text data. Specifically, the conversion unit 17 converts the voice data generated by the voice data generation unit 16 into text data by performing processes such as acoustic analysis and phoneme identification, and converting the identified phonemes into words.

The extraction unit 18 extracts words that meet a predetermined condition from the text data converted by the conversion unit 17. The predetermined condition is a condition for obtaining ticketing information required when the user purchases a ticket, and is set according to the content of the guidance that the AI automatic response unit 15 provides to the user. For example, when the AI automatic response unit 15 is guiding the user to the question "Welcome, where are you going?" described above, the predetermined condition is set to "existing station name" in order to obtain the name of the destination station as an answer from the content of the utterance. Then, "Tokyo", which is a word that meets the condition, is extracted from the content of the user's utterance that has been converted into text data (the content of the utterance "to Tokyo" in the above example). In this case, "Tokyo" is the extracted word.

In addition, for example, when the AI automatic response unit 15 guides the user to the question "Is the ticket one-way?" to confirm whether the ticket purchase request is for a one-way trip, the extraction unit 18 sets "one-way", "round trip", "no need" and "no ticket" as predetermined conditions to obtain an answer indicating either "one-way" indicating a one-way ticket, "round trip" indicating a round trip ticket, or "no ticket required" indicating that a ticket is not required from the content of the utterance. Then, for example, if the user utters "one-way" in response to the question, "one-way", which is a word matching the condition, is extracted from the content of the user's utterance converted into text data (the content of the utterance of "one-way"). In this case, "one-way" is the extracted word.
If the extraction unit 18 is able to extract an extracted word, it notifies the reception unit 19 of the extracted word, and if the extraction unit 18 is unable to extract an extracted word, it notifies the reception unit 19 of that effect.

The reception unit 19 receives the extracted word extracted by the extraction unit 18, and notifies the AI automatic response unit 15, the voice generation output unit 21, the display unit 22, the fare calculation unit 26, the ticket issuing device 27, etc. of the extracted word or information corresponding to the extracted word, and updates the ticket issuing information. Specifically, for example, if the extracted word is "Tokyo" indicating the name of the destination station, the AI automatic response unit 15, the voice generation output unit 21, the display unit 22, the fare calculation unit 26, the ticket issuing device 27, etc. of the extracted word, and updates the "arrival station" item in the ticket issuing information. Also, for example, if the extracted word is "one way" indicating a one-way ticket, the voice generation output unit 21 and the display unit 22 are notified of the extracted word, and the AI automatic response unit 15, the fare calculation unit 26, the ticket issuing device 27, etc. are notified of the desire to purchase a one-way ticket as information corresponding to the extracted word, and the ticket issuing information is updated.

The voice generation output unit 21 synthesizes a response phrase, which is a word indicating that the reception unit 19 has received the extracted word, and outputs the response phrase as a voice from the speaker 20, which is a voice output unit. The voice generation output unit 21 also has a sentence generation unit 21a (FIG. 2) that can generate a sentence including the extracted word, and can output the sentence generated by the sentence generation unit 21a as a response phrase from the speaker 20. Specifically, when "Tokyo" is received by the reception unit 19 as the extracted word, the sentence generation unit 21a generates "We received it in Tokyo" as a sentence including "Tokyo". Then, the voice generation output unit 21 synthesizes the sentence "We received it in Tokyo" as a response phrase and outputs it as a voice from the speaker 20, as shown in FIG. 3B.
Also, for example, when the reception unit 19 receives the extracted word "one way", the sentence generation means 21a generates "We received it in one way" as a sentence including "one way", and the voice generation output unit 21 synthesizes the sentence "We received it in one way" as a response phrase and outputs it as voice from the speaker 20.

Based on notifications from the AI automatic response unit 15 and the reception unit 19, the display unit 22 controls the display content of the display 2, which is a display means, and updates or changes the screen displayed on the display 2. Specifically, when the reception unit 19 receives "Tokyo" as an extracted word, the display unit 22 displays the sentence "Accepted in Tokyo" in the eye-catching display unit 2b of the display 2, similar to the sentence "Accepted in Tokyo" generated by the sentence generation means 21a, and displays the extracted word "Tokyo" accepted by the reception unit 19 in the "Arrival Station" item of the ticket issuance information (FIG. 3(b)).

In addition, after a predetermined time has elapsed for the user to check the display on display 2 and the voice output by the voice generation output unit 21, the display unit 22, upon receiving a notification from the AI automatic response unit 15, transitions the display screen of the display 2 to the screen shown in Figure 3 (c) and displays a confirmation message in the eye-catching display unit 2b saying "From Osaka to Tokyo, departing at 20:45 on March 14th, is it OK for one adult to travel?"
Thereafter, when the user touches the “OK” button displayed on the display 2, the fare is calculated by the fare calculation unit 26, and the ticket issued by the ticket issuing device 27 upon settlement of the fare is discharged from the ticket discharge port 12.

The communication unit 23 communicates with an external device such as a server installed outside the ticket vending machine 1. For example, when purchasing a ticket that requires confirmation of available seats, such as a reserved seat on a Shinkansen train, the communication unit 23 communicates with a reservation sales management center (not shown) that manages seat reservations and sales, and inquires about available seats. Note that in this embodiment, an example is shown in which the ticket vending machine 1 is equipped with the AI automatic response unit 15, but for example, the AI automatic response unit 15 may be provided in an external device such as a server connected to a network, and the server and the ticket vending machine 1 may communicate with each other via the communication unit 23 to operate the AI automatic response unit 15 and realize AI automatic response in the ticket vending machine 1.

The control unit 24 controls the above-mentioned functions and various devices in the ticket vending machine 1. The control unit 24 has a processor such as a CPU (Central Processing Unit) that performs various processes according to a program, and controls the AI automatic response unit 15, voice data generation unit 16, conversion unit 17, extraction unit 18, reception unit 19, voice generation output unit 21, display unit 22, and communication unit 23 by using an OS (Operating System) program or application program stored in a RAM (Random Access Memory) according to a program stored in a ROM (Read Only Memory), and also controls various devices, etc. required for selling tickets (for example, devices or means such as the currency identification device 25, fare calculation unit 26, and ticket issuing device 27).

The ticket vending machine 1 is configured as described above. Next, we will explain the processing flow in the ticket vending machine 1.

FIG. 4 is a flow diagram showing the flow of processing (repetition processing) from when voice input (utterance operation) is made in the ticket vending machine 1 to when a response phrase is voice output.
When a voice input (speech operation) is first made by the user (S101), the voice data generating unit 16 generates voice data from the voice input utterance (S102). After that, the voice data generated by the voice data generating unit 16 is converted into text data by the converting unit 17 (S103). Next, the extracting unit 18 extracts extracted words that match a predetermined condition according to the content of the guidance given to the user by the AI automatic response unit 15 from the text data converted by the converting unit 17 (S104). After that, the extracted words extracted by the extracting unit 18 are notified to the receiving unit 19, and the receiving unit 19 receives the extracted words (S105). Then, the extracted words received by the receiving unit 19 are notified to the voice generating output unit 21, etc., and the sentence generating means 21a in the voice generating output unit 21 generates sentences including the extracted words as "words (response phrases) indicating that the receiving unit 19 has received the extracted words" (S106). Next, the voice generating and outputting unit 21 synthesizes a sentence including the extracted word generated by the sentence generating means 21a as a response phrase and outputs the response phrase from the speaker 20 (S107).

Above, the flow of the recitation process in the ticket vending machine 1 has been explained. Next, the relationship between the screen of the eye-catching display section 2b of the display 2, which transitions based on the user's speech operation or touch operation, and the recitation process in the ticket vending machine 1 will be explained. Note that below, separate explanations will be given for cases where the AI automatic response unit 15 asks questions to update each item of the main ticket issuing information ("departure station", "arrival station", "date and time", and "number of people") displayed on the display 2 shown in Figures 3(a) to 3(c) in order to obtain the user's desired conditions as an answer, and cases where questions are asked to confirm specific information other than each of the items.

Figure 5 is an explanatory diagram showing a case where the AI automatic response unit 15 asks a question to update each item of the main ticket issuing information ("Departure station", "Arrival station", "Date and time", and "Number of people"). Specifically, Figure 5 shows a case where the AI automatic response unit 15 asks the question "Welcome. Where are you going?" in order to update the "Arrival station" item (obtain the name of the destination station) from among the main ticket issuing information in a ticket vending machine 1 installed at Osaka Station. First, the AI automatic response unit 15 displays "Welcome. Where are you going?" on the eye-catching display unit 2b via the display unit 22, and the voice data generation unit 16 starts collecting the user's speech using the microphone 5.

When the user speaks "To Tokyo" to perform voice input (speech operation), the process is carried out by the voice data generation unit 16, conversion unit 17, extraction unit 18, and reception unit 19, and the sentence "Accepted in Tokyo" is output by the voice generation output unit 21 as a response phrase. The AI automatic response unit 15 also transitions the screen of the eye-catching display unit 2b via the display unit 22, and the sentence "Accepted in Tokyo" is displayed in the eye-catching display unit 2b. Note that while the response phrase is being output by the voice generation output unit 21, sound collection (voice recognition) is stopped so that only touch operations by the user can be accepted and no new voice operations by the user are accepted.

After that, when a predetermined time has passed, the AI automatic response unit 15 changes the screen of the eye-catching display unit 2b via the display unit 22, and a confirmation message is displayed on the eye-catching display unit 2b saying, "Is it OK for one adult and one child to travel from Osaka to Tokyo, departing at 12:34 on December 25th?" At this time, the voice generation output unit 21 may be configured to generate a sentence including the extracted word (response phrase) such as "Is it OK for one adult and one child to travel from Osaka to Tokyo, departing at 12:34 on December 25th?" and output this sentence by voice. In this case, the voice generation output unit 21 may be configured to generate a sentence using a conjunction such as "Well then" so that the user feels that the AI automatic response by the ticket vending machine 1 and the voice output by the voice generation output unit 21 are similar to the response by a station attendant.

On the other hand, if the user touches the "Tokyo" button indicating Tokyo Station among the buttons indicating the names of destination stations displayed on the display 2, the AI automatic response unit 15 transitions the screen of the eye-catching display unit 2b via the display unit 22, and a confirmation message is displayed on the eye-catching display unit 2b saying "From Osaka to Tokyo, departing at 12:34 on December 25th, one adult and one child, is this OK?". In this case, as in the case of speech operation, the voice generation output unit 21 may be configured to generate the sentence "From Osaka to Tokyo, departing at 12:34 on December 25th, is this OK for one adult and one child?" and output this sentence by voice. Note that in this case, since it is a touch operation, the voice generation output unit 21 may be configured to generate a sentence without a conjunction such as "Well then" in order to distinguish it from the case of speech operation.

In addition, in FIG. 5, a case is shown in which one item (only "arrival station") is updated among the main ticket issuing information ("departure station", "arrival station", "date and time" and "number of people") in one speech operation by the user, but the extraction unit 18 may extract multiple extracted words in one speech operation, and the reception unit 19 may receive the multiple extracted words at once. For example, when the user utters "from Osaka to Tokyo" by a speech operation, the extraction unit 18 may extract the station name of the "departure station" from the utterance of "from XX" and the station name of the "arrival station" from the utterance of "to XX", and the reception unit 19 may receive these station names of the "departure station" and the "arrival station" at once. In that case, since multiple items of the main ticket issuing information are updated at once, if the voice output of a response phrase such as "We have received your order at XX" is repeated multiple times for each updated item, the user may have to wait a long time until the voice output is completed. Therefore, when two or more items are updated at once, the voice generation output unit 21 may generate and output only the word “I understand,” which is a general-purpose phrase used by station staff when accepting a request from a user, as a response phrase indicating that the reception unit 19 has received the extracted word.
In addition, even while the sentence "I understand" is being output by the voice generation output unit 21 as a response phrase, sound collection (voice recognition) may be stopped so that only touch operations by the user can be accepted and new speech operations by the user are not accepted.

Figure 6 is an explanatory diagram showing a case where the AI automatic response unit 15 asks a question to confirm predetermined information other than the main ticketing information ("departure station", "arrival station", "date and time", and "number of people"). Confirmation of predetermined information other than the main ticketing information includes, for example, confirmation of information on "seat facilities", "seat position", "whether or not discounts are applied", and "desired ticket purchase". Specifically, in the case of information on "seat facilities", confirmation is made that the seat facilities are either "ordinary car", "green car", "unreserved seat", or "no seat type preference", in the case of information on "seat position", confirmation is made that the seat position is either "window seat", "aisle seat", or "no seat position preference", in the case of information on "whether or not discounts are applied", confirmation is made that either "no discounts are used" or "discounts are used", and in the case of information on "desired ticket purchase", confirmation is made that the desired ticket purchase is either "one-way", "round trip", or "no ticket required".

Figure 6 shows a case where the AI automatic response unit 15 asks the question "Is the ticket a one-way ticket?" to obtain confirmation of whether the "desired ticket purchase" is a "one-way ticket" among other predetermined information other than the main ticket issuing information, on a ticket vending machine 1 installed at Osaka Station. First, the AI automatic response unit 15 displays the message "Is the ticket a one-way ticket?" on the eye-catching display unit 2b via the display unit 22, and the voice data generation unit 16 starts collecting the user's speech using the microphone 5.

When the user speaks "One way" to perform voice input (speech operation), the speech data generation unit 16, conversion unit 17, extraction unit 18, and reception unit 19 process the speech output unit 21 to output the sentence "One way confirmed" as a response phrase. The AI automatic response unit 15 also transitions the screen of the eye-catching display unit 2b via the display unit 22, and the sentence "One way confirmed" is displayed on the eye-catching display unit 2b. Note that while the response phrase is being output by the speech output unit 21, sound collection (voice recognition) is stopped so that only touch operations by the user can be accepted and no new speech operations by the user are accepted.

After that, when a predetermined time has passed, the AI automatic response unit 15 transitions the screen of the eye-catching display unit 2b via the display unit 22, and the eye-catching display unit 2b displays a confirmation message saying, "Are you sure you want a one-way ticket from Osaka to Kaihin Makuhari, December 25th, for one adult and one child, in a regular reserved seat?". At this time, the voice generation output unit 21 may be configured to generate a sentence, "Are you sure you want a one-way ticket from Osaka to Kaihin Makuhari, December 25th, for one adult and one child, in a regular reserved seat?" as a word (response phrase) indicating that the reception unit 19 has accepted the extracted word "one-way", and to output this sentence by voice. In this case, since this is a confirmation of specified information, the sentence generated by the voice generation output unit 21 may be a sentence without a conjunction such as "Well then", in order to distinguish it from a case where a question is asked to update each item of the main ticket issuance information.

On the other hand, if the user touches the "One Way" button, which indicates a one-way ticket, among the buttons displayed on the display 2 indicating "One Way," "Round Trip," and "No Ticket Required," the AI automatic response unit 15 transitions the screen of the eye-catching display unit 2b via the display unit 22, and a confirmation message is displayed on the eye-catching display unit 2b, saying, "Are you sure you want a one-way ticket from Osaka to Kaihin Makuhari, on December 25th, for one adult and one child, in a regular reserved seat?". In this case, as in the case of speech operation, the voice generation output unit 21 may be configured to generate the sentence "Are you sure you want a one-way ticket from Osaka to Kaihin Makuhari, on December 25th, for one adult and one child, in a regular reserved seat?" and output this sentence by voice.

Next, we will explain the function and effect of the ticket vending machine 1.

In the ticket vending machine 1, when the reception unit 19 receives an extracted word, which is information input by the user through speech operation (voice input), a word indicating that the reception unit 19 has received the extracted word, such as a sentence such as "I understand that it is ____" or "I understand," is output as a voice from the speaker 20 as a response phrase. Therefore, it is easy to confirm from the response phrase that the information input by voice input (extracted word) has been accepted. In addition, the user's speech and the response phrase output from the speaker 20 in response to the content of the speech allow the user to purchase a ticket with a feeling similar to that of purchasing a ticket while talking to a station attendant. Furthermore, because the response phrase makes it easy to grasp the update status of the information, there is no need to worry about whether the information has been accepted properly, and voice input (speech operation) can be performed with confidence.

In addition, in the ticket vending machine 1, a sentence including the extracted word, such as "We have received your order at XX," is generated by the sentence generation means 21a of the voice generation output unit 21, and the sentence is output as a voice response phrase. Therefore, the extracted word that has been received can be more clearly communicated to the user than when a response phrase that does not include the extracted word is output as a voice response.

Furthermore, in the ticket vending machine 1, when multiple extracted words are received by the reception unit 19, the voice generation output unit 21 can use only one general-purpose phrase, "Understood," as the response phrase. Therefore, even if multiple extracted words are received by the reception unit 19, multiple response phrases are not output in voice corresponding to the multiple extracted words, so the user can smoothly purchase a ticket without having to wait for the end of the voice output.

In the above explanation, we have described the case where the extraction unit 18 extracts extracted words. However, if words related to ticketing information such as the names of the departure or arrival stations, the departure time, the number of passengers, etc. cannot be accurately extracted, the response phrase from the voice generation output unit 21 may be configured to output a voice phrase encouraging the user to try again, such as "Please select a time after the current time."
In addition, when the user's speech input results in voice input of an utterance that is unlikely to be expected under normal circumstances, for example, when "10 adults" is input in response to a question about the number of passengers, the AI automatic response unit 15 is configured to prompt the user with a question to confirm whether the desired number of passengers on the ticket is 10 adults, and it is preferable to configure the voice generation output unit 21 to temporarily stop outputting the response phrase until an answer to the question is obtained when such guidance is given.

Although the embodiment of the present invention has been described above in detail, the present invention is not limited to the above embodiment. Furthermore, the present invention can be modified in various ways without departing from the matters described in the scope of the utility model registration claims.

For example, in the ticket vending machine 1 according to the present embodiment, an example has been shown in which sentences such as "We received it at XX" or "Understood" are output as voice phrases indicating that the reception unit 19 has received the extracted word, but the response phrase is arbitrary as long as the user can recognize that the reception unit 19 has received the extracted word. For example, sentences such as "We received it at XX" or "XX, right? Understood" may be used, or the voice generation output unit 21, which does not have a sentence generation means 21a, may output only preset standard phrases as response phrases. In addition, when there are multiple extracted words received by the reception unit 19, a sentence containing each extracted word may be generated corresponding to each extracted word, and all of the sentences may be output as voice. Furthermore, the generic phrase is not limited to "Understood", and may be a phrase such as "Understood" or "Got it".

1 Ticket vending machine 2 Display (display means)
2a Menu button 2b Eye-catching display section 3 Monitor 4 Camera 5 Microphone (sound collection means)
6 Receiver 7 Coin slot 8 Bill slot 9 Bill return slot 10 Coin return slot 11 Card slot 12 Ticket ejection slot 13 Reading device 14 Numeric keypad 15 AI automatic response unit 16 Voice data generation unit 17 Conversion unit 18 Extraction unit 19 Reception unit 20 Speaker (voice output means)
21 Voice generation output unit 21a Text generation means 22 Display unit 23 Communication unit 24 Control unit 25 Currency identification device 26 Fare calculation unit 27 Ticket issuing device 30 Operator X Left-right direction Y Up-down direction

Claims (3)

A ticket vending machine capable of issuing tickets in response to the content of a user's speech,
a voice data generating unit that generates voice data from the utterance collected by the voice collecting means;
A conversion unit that converts the voice data into text data;
an extraction unit that extracts words that meet a predetermined condition from the text data;
a receiving unit that receives the extracted words, which are the words extracted by the extraction unit;
a voice generating and outputting unit that synthesizes a response phrase, which is a word indicating that the receiving unit has received the extracted word, and outputs the response phrase as a voice from a voice output means,
The response phrase outputted by the voice output means can confirm that the extracted word has been received by the reception unit.
A ticket vending machine characterized by the above. the speech generation and output unit has a sentence generation means for generating a sentence including the extracted word, and the sentence generated by the sentence generation means is set as the response phrase.
2. The ticket vending machine according to claim 1 . the speech generation and output unit, when the reception unit receives a plurality of the extracted words, sets only one general phrase as the response phrase;
3. A ticket vending machine according to claim 1 or 2.

Images