KR20220154532A - Automatic call response system for orders and reservations - Google Patents

Abstract

The present invention relates to an automatic call response system for ordering and reservation. By extracting only the voice of a sentence including an entity from which order reservation information was derived and processing automatic call response, the loss of orders and reservations due to voice recognition errors is reduced, and time and manpower consumption for reconfirming consultation details is reduced. As an example, disclosed is an automatic call answering system for ordering and reservation, including an integrated management server unit which performs: receiving the voice query request data from the user communication terminal when the call is connected with the user communication terminal; inferring a query request intention for the voice query request data; extracting text answer data and entity information from a pre-established database based on the result of intention inference; and converting the extracted text response data into voice response data and transmitting it to the user communication terminal.

Description

Automatic call response system for ordering and reservation {AUTOMATIC CALL RESPONSE SYSTEM FOR ORDERS AND RESERVATIONS}

Embodiments of the present invention relate to an automatic call answering system for ordering and reservations, and more specifically, by utilizing an entity for a user's order and reservation details to easily check the order and reservation, automatically ordering and reservation It relates to a system that responds to calls for.

In an automatic call answering system, the most problematic factor for automatically responding to a customer's telephone inquiry and accepting an order or reservation is a reliability problem in that the customer's voice recognition rate cannot be recognized 100%. In this way, if an order or reservation is processed without reconfirmation in a situation where the customer's voice recognition is unreliable for automatic call response, a large loss is inevitable as incorrect information is recognized and processed.

Accordingly, in order to confirm the order or reservation again, it is necessary for the counselor to listen to the entire recorded file or call the customer again to confirm whether the order or reservation details are correct.

Accordingly, a new type of automatic call answering system is required to solve the loss due to voice recognition error, unnecessary manpower and time consumption due to the counselor's reconfirmation, and customer's inconvenience.

Patent Publication No. 10-2020-0092448 (published date: August 04, 2020) Patent Publication No. 10-2019-0061706 (published date: June 05, 2019)

An embodiment of the present invention extracts only the voice of a sentence including an entity from which order reservation information is derived and handles automatic call response, thereby reducing loss of orders and reservations due to voice recognition errors, and reconfirming consultation details. It provides an automatic call response system for ordering and reservations that can minimize time and manpower consumption for

An automatic call answering system for ordering and reservation according to an embodiment of the present invention is connected to a user communication terminal, receives the voice query request data from the user communication terminal, and determines the intention of the inquiry request for the voice query request data. Inference, extract text answer data and entity information from a pre-established database based on the result of intention inference, convert the extracted text answer data into voice answer data and transmit it to the user communication terminal, and transmit the entity information to the user's communication terminal. At least one unit entity information is generated by classifying the address information, order request information, and reservation request information, and the voice query request data is classified into unit voice query request data corresponding to the unit entity information. Unit entity information and the unit voice query request data are matched, matched order request information is classified and stored for each user, and when the stored and managed order request information is requested, the order request information is displayed for each user. An integrated management server that divides and displays the unit voice query request data matched with the unit entity information through a speaker when displaying the unit information by classifying it according to the entity information within the unit of user and performing voice confirmation on the unit entity information. includes wealth

The integrated management server unit may include a call response server unit connected to a mobile communication network, receiving the voice query request data from the user communication terminal, and transmitting the voice response data to the user communication terminal; a voice recognition unit converting the voice query request data into text query request data; a learning database unit that takes user query request data as an input value, learns query request intention in the user query request data as a feature value, and outputs query request intention data for the user query request data as a result value; An inference engine that receives query request intent data for the text query request data in conjunction with the learning database unit, and generates intent inference result data for the text query request data based on the received query request intent data. wealth; an answer database unit extracting the text answer data and entity information matching the intention inference result data and providing them to the inference engine unit; a voice synthesis unit which converts the text response data into the voice response data and provides the converted text response data to the user communication terminal through the call response server unit; At least one unit entity information is generated by classifying the entity information according to the user's address information, order request information, and reservation request information, and the voice query request data is classified into unit voice query request data corresponding to the unit entity information. an entity manager for matching the classified unit entity information and the unit voice query request data, classifying matched order request information for each user, and storing and managing the matched order request information; and when requesting the order request information stored and managed through the entity management unit, the order request information is divided and displayed in units of users through a display, but is divided and displayed by the entity information within the unit of users, and displayed in the unit entity information. and an entity monitoring unit that outputs the unit voice query request data matched with the unit entity information through a speaker when voice confirmation is performed.

According to the present invention, by extracting only the voice of a sentence including an entity from which order reservation information is derived and automatically processing a call, reducing the loss of orders and reservations due to voice recognition errors, and for reconfirming consultation details An automatic call answering system for ordering and reservations that can minimize time and manpower consumption can be provided.

1 is a schematic diagram showing the overall configuration of an automatic call answering system for ordering and reservation according to an embodiment of the present invention.
2 is a diagram illustrating the detailed configuration and operation method of an automatic call answering system for ordering and reservation according to an embodiment of the present invention.
3 is a diagram illustrating an operation method of an entity management unit and an entity monitoring unit according to an embodiment of the present invention.

The terms used in this specification will be briefly described, and the present invention will be described in detail.

The terms used in the present invention have been selected from general terms that are currently widely used as much as possible while considering the functions in the present invention, but these may vary depending on the intention of a person skilled in the art or precedent, the emergence of new technologies, and the like. In addition, in a specific case, there is also a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the invention. Therefore, the term used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, not simply the name of the term.

When it is said that a certain part "includes" a certain component throughout the specification, it means that it may further include other components without excluding other components unless otherwise stated. In addition, terms such as "...unit" and "module" described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software or a combination of hardware and software. .

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in many different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

1 is a schematic diagram showing the overall configuration of an automatic call response system for ordering and reservation according to an embodiment of the present invention, and FIG. 2 is a detailed configuration of an automatic call response system for ordering and reservation according to an embodiment of the present invention 3 is a diagram shown to explain an operating method of an entity management unit and an entity monitoring unit according to an embodiment of the present invention.

Referring to FIG. 1 , an automatic call response system 1000 for ordering and reservation according to an embodiment of the present invention may include an integrated management server unit 100 connected to a user communication terminal 10 .

The user communication terminal 10 receives voice query request data from the user, transmits the input to the integrated management server unit 100, and receives voice answer data according to the voice query request data from the integrated management server unit 100, thereby allowing the user to can be provided to

In this embodiment, the user communication terminal 10 includes a smart phone, a tablet PC, a laptop computer, an electronic book terminal, a digital broadcasting terminal, a PDA (Personal Digital Assistants), a PMP (Portable Multimedia Player), It may be an MP3 player, digital camera, or navigation device, but is not limited thereto.

The integrated management server unit 100 is connected to the user communication terminal 10 to receive voice query request data, infer the query request intention for the voice query request data, and pre-established based on the result of the inference of intention. Text response data and entity information may be extracted from the database, and the extracted text response data may be converted into voice response data and transmitted to the user terminal 10 .

The integrated management server unit 100 according to this embodiment has the same configuration as a conventional web server in terms of hardware, and in terms of software, through various types of languages such as C, C++, Java, Visual Basic, Visual C, etc. It may include program modules that are implemented and perform various functions. In addition, general server hardware is available in various ways depending on operating systems such as DOS, Windows, Linux, Unix, Macintosh, Android, and iOS. It can be implemented using a web server program that is being provided.

An example of a wireless communication network of an Internet network connecting the user communication terminal 10 and the integrated management server unit 100 according to the present embodiment is technical standards or communication methods for mobile communication (eg, GSM ( Global System for Mobile communication), CDMA (Code Division Multi Access), CDMA2000 (Code Division Multi Access 2000), EV-DO (Enhanced Voice-Data Optimized or Enhanced Voice-Data Only), WCDMA (Wideband CDMA), HSDPA (High Speed Downlink Packet Access), HSUPA (High Speed Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), 5G, etc.) It is not. In addition, an example of the wired communication network may be a closed network such as a local area network (LAN) and a wide area network (WAN), preferably an open network such as the Internet. Internet consists of the TCP/IP protocol and several services that exist on its upper layer, namely HTTP (HyperText Transfer Protocol), Telnet, FTP (File Transfer Protocol), DNS (Domain Name System), SMTP (Simple Mail Transfer Protocol), SNMP ( Simple Network Management Protocol), Network File Service (NFS), and Network Information Service (NIS).

As shown in FIG. 2, the integrated management server unit 100 includes a call response server unit 110, a voice recognition unit 120, a learning database unit 130, an inference engine unit 140, an answer database unit ( 150), a voice synthesis unit 160, an entity management unit 170, and an entity monitoring unit 180.

The call response server unit 110 is connected to the mobile communication network 20 to receive voice query request data from the user communication terminal 10 and transmit voice response data to the user communication terminal 10 .

The voice recognition unit 120 receives voice query request data from the call response server 110 and converts the received voice query request data into text query request data through a speech to text (STT) conversion module. can

The learning database unit 130 takes the user query request data as an input value, learns the query request intention in the input user query request data as a feature value, and obtains the query request intention data for the user query request data as a result. value can be output. Here, the user query request data may include the user's voice data recorded at the time of ordering or making a reservation. Accordingly, the learning database unit 130 may learn words representing the intention of the query request from the voice data as feature values, and output the learned feature values as result values. The learning database unit 130 may store the result of learning each characteristic value and the corresponding characteristic value for various intentions for various query requests in advance, and the inference engine unit 140 accurately determines the intention. enable inference.

The reasoning engine unit 140 interworks with the learning database unit 130 to receive query request intention data for the text query request data, and the intention for the text query request data based on the received query request intention data. Inference result data can be created.

That is, when the text query request data is used as an input value of the learning database unit 130 and the pre-learned query request intention data is returned to the inference engine unit 140, the query request intention data among the returned query request intention data Intention inference result data can be created by finally selecting the value with the highest possibility as the intention. Here, the highest possibility as query intention means that when specific text query request data is an input value and specific query request intention data is output as a result value, the final selection frequency is high, that is, the query has a statistically high probability of selection. That is, the request intention data is determined as the intention inference result data.

The answer database unit 150 may extract text answer data and entity information that match the intention inference result data of the reasoning engine unit 140 and provide the extracted entity information to the reasoning engine unit 140 . Here, the entity information may include at least one of the user's address information, order request information, and reservation request information. Appropriately classified concept information may be set as entity information.

The voice synthesis unit 160 converts the text response data provided from the reasoning engine unit 140 into voice response data through a text to speech (TTS) conversion module, and provides it to the call response server unit 110 to respond to a call. It is provided to the user communication terminal 10 through the server unit 110.

The entity management unit 170 may store and manage entity information and voice query request data corresponding to the entity information by matching them. More specifically, the entity manager 170 classifies the entity information according to the user's address information, order request information, and reservation request information to generate at least one unit entity information, and the voice query request data corresponds to the unit entity information. After classifying each of the unit voice query request data, the classified unit entity information and the unit voice query request data are matched, and the matched order request information is classified for each user and stored and managed.

For example, as shown in FIG. 3, entity information is a concept of a category classified into user's address information, order request information, and reservation request information (not shown), and each of these is defined as unit entity information. Substantive content of voice query request data can be classified into unit entity information. For example, voice content related to an address is classified into unit entity information called address information of a user, and voice content related to order details is user order request information ( Order details) can be classified as unit entity information. At this time, as a classification method, based on each entity of the voice content, it is possible to identify and classify the unit entity information of which classification the corresponding entity belongs to. The unit entity information A1 classified as above and the unit voice query request data B1 may be matched, and another unit entity information A2 and the unit voice query request data B2 may be matched with each other. The matched 'A1-B1' and 'A2-B2' can be classified and stored for each user, and being classified and stored for each user means that 'A1-B1' and 'A2-B2' are classified based on the user's contact information. It can mean that information is stored in a state that can be identified with other users. In this case, each information may be stored and managed by identifying not only the user's contact information but also the user's name, call connection time, and the like.

When a manager requests order request information stored and managed through the entity management unit 170, the entity monitoring unit 180 classifies and displays the order request information by user unit through a display, but displays the entity information within the user unit. It is displayed separately, and unit voice query request data matched with the unit entity information can be output through the speaker when voice confirmation is performed on the unit entity information.

For example, if information on the status of AI call order/reservation acceptance is requested based on a specific user, the first unit entity information A1 and the first unit voice query request data matched with each other as shown in FIG. 3 (B1), second unit entity information (A2) and second unit voice query request data (B2) matched with each other are displayed, and when the play icon of each unit voice query request data is clicked or touched to select The user's voice part included in the unit voice query request data matched with the corresponding unit entity information is output through the speaker so that it can be easily confirmed whether it matches the corresponding unit entity information.

Referring to FIG. 2, the operation method and sequence of the automatic call answering system 1000 for ordering and reservation according to an embodiment of the present invention will be described below.

First, when the user communication terminal 10 sends a call connection request signal to the call response server unit 110 through the mobile communication network 20 (①), the call response server unit 110 makes a call through the mobile communication network 20. A connection approval signal is transmitted to the user communication terminal 10 to establish a connection between the user communication terminal 10 and the call response server 110 .

Thereafter, when the user transmits an order request signal (voice quality query request data) to the call response server unit 110 through the user communication terminal 10 (③), the call response server unit 110 receives it and the voice recognition unit Transfer to (120) (④). At this time, the voice recognition unit 120 converts the voice query request data into text query request data, and then transmits the voice query request data (voice data) to the reasoning engine unit 140 (⑤).

Thereafter, when the reasoning engine unit 140 transfers the text query request data among the received text query request data and voice query request data (voice data) to the learning database unit 130 (⑥), the learning database unit 130 Text query request data is used as an input value, and query request intention data related thereto is provided to the inference engine unit 140 as a result value based on previously learned data (⑦).

Thereafter, the reasoning engine unit 140 transfers the query request intention data received from the learning database unit 130, that is, the intention inference result to the answer database unit 150 (⑧), and the answer database unit 150 infers the intention Text answer data suitable for the result is extracted from the previously stored answer data and provided again to the inference engine unit 140 (⑨).

Thereafter, the inference engine unit 140 classifies entity information from the text answer data received from the answer database unit 150, and if there is user's voice query request data corresponding to the classified entity information, the corresponding voice query request data and entity information to the entity management unit 170. At this time, the entity management unit 170 matches the voice query request data received from the reasoning engine unit 140 with the entity information, and stores and manages the matching.

Thereafter, the reasoning engine unit 140 transfers the text answer data received from the answer database unit 150 to the voice synthesis unit 160 (⑪), and the voice synthesis unit 160 transmits the text response data from the reasoning engine unit 140. The received text response data may be converted into voice response data and provided to the call response server unit 110 (⑫). At this time, the call response server unit 110 converts the voice response data received from the voice synthesis unit 160 It is transmitted to the user communication terminal 10 through the mobile communication network 20 to complete the response to the AI call response (⑬).

Thereafter, when a request is generated from the entity monitoring unit 180 to the entity management unit 170 for an entity, the entity management unit 170 transmits mutually matched voice query request data and entity information for the corresponding user to the entity monitoring unit 180. ) is transmitted (⑭).

Accordingly, as shown in FIG. 3, the entity monitoring unit 180 can check entity information (text information) classified according to address and order details for each call time and contact, and play icon matched with each entity information. If is selected, the user can output voice data related to the corresponding entity information. At this time, the output voice data is not the entire voice data recorded between the user's phone calls, but corresponds to partial voice data corresponding to each entity information, and each voice data is matched and divided for each entity information, It is possible to check the voice data divided according to the corresponding entity information.

What has been described above is only one embodiment for implementing an automatic call answering system for ordering and reservation according to the present invention, and the present invention is not limited to the above embodiment, and as claimed in the following claims Anyone with ordinary knowledge in the field to which the present invention pertains without departing from the gist of the present invention will say that the technical spirit of the present invention exists to the extent that various changes can be made.

1000: Automated call answering system for ordering and reservations
100: integrated management server unit
110: call response server unit
120: voice recognition unit
130: learning database unit
140: reasoning engine unit
150: answer database unit
160: voice synthesis unit
170: entity management unit
180: entity monitoring unit
10: user communication terminal
20: mobile communication network

Claims (2)

The voice query request data is received from the user communication terminal through a call connection with the user communication terminal, the query request intention is inferred from the voice query request data, and the text answer data and Entity information is extracted, the extracted text response data is converted into voice response data and transmitted to the user communication terminal, and the entity information is classified according to the user's address information, order request information, and reservation request information to form at least one unit entity. After generating information, classifying the voice query request data into unit voice query request data corresponding to the unit entity information, matching between the classified unit entity information and the unit voice query request data, and matching the matched order The request information is classified and stored and managed by user, and when the order request information stored and managed is requested, the order request information is classified and displayed by user unit through a display, but is classified and displayed by the entity information within the user unit, and an integrated management server unit outputting the unit voice query request data matched with the unit entity information through a speaker upon voice confirmation of the unit entity information.
According to claim 1,
The integrated management server unit,
a call response server unit connected to a mobile communication network, receiving the voice query request data from the user communication terminal, and transmitting the voice response data to the user communication terminal;
a voice recognition unit converting the voice query request data into text query request data;
a learning database unit that takes user query request data as an input value, learns query request intention in the user query request data as a feature value, and outputs query request intention data for the user query request data as a result value;
An inference engine that receives query request intent data for the text query request data in conjunction with the learning database unit, and generates intent inference result data for the text query request data based on the received query request intent data. wealth;
an answer database unit extracting the text answer data and entity information matching the intention inference result data and providing them to the inference engine unit;
a voice synthesis unit which converts the text response data into the voice response data and provides the converted text response data to the user communication terminal through the call response server unit;
At least one unit entity information is generated by classifying the entity information according to the user's address information, order request information, and reservation request information, and the voice query request data is classified into unit voice query request data corresponding to the unit entity information. an entity manager for matching the classified unit entity information and the unit voice query request data, classifying matched order request information for each user, and storing and managing the matched order request information; and
When the order request information stored and managed through the entity management unit is requested, the order request information is divided and displayed on a user basis through a display, but is divided and displayed by the entity information within the user unit, and the unit entity information and an entity monitoring unit outputting the unit voice query request data matched with the unit entity information through a speaker when voice confirmation is performed.

Images