KR20210078038A - Control method, device and program of chatbot system - Google Patents

Abstract

Provided is a control method of a chatbot system. The control method of the chatbot system comprises: a step of receiving, by a server, a query from a second user terminal through a first chat room wherein a first user terminal and a second user terminal participate; a step of transmitting, by the server, the query to the first user terminal; a step of obtaining, by the server, a first preliminary response message by analyzing the query; and a step of transmitting, by the server, the first preliminary response message to at least one user terminal among the first user terminal and the second user terminal in which the first preliminary response message is in accordance with a preset condition. Therefore, the present invention is capable of having an effect of improving a user convenience.

Description

Control method, device and program of chatbot system {CONTROL METHOD, DEVICE AND PROGRAM OF CHATBOT SYSTEM}

The present invention relates to a method, apparatus and program for controlling a chatbot system.

An artificial intelligence (AI) system is a computer system that implements human-level intelligence, and unlike the existing rule-based smart system, the machine learns, judges, and becomes smarter by itself. The more the artificial intelligence system is used, the better the recognition rate and the more accurate understanding of user preferences, and the existing rule-based smart systems are gradually being replaced by deep learning-based artificial intelligence systems.

Speech recognition and image analysis are representative fields to which artificial intelligence systems are applied. In particular, in the field of speech recognition, a natural language processing technology for analyzing a user's conversation to understand a user's intention and a natural language processing technology for generating a natural answer based on the user's intention are required.

However, for accurate speech recognition, there is a problem that, apart from technological development, a lot of learning data and high-performance GPU are required, so the effect is not large compared to the cost input.

Therefore, in a chatbot system using a technology element in the field of voice recognition, it is necessary to discuss a method for constructing an artificial intelligence model for conversation with a user.

Registered Patent Publication No. 10-1756261 (Registered on 2017.07.04)

An object of the present invention is to provide a control method, an apparatus, and a program for a chatbot system.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

In a method for controlling a chatbot system according to an aspect of the present invention for solving the above problems, a server receives a query from a second user terminal through a first chat room in which a first user terminal and a second user terminal participate to do; transmitting, by the server, the query to a first user terminal; obtaining, by the server, a first preliminary response message by analyzing the query; transmitting, by the server, the first preliminary response message to at least one of the first user terminal and the second user terminal according to a condition in which the first preliminary response message is preset; includes

In this case, the control method may include: obtaining, by the server, related data by analyzing the query; and transmitting, by the server, the related data to at least one of the first user terminal and the second user terminal.

In this case, the acquiring of the related data may include: extracting a keyword based on the query; searching for first related data corresponding to the keyword in an internal DB of the server; searching an external server for second related data corresponding to the keyword; and transmitting at least one of the first related data and the second related data to at least one of the first user terminal and the second user terminal. The method comprising: receiving, by the second user terminal, at least one of the first related data and the second related data from the server; and displaying, by the second user terminal, the at least one related data in a first area of the first chat room; Including, the related data displayed in the first area may be data downloadable to the second user terminal and url information accessible to the user terminal.

In this case, the transmitting to at least one of the first user terminal and the second user terminal may include: obtaining a security level of the first related data; obtaining a user rating of the second user terminal; when the security level of the first related data is higher than the user level of the second user terminal, transmitting the first related data to the first user terminal; and when the security level of the first related data is lower than the user level of the second user terminal, transmitting the first related data to the second user terminal. In the step of transmitting to the first user terminal, when a user command for transmitting the first related data to the second user terminal is input from the first user terminal, the server transmits the first related data to the second transmitting to a user terminal; may include.

In this case, the preset condition is a condition for whether the first user terminal is connected to the server, and sends the first preliminary response message to at least one of the first user terminal and the second user terminal. The transmitting may include, when the first user terminal is connected to the server, transmitting, by the server, the first preliminary response message to the first user terminal; obtaining, by the first user terminal, a first response message based on the first preliminary response message and transmitting the first response message to the server; transmitting, by the server, the first response message to the second user terminal; and displaying, by the second user terminal, the first response message in a second area of the first chat room, wherein when the first user terminal is not connected to the server, the server transmitting a first preliminary response message to the first user terminal and the second user terminal; and displaying, by the first user terminal and the second user terminal, the first preliminary response message in a second area of the first chat room.

In this case, the acquiring of the first related data may include: acquiring personal information and past conversation history corresponding to the second user terminal; acquiring the first related data based on the acquired personal information, past conversation history, and the keyword; Inviting, by the server, to a second chat room including the first user terminal, the second user terminal, and the third user terminal when the first related data is the related data received from the third user terminal ; and obtaining, by the server, the conversation contents of the second user terminal and the third user terminal collected through the second chat room.

In this case, the step of obtaining the conversation contents of the second user terminal and the third user terminal may include: obtaining, by the server, query information of the second user terminal through the second chat room; obtaining, by the server, a second preliminary response message by analyzing the query; obtaining, by the server, a security level of the second preliminary response message; and outputting, by the server, the second preliminary response message to the first chat room when the security level of the second preliminary response message exceeds a preset grade; and outputting, by the server, the second preliminary response message to the second chat room when the security level of the second preliminary response message is less than or equal to a preset grade.

Other specific details of the invention are included in the detailed description and drawings.

According to various embodiments of the present invention described above, a chatbot system in which a counselor and artificial intelligence are combined is provided, thereby improving user convenience.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a system diagram illustrating a system according to an embodiment of the present invention.
2 is a flowchart illustrating a chatbot system according to an embodiment of the present invention.
3 is a flowchart illustrating a method of providing related data according to an embodiment of the present invention.
4 is a flowchart illustrating a method of acquiring a type of related data according to an embodiment of the present invention.
5 is a flowchart illustrating a method of providing related data according to a type of related data according to an embodiment of the present invention.
6 is a flowchart illustrating a message provided to a second user terminal according to whether the first user terminal is connected according to an embodiment of the present invention.
7 is a flowchart illustrating a method of determining a related user based on a past history of a first user terminal according to an embodiment of the present invention.
8 is a flowchart illustrating a method of utilizing a plurality of chat rooms according to an embodiment of the present invention.
9 is an exemplary diagram for explaining a chat room according to an embodiment of the present invention.
10 is a block diagram of an apparatus according to an embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention belongs. It is provided to fully understand the scope of the present invention to those skilled in the art, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. As used herein, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless specifically defined explicitly.

As used herein, the term “unit” or “module” refers to a hardware component such as software, FPGA, or ASIC, and “unit” or “module” performs certain roles. However, “part” or “module” is not meant to be limited to software or hardware. A “unit” or “module” may be configured to reside on an addressable storage medium or to reproduce one or more processors. Thus, by way of example, “part” or “module” refers to components such as software components, object-oriented software components, class components and task components, processes, functions, properties, Includes procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. Components and functionality provided within “parts” or “modules” may be combined into a smaller number of components and “parts” or “modules” or as additional components and “parts” or “modules”. can be further separated.

Spatially relative terms "below", "beneath", "lower", "above", "upper", etc. It can be used to easily describe the correlation between a component and other components. A spatially relative term should be understood as a term that includes different directions of components during use or operation in addition to the directions shown in the drawings. For example, when a component shown in the drawing is turned over, a component described as “beneath” or “beneath” of another component may be placed “above” of the other component. can Accordingly, the exemplary term “below” may include both directions below and above. Components may also be oriented in other orientations, and thus spatially relative terms may be interpreted according to orientation.

In this specification, a computer means all types of hardware devices including at least one processor, and may be understood as encompassing software configurations operating in the corresponding hardware device according to embodiments. For example, a computer may be understood to include a smartphone, a tablet PC, a desktop, a notebook computer, and a user client and an application running on each device, but is not limited thereto.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Each step described in this specification is described as being performed by a computer, but the subject of each step is not limited thereto, and at least a portion of each step may be performed in different devices according to embodiments.

1 is a system diagram illustrating a system according to an embodiment of the present invention.

The system according to the present invention may include a server 10 , a first user terminal 20 , and a second user terminal 30 .

The server 10 is configured to provide a chatbot service to the first user terminal 20 and the second user terminal 30 . As an embodiment, the chatbot service may be provided to the first user terminal 20 and the second user terminal 30 in the form of an application.

The server 10 may use an artificial intelligence model to obtain a preliminary response message and related data by analyzing a query received from at least one of the first user terminal 20 and the second user terminal 30 .

In an embodiment, the server may train the artificial intelligence model by acquiring a query message for learning, a preliminary response message to the query message, and data related to the query message as training data. Specifically, the server 10 may train the AI model to output a preliminary response message and related data by inputting a query message into the AI model.

Furthermore, the server 10 may additionally acquire a query message received from a user using the chatbot, a preliminary response message thereto, and a related data message as learning data. In this case, the server 10 may generate answers to all chatbot users through the first general purpose artificial intelligence model, and further, may generate answers to specific users through the personalized second artificial intelligence model. .

That is, the server 10 acts as the training data of the first artificial intelligence model in which the initially set arbitrary learning data and the learning data obtained from each user'D, but the learning data learned from a specific user is the second artificial intelligence It can be used as training data for the model.

In one embodiment, when the server 10 receives a query for a specific user, the server 10 independently inputs the query as an input value to the first AI model and the second AI model to output values (ie, preliminary response message and related data), and at least one of an output value output through the first artificial intelligence model and an output value output through the second artificial intelligence model may be obtained as a final output value. For example, when the accuracy of the output value output through the second artificial intelligence model is less than or equal to a preset value, the server 10 acquires the output value output through the first artificial intelligence model as a final output value, and the second artificial intelligence model When the accuracy of the output value output through the ? is greater than the preset value, the output value output through the second artificial intelligence model may be obtained as the final output value. Through the above-described method, the server 10 provides a personalized response using the second artificial intelligence model, but uses a second artificial intelligence model that does not guarantee accurate result output due to a relatively insufficient amount of learning data. This can be supplemented by the model.

On the other hand, the server 10 according to the present invention analyzes a query received from at least one user terminal among the first user terminal 20 and the second user terminal 30, a natural language understanding unit (NLU) ), a query management management module, a natural language generator (NLG), and a database.

The natural language understanding unit is configured to identify a task related to an entity and an intent of the query based on the received query. Specifically, the natural language understanding unit may interpret the sentence by analyzing the structure and main components of the sentence, and may perform the sentence analysis using statistics/analysis.

The query management management module is a configuration for acquiring information on responses to queries based on the natural language understanding results and data stored in the database. It goes without saying that the query management management module can be implemented using Frame-based, Agent-based, etc., and it can also be implemented through MDP (Markov Decision Process) and Reinforcement Learning-based modeling. In this case, the query management management module may acquire information for generating a response (preliminary response message or related data), and as described above, the acquired information includes the task identified through the natural language understanding unit and data stored in the database. can be determined based on The natural language generator may acquire a natural language as a response to a query based on information obtained through the query management management module.

The database is a configuration for storing information necessary to generate a response in the query management management module. In this case, the data stored in the database may be various. In an embodiment, the data stored in the database includes a history of the past conversation contents of the first user terminal 20 and the second user terminal 30, the type of related data transmitted according to the past conversation contents, the second It may be data received from the user terminal 30 . However, of course, in addition to the above data, the server 10 may store various data for determining data to be transmitted to the first user terminal 20 and the second user terminal 30 .

In this case, it may be an artificial intelligence model based on a convolutional deep neural network (CNN) that is learned by inputting the artificial intelligence model learning data according to the present invention. However, the present invention is not limited thereto, and it goes without saying that various artificial intelligence models may be applied to the present invention. For example, a model such as a Deep Neural Network (DNN), a Recurrent Neural Network (RNN), or a Bidirectional Recurrent Deep Neural Network (BRDNN) may be used as the AI model, but is not limited thereto.

In this case, convolutional deep neural networks (CNNs) are a type of multilayer perceptrons designed to use minimal preprocessing. A convolutional neural network consists of one or several convolutional layers and general artificial neural network layers on top of it, and additionally utilizes weights and pooling layers. Thanks to this structure, the convolutional neural network can fully utilize the input data of the two-dimensional structure. In addition, convolutional neural networks can be trained via standard backpropagation. Convolutional neural networks are easier to train than other feedforward neural network techniques and have the advantage of using fewer parameters.

In addition, deep neural networks (DNNs) are artificial neural networks (ANNs) composed of a plurality of hidden layers between an input layer and an output layer.

In this case, the structure of the deep neural network may be composed of a perceptron. A perceptron consists of several inputs, one processor, and one output value. The processor multiplies each input value by a weight, and then sums the input values multiplied by the weight. Then, the processor outputs one output value by substituting the summed value into the activation function. If a specific value is desired as the output value of the activation function, the weight multiplied by each input value may be corrected, and the output value may be recalculated using the modified weight. In this case, each perceptron may use a different activation function. In addition, each perceptron receives the outputs from the previous layer as input, and then uses the activation function to obtain the output. The obtained output is transferred to the input of the next layer. Through the process as described above, some output values can be finally obtained.

In addition, returning to the description of the deep learning technique, a Reccurent Neural Network (RNN) refers to a neural network in which connections between units constituting an artificial neural network constitute a directed cycle. Unlike forward neural networks, recurrent neural networks can utilize the memory inside the neural network to process arbitrary inputs.

Deep Belief Networks (DBN) is a generative graphical model used in machine learning, and in deep learning, it means a deep neural network composed of multiple layers of latent variables. There is a connection between layers, but there is no connection between units within a layer.

The deep trust neural network can be used for prior learning due to the nature of the generative model, and after learning the initial weight through prior learning, the weights can be fine-tuned through backpropagation or other discrimination algorithms. This feature is very useful when the training data is small, because the smaller the training data, the stronger the effect of the initial value of the weight on the resulting model. The pre-learned initial weight value is closer to the optimal weight compared to the arbitrarily set initial weight value, which enables performance and speed improvement in the fine-tuning stage.

The above-described artificial intelligence and its learning method have been described for illustrative purposes, and the artificial intelligence and its learning method used in the embodiments described below are not limited. For example, all kinds of artificial intelligence technology and a learning method thereof that a person skilled in the art can apply to solve the same problem may be utilized to implement the system according to the disclosed embodiment.

The first user terminal 20 may be a terminal of a service provider for providing a service. Of course, according to an embodiment, the server 10 and the first user terminal 20 may be implemented as one device.

The second user terminal 30 may be a user terminal for making a query to the server 10 . The second user terminal 30 may communicate with the server 10 and the first user terminal 20 based on the chatbot application provided from the server 10 , and specifically, through a chat room provided from the server 10 . It can communicate with the server 10 and the first user terminal 20 .

In one embodiment, the first user terminal 20 and the second user terminal 30 are a smartphone, a tablet personal computer (PC), a mobile phone, a video phone, an e-book reader (e) -book reader), desktop PC (desktop PC), laptop PC (laptop PC), netbook computer (netbook computer), workstation (workstation), server, PDA (personal digital assistant), at least one of PMP (portable multimedia player) may include.

Hereinafter, various embodiments of the present invention will be described with reference to FIGS. 2 to 9 .

2 is a flowchart illustrating a chatbot system according to an embodiment of the present invention.

In step S110 , the server 10 may receive a query from the second user terminal 30 through the first chat room in which the first user terminal 20 and the second user terminal 30 participate.

In an embodiment, the second user terminal 30 may input a query into the first chat room created by executing the chatbot application, and the server 10 may receive the query registered in the first chat room.

In step S120 , the server 10 may transmit a query to the first user terminal 20 . Specifically, the server 10 may transmit the query generated by the second user terminal 30 to the first chat room created by the chatbot application of the first user terminal 20 .

In step S130, the server 10 may obtain a first preliminary response message by analyzing the query.

In this case, the preliminary response message may be a message to be provided to the second user terminal 30 based on the query analyzed by the server 10 .

In one embodiment, when the query is "I want to request a design", the server 10 pre-responds to the message "Which one is web, app, icon, or UI/UX design?" as shown in FIG. 9 . It can be obtained by message.

In step S140 , the server 10 transmits the first preliminary response message to at least one of the first user terminal 20 and the second user terminal 30 according to a condition that the first preliminary response message is preset. can be transmitted

As will be described later, the preset condition may be whether the first user terminal 20 is connected to the server 10 . In one embodiment, when the first user terminal 20 is connected to the server 10, the server 10 transmits a preliminary response message to the first user terminal 20, otherwise, the server 10 may transmit the preliminary response message to the second user terminal 30 .

3 is a flowchart illustrating a method of providing related data according to an embodiment of the present invention.

In step S210 , the server 10 may obtain related data by analyzing the query.

As an embodiment, the related data may mean data related to the query of the second user terminal 30 . For example, as shown in the first area 910 of FIG. 9 , the related data may mean a file for a design proposal.

In step S220 , the server 10 may transmit the related data to at least one of the first user terminal 20 and the second user terminal 30 .

As will be described later, when the first user terminal 20 is not connected to the server 10 , the server 10 transmits the related data to the second user terminal 30 according to the security level of the related data. can determine whether

Meanwhile, according to various embodiments of the present invention, the preliminary response message and related data may be classified based on the type of the chatbot's response to the user's query.

For example, when the chatbot's response to the user's query is an additional query, the server 10 may acquire the additional query as preliminary response data. Also, when the chatbot's response to the user's query is an answer response, the server 10 may acquire the response response as related data.

In this case, if there is no additional inquiry from the second user terminal 30 for a preset time after the related data is transmitted, the server 10 determines that the conversation has ended, and from the first user inquiry to the last related data transmission User queries, preliminary response data, and related data can be stored and managed as one data set.

4 is a flowchart illustrating a method of acquiring a type of related data according to an embodiment of the present invention.

In step S310, the server 10 may extract a keyword based on the query.

In one embodiment, the server 10 may extract a keyword by morphological analysis of the user's query.

In another embodiment, the server 10 inputs a user query to the natural language understanding unit to obtain a task related to the entity for the query and the intent of the query, and uses the obtained task as a keyword. can judge In this case, the task related to the intent of the query may be information necessary for responding to the user's query. For example, if the user query is a query to inquire about a design, the task may be a design query. Alternatively, when the user query is a query for requesting an estimate, the task may be an estimate creation target and an estimate cost inquiry for the target.

In step S320 , the server 10 may search for the first related data corresponding to the keyword in the internal DB of the server 10 .

In step S330 , the server 10 may search the external server 10 for second related data corresponding to the keyword.

That is, the server 10 may search for the acquired keywords not only in the internal DB stored in the server 10 but also in the external server to deliver more accurate and rich results to the user. In this case, the external server may include various sites such as a portal site, an SNS site, and a website.

In step S340 , the server 10 may transmit at least one of the first related data and the second related data to at least one of the first user terminal 20 and the second user terminal 30 . have.

In step S350 , the second user terminal 30 may receive at least one of the first related data and the second related data from the server 10 .

In step S360, the second user terminal 30 may display at least one piece of related data in the first area 910 of the first chat room.

In this case, the related data displayed in the first area 910 may be data downloadable by the second user terminal 30 and url information accessible by the user terminal.

In one embodiment, as shown in FIG. 9 , when the second user terminal 30 acquires "Gangwon Open Market_Design Proposal.pdf" as related data, the related data is transferred to the upper area of the first chat room. It may be displayed in the first area 910 .

As another embodiment, when the related data is url information accessible to the user terminal, the second user terminal 30 may receive the url information from the server 10 and display it in the first area 910 .

In another embodiment, when the related data is data for an answer response rather than an additional question response of the chatbot as described above, the second user terminal 30 receives the response response from the server 10 and receives the first It may be displayed in the area 910 . For example, in the embodiment of Fig. 9, if the response is not an additional query to the user's query (which one is web, app, icon, or UI/UX design?), such as "design request is a service not supported", chatbot may be displayed in the first area 910 instead of the second area 920 .

5 is a flowchart illustrating a method of providing related data according to a type of related data according to an embodiment of the present invention.

In step S410, the server 10 may obtain a security level of the first related data.

Specifically, the first related data obtained from the internal DB of the server 10 may need security unlike the second related data obtained from the external server. Accordingly, the server 10 may determine the security level of the first related data and determine whether to transmit the first related data to the second user terminal 30 .

In step S420 , the server 10 may acquire a user rating of the second user terminal 30 .

In this case, the security level may mean a value expressed as an absolute numerical value, such as a first-class security level or a second-level security level, but may mean a relative security level that is set differently depending on a user using the chatbot.

For example, when the first related data is data received from the second user terminal 30 in the past, the server 10 transmits the first related data to the second user regardless of the security level displayed as an absolute numerical value. It may be transmitted to the terminal 30 .

The user rating is also a concept corresponding to the security level, and may be an absolute numerical user rating, but may be a relative user rating set differently depending on the user using the chatbot.

Meanwhile, the server 10 may store and manage data for use as the first related data in a separate database. For example, the server 10 may include a first database, and data stored in the first database may be used to obtain the first associated data.

In this case, the server 10 may include acquiring data to be stored in the first database and adding an index indicating the source of the acquired data. For example, when the data is data obtained from an external server, the server 10 adds an external server index to the data, and when the data is data obtained from the second user terminal 30, the server 10 adds a second user terminal index to the corresponding data, and when the data is data obtained from the third user terminal, the server 10 adds a third user terminal index to the corresponding data, and the data is sent to the server 10 In the case of data obtained from the corresponding user terminal, the server 10 may add a manager index to the corresponding data. The above-described external server index, the second user terminal index, the third user terminal index, and the manager index are indexes that are distinct from each other and do not depend on their names. For example, the server 10 may set the external server index to 0000, the second user terminal index to 0001, the third user terminal index to 0010, and the administrator index to 0011.

In step S430 , when the security level of the first related data is higher than the user level of the second user terminal 30 , the server 10 may transmit the first related data to the first user terminal 20 .

That is, when it is determined that the risk of transmitting the first related data to the second user terminal 30 without any verification is high, the server 10 may transmit the first related data to the first user terminal 20 .

In step S440 , when the security level of the first related data is lower than the user level of the second user terminal 30 , the server 10 may transmit the first related data to the second user terminal 30 .

The meaning of the case where the security level of the first related data is lower than the user level of the second user terminal 30 may be interpreted literally when the security level and the user level are related to absolute values. For example, when the security level is level 2 and the user level is level 1 , the server 10 may transmit the first related data to the second user terminal 30 .

However, when the security level and the user level are relatively determined, the meaning of the case that the security level of the first related data is lower than the user level of the second user terminal 30 means that the second user terminal 30 is the first related data It may mean that you are entitled to receive For example, when the first related data is data received by the second user terminal 30 , the server 10 determines that the security level of the first related data is lower than the user level of the second user terminal 30 . and the server 10 may transmit the first related data to the second user terminal 30 . Of course, the above-described example is equally applied to the above-described step S430.

In step S450, when a user command for transmitting the first related data to the second user terminal 30 is input from the first user terminal 20, the server 10 transmits the first related data to the second user terminal (30) can be transmitted.

That is, in the above-described step S430, the first user terminal 20 receives the first related data, and transmits the received first related data to the second user terminal 30, so that it can be operated by the server 10 There is an effect that an operation that is not there can be performed through the first user terminal 20 .

Meanwhile, according to various embodiments of the present disclosure, it goes without saying that the operation of step S430 may be performed only when the first related data is related data to which the manager index is added. That is, when the security level of the first related data is higher than the user level of the second user terminal 30 and the manager index is added to the first related data, the server 10 transmits the first related data to the first user terminal (20) can be transmitted.

For example, when the third user terminal index is added to the first related data, the server 10 and the first user terminal 20 determine whether to transmit the first related data to the second user terminal 30 . It may not be desirable to do Accordingly, the server 10 may perform the operation of step S430 only when the first related data is related data to which the manager index is added.

6 is a flowchart illustrating a message provided to a second user terminal according to whether the first user terminal is connected according to an embodiment of the present invention.

In step S510 , the server 10 may determine whether the first user terminal 20 is connected to the server 10 .

When the first user terminal 20 is connected to the server 10 (S510-Y), in step S520 , the server 10 may transmit a first preliminary response message to the first user terminal 20 .

In step S530 , the first user terminal 20 may obtain a first response message based on the first preliminary response message and transmit it to the server 10 .

That is, when the first user terminal 20 is connected to the server 20 , the server 10 may respond to the second user terminal 30 through the first user terminal 20 . This is because in some cases, it is desirable for the user to directly participate in the conversation due to the nature of chatting and customer service.

In step S540 , the server 10 may transmit the first response message to the second user terminal 30 .

In step S550 , the second user terminal 30 may display the first response message in the second area 920 of the first chat room.

If the first user terminal 20 is not connected to the server 10 (S510-N), in step S560, the server 10 sends a first preliminary response message to the first user terminal 20 and the second user It can be transmitted to the terminal 30 .

That is, when the first user terminal 20 is not connected to the server 10 (S510-N), the server 10 processes the query of the second user terminal 30 received in real time by the first user It is possible to provide an answer to a query without the participation of the terminal 20 .

In step S570 , the first user terminal 20 and the second user terminal 30 may display the first preliminary response message in the second area 920 of the first chat room.

Meanwhile, according to various embodiments of the present disclosure, the server 10 learns an artificial intelligence model for obtaining a response to a user query based on a response message transmitted by the first user terminal 20 to the server 10 . can do it

In one embodiment, the server 10 compares the preliminary response message to the query sent by the server 10 to the first user terminal 20 with the response message received from the first user terminal 20, As a result of the comparison, when the preliminary response message and the response message are different, the AI model may be trained using the response message transmitted by the first user terminal 20 as training data. Specifically, the server 10 may learn the artificial intelligence model by receiving the user query and the response message transmitted by the first user terminal 20 as training data, and related data may be added as training data as necessary. of course there is

That is, the first user terminal 20 achieves the purpose of generating a response to the second user terminal 30 and obtains learning data for updating the artificial intelligence model of the server 10, so that the server 10 There is a new effect that can improve the response generation of

7 is a flowchart illustrating a method of determining a related user based on a past history of a first user terminal according to an embodiment of the present invention.

In step S610 , the server 10 may acquire personal information and past conversation history corresponding to the second user terminal 30 .

In step S620 , the server 10 may acquire first related data based on the acquired personal information, past conversation history, and keywords.

In an embodiment, the server 10 may acquire the first related data based on the past conversation contents of the second user terminal 30 . For example, when the query of the second user terminal 30 is similar to a query performed in the past, the server 10 may obtain the first related data based on the contents of the past query and the contents of the answer.

In step S630, if the first related data is the related data received from the third user terminal, the server 10, including the first user terminal 20, the second user terminal 30 and the third user terminal You can invite them to the second chat room.

In one embodiment, as described above, when the third user terminal index is added to the first related data, the server 10 determines the corresponding data as data of the third user terminal, and selects the third user terminal as the data of the third user terminal. 2 You can invite to chat room. Unlike the first chat room including the server 10 , the first user terminal 20 and the second user terminal 30 , the second chat room includes the server 10 , the first user terminal 20 , and the second user terminal (30) and a third user terminal may be included.

In step S640, the server 10 may obtain the conversation contents of the second user terminal 30 and the third user terminal collected through the second chat room.

As will be described later in FIG. 8 , the server 10 communicates with at least one of the second user terminal 30 and the third user terminal based on the conversation contents of the second user terminal 30 and the third user terminal. can provide a response.

Specifically, unlike the first chat room in which the server 10 (and the first user terminal 20) and the second user terminal 30 participate, there may be a plurality of recipients of the conversation contents obtained in the second chat room.

In one embodiment, the server 10 responds through the method of FIGS. 2 to 6 when the receiver of the conversation received from at least one of the second user terminal 30 and the third user terminal is a chatbot. can do.

In another embodiment, when the receiver of the conversation received from the second user terminal 30 is the third user terminal or vice versa, the server 10 responds to assisting the sender or receiver of the conversation through the chatbot can create

For example, the server 10 summarizes a part of the conversation performed by the server 10 and the second user terminal 30 in the first chat room for the invited third user terminal and provides it to the third user terminal. can As another example, the server 10 may analyze the conversation content between the second user terminal 30 and the third user terminal, and provide necessary related data to the first area of the second chat room. That is, when the second user terminal 30 and the third user terminal are having a conversation, the server 10 may provide only related data to the first area of the second chat room without providing preliminary response data.

8 is a flowchart illustrating a method of utilizing a plurality of chat rooms according to an embodiment of the present invention.

In step S710 , the server 10 may obtain query information of the second user terminal 30 through the second chat room.

According to various embodiments of the present disclosure, the server 10 may obtain query information of at least one of the second user terminal 30 and the third user terminal through the second chat room.

In step S720, the server 10 may obtain a second preliminary response message by analyzing the query.

As an embodiment, the second preliminary response message may be a message that the second user terminal 30 may provide to the third user terminal.

In step S730, the server 10 may acquire the security level of the second preliminary response message.

In step S740, when the security level of the second preliminary response message exceeds the preset level, the server 10 may output the second preliminary response message to the first chat room.

In step S750, if the security level of the second preliminary response message is less than or equal to the preset level, the server 10 may output the second preliminary response message to the second chat room.

In one embodiment, the server 10, when the second preliminary response message includes personal information or security information of the second user terminal 30, it is inappropriate to output the second preliminary response message to the second chat room. There are cases. Therefore, the server 10, when the security level of the second preliminary response message exceeds the preset level, the server 10 outputs the second preliminary response message to the first chat room, and the security level of the second preliminary response message is If the level is lower than the preset level, the server 10 may output a second preliminary response message to the second chat room.

In this case, the security level may be set by the second user terminal 30 . As an embodiment, the second user terminal 30 may preset a keyword that requires security, and if the keyword exists in the second preliminary response message, the server 10 transmits the second preliminary response message to the first You can print it out in the chat room.

10 is a block diagram of an apparatus according to an embodiment of the present invention.

The processor 102 may include one or more cores (not shown) and a graphic processing unit (not shown) and/or a connection path (eg, a bus, etc.) for transmitting and receiving signals with and/or other components. .

The processor 102 according to an embodiment performs the method described with reference to FIGS. 2 to 8 by executing one or more instructions stored in the memory 104 .

For example, the processor 102 obtains new training data by executing one or more instructions stored in the memory, performs a test on the acquired new training data by using the learned model, and performs the test result, labeling Extracting the first learning data in which the obtained information is obtained with an accuracy greater than or equal to a predetermined first reference value, deleting the extracted first learning data from the new learning data, and removing the new learning data from which the extracted learning data is deleted It is possible to retrain the learned model using

On the other hand, the processor 102 is a RAM (Random Access Memory, not shown) and ROM (Read-Only Memory: ROM) for temporarily and / or permanently storing a signal (or data) processed inside the processor 102. , not shown) may be further included. In addition, the processor 102 may be implemented in the form of a system on chip (SoC) including at least one of a graphic processing unit, a RAM, and a ROM.

The memory 104 may store programs (one or more instructions) for processing and controlling the processor 102 . Programs stored in the memory 104 may be divided into a plurality of modules according to functions.

The steps of a method or algorithm described in relation to an embodiment of the present invention may be implemented directly in hardware, as a software module executed by hardware, or by a combination thereof. A software module may contain random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

The components of the present invention may be implemented as a program (or application) to be executed in combination with a computer, which is hardware, and stored in a medium. Components of the invention may be implemented as software programming or software components, and similarly, embodiments may include various algorithms implemented as data structures, processes, routines, or combinations of other programming constructs, including C, C++ , Java, assembler, etc. may be implemented in a programming or scripting language. Functional aspects may be implemented in an algorithm running on one or more processors.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can realize that the present invention can be embodied in other specific forms without changing its technical spirit or essential features. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

10 : Server
20: first user terminal
30: second user terminal

Claims (7)

In the control method of the chatbot system,
receiving, by the server, a query from the second user terminal through the first chat room in which the first user terminal and the second user terminal participate;
transmitting, by the server, the query to a first user terminal;
obtaining, by the server, a first preliminary response message by analyzing the query;
transmitting, by the server, the first preliminary response message to at least one of the first user terminal and the second user terminal according to a condition in which the first preliminary response message is preset; A control method comprising a. According to claim 1,
The control method is
obtaining, by the server, related data by analyzing the query; and
and transmitting, by the server, the related data to at least one of the first user terminal and the second user terminal. 3. The method of claim 2,
The step of obtaining the relevant data includes:
extracting keywords based on the query;
searching for first related data corresponding to the keyword in an internal DB of the server;
searching an external server for second related data corresponding to the keyword; and
transmitting at least one of the first related data and the second related data to at least one of the first user terminal and the second user terminal; including,
The control method is
receiving, by the second user terminal, at least one of the first related data and the second related data from the server; and
displaying, by the second user terminal, the at least one related data in a first area of the first chat room; including,
The related data displayed in the first area is data downloadable by the second user terminal and url information accessible by the user terminal. 4. The method of claim 3,
The step of transmitting to at least one user terminal of the first user terminal and the second user terminal,
obtaining a security level of the first related data;
obtaining a user rating of the second user terminal;
when the security level of the first related data is higher than the user level of the second user terminal, transmitting the first related data to the first user terminal; and
When the security level of the first related data is lower than the user level of the second user terminal, transmitting the first related data to the second user terminal;
Transmitting the first related data to the first user terminal comprises:
transmitting, by the server, the first related data to the second user terminal when a user command for transmitting the first related data to the second user terminal is input from the first user terminal; A control method comprising a. According to claim 1,
The preset condition is a condition for whether the first user terminal is connected to the server,
Transmitting the first preliminary response message to at least one of the first user terminal and the second user terminal comprises:
transmitting, by the server, the first preliminary response message to the first user terminal when the first user terminal is connected to the server;
obtaining, by the first user terminal, a first response message based on the first preliminary response message and transmitting the first response message to the server;
transmitting, by the server, the first response message to the second user terminal; and
Including, by the second user terminal, displaying the first response message in a second area of the first chat room;
When the first user terminal is not connected to the server,
transmitting, by the server, the first preliminary response message to the first user terminal and the second user terminal; and
and displaying, by the first user terminal and the second user terminal, the first preliminary response message in a second area of the first chat room. 4. The method of claim 3,
The step of obtaining the first related data includes:
obtaining personal information and past conversation history corresponding to the second user terminal;
acquiring the first related data based on the acquired personal information, past conversation history, and the keyword;
Inviting, by the server, to a second chat room including the first user terminal, the second user terminal, and the third user terminal when the first related data is the related data received from the third user terminal ; and
Control method comprising a; obtaining, by the server, the conversation contents of the second user terminal and the third user terminal collected through the second chat room. 7. The method of claim 6,
The step of obtaining the conversation contents of the second user terminal and the third user terminal,
obtaining, by the server, query information of the second user terminal through the second chat room;
obtaining, by the server, a second preliminary response message by analyzing the query;
obtaining, by the server, a security level of the second preliminary response message; and
outputting, by the server, the second preliminary response message to the first chat room when the security level of the second preliminary response message exceeds a preset grade; and
and outputting, by the server, the second preliminary response message to the second chat room when the security level of the second preliminary response message is less than or equal to a preset grade.

Images