KR20190059731A - Chatbot system based on multi-domain and method for providing chatbot service - Google Patents

Abstract

Disclosed are a multi-domain based chatbot system and a method for providing a chatbot service. The system includes a chatbot management server and a chatbot server. The chatbot management server stores and manages a plurality of routing information, which is an access path to a plurality of external chatbot servers, searches for an external chatbot server corresponding to an object name and an intention for a natural language query received from the outside and provides routing information on the searched external chatbot server. In order to answer to a user′s question, the chatbot server stores information on the basis of knowledge base or machine learning, searches for an answer corresponding to a natural language query received from a user terminal and provides the retrieved answer to the user terminal as a reply to the natural language. When the answer corresponding to the natural language query is not found, routing information corresponding to the object name and the intention of the natural language query is received through the chatbot management server, the natural language query is transmitted to the external chat server corresponding to the provided routing information and thereafter, an answer is received from the external chat server and provided to the user terminal.

Description

Field of the Invention < RTI ID = 0.0 > [0001] < / RTI > The present invention relates to a multi-

The present invention relates to a multi-domain based chatbot system and a method for providing a chatbot service.

Recently, as a result of expanding the usage environment of messenger via smartphone and development of artificial intelligence, customers and companies want to be connected through messaging service. Therefore, chatbot (messenger) ChatBot) system is in the spotlight. Such a chatbot system is originally a robot that communicates with a system that means a chatter robot system.

In addition, artificial intelligence chatbots that provide information by constructing a knowledge base based on artificial intelligence have been used in recent years due to the development of artificial intelligence (AI) technology, but such artificial intelligence chatbots also require additional information In this case, it is necessary to interwork with external chatbots only to the extent that communication interface protocol with external system supports it. This is due to the increasing number of chatbots per business domain, it is difficult for a single chatbot to cover the knowledge of all the subject areas that the user queries, and thus it is necessary to combine it with the knowledge of the chatbots built on other systems or platforms This is because interfaces that depend on communication protocols have limitations in terms of cost and schedule, such as the necessity of developing interfaces one by one.

A problem to be solved by the present invention is to solve problems of existing methods of developing an interface depending on a communication protocol when a user queries a knowledge domain that can not be covered by knowledge of a single chatbot, By acquiring the routing information of the target chatbot to be linked through the multi-chatbot information acquired based on the Named Entity Recognition (NER) and the Intent Classification, it is possible to inquire and reply to the external chatbot by the artificial intelligence And a method for providing a chattb service based on an advanced multi-domain.

In a chatbot system according to one aspect of the present invention,

A chattbot system, which stores and manages a plurality of routing information, which is an access path to a plurality of external chatbot servers, searches an external chatbot server corresponding to an object name and an intention of a natural language query received from the outside, Based on a knowledge base or a machine learning basis by answering a question of a chatbot management server and a user providing routing information for a natural language query received from a user terminal and searching for an answer corresponding to a natural language query received from a user terminal, Providing a routing information corresponding to an object name and an intention for the natural language query through the chattube management server when an answer corresponding to the natural language query is not retrieved, Corresponding external chatbot After passing by the natural language query receives a response that the service from the external server and a chaetbot chaetbot server that is available to the user terminal.

Here, the chatbot server may be an information database for storing information based on knowledge base or machine learning. An answer generator for searching for an answer corresponding to the natural language query using the information database, analyzes the natural language query, And an intention to be classified with respect to the natural language query term is transmitted to the chatbot management server and the entity name and the intention of the natural language query recognized by the analysis unit are transmitted from the chatbot management server to the chatbot management server, A response acquiring unit that acquires an answer corresponding to the natural language query through the external chatbot server after accessing an external chatbot server corresponding to the routing information; Found by Configuring the answers to the natural language query from the response obtained by the response obtaining unit to the side or to include a processing in response to provided to the user terminal.

In addition, when the search for an answer to the natural language query fails using the information database, the answer generator transmits an entity name and an intention for the natural language query to the routing information acquisition unit to acquire routing information corresponding to the natural language query do.

The chatbot management server may further include a routing table for storing and managing a plurality of routing information which is an access path to the plurality of external chatbot servers, and a routing table for storing the routing information, And a chatbot searching unit searching for an external chatbot server corresponding to the natural language query and providing routing information for the searched external chatbot server to the chatbot server.

The routing table may further include a response rate for the plurality of external chatbot servers, and the chatbot search unit may further transmit a response rate to the external chatbot server corresponding to the natural language query word to the chatbot server.

Also, when the response rate provided from the chattube management server is equal to or greater than a predetermined threshold, the chatbot server accesses an external chat server corresponding to the routing information provided with the response rate, and receives a response to the natural language query.

If the natural language query term received from the user terminal is a query having a plurality of intentions, the chatbot server generates a query term for each of the plurality of intentions, And reconstructs the response as a composite response, and provides the response to the user terminal in response to the natural language query.

Also, the chatbot server displays, through a display of the user terminal, a process of obtaining a response to the natural language query through a chat with the external chatbot server.

The step of acquiring an answer to the natural language query through the chat with the external chatbot server may include acquiring an answer to the natural language query through the chat window, Through the window displayed on the screen.

According to another aspect of the present invention,

A method of providing a chat service to a user terminal, the method comprising: searching for an answer through an information database with respect to a natural language query received from a user terminal, the information database storing information on a knowledge base or machine learning basis; Receiving routing information through a chatbot management server when an answer search through the information database fails for a natural language query, the chatbot management server storing and managing a plurality of routing information, which is an access path to a plurality of external chatbot servers, - receiving a response to the natural language query through an external chat server corresponding to the routing information provided from the chatbot management server, and receiving a response searched in the step of searching for the answer or an answer provided in the step of receiving the answerProcessed as a group in response to a natural language query phrase includes the step of providing to the user terminal.

Here, the chatbot management server further stores and manages the response rates for the plurality of external chatbot servers, and receives the routing information from the chatbot management server in addition to providing the response rate with the routing information Further comprising determining whether the response rate is equal to or greater than a predetermined threshold value between the step of receiving the routing information and the step of receiving the response, and determining whether the response rate is equal to or greater than a preset threshold value A step of receiving the answer when the answer rate is equal to or greater than the preset threshold value.

If the answer rate is smaller than the preset threshold value, an answer detection failure is provided to the user terminal in response to the query term.

Also, in the step of receiving the routing information, the chatbot server transmits an entity name and an intention for the natural language query term to the chatbot management server, and the chatbot management server, based on the object name and the intention of the natural language query, Search the server.

In a chatbot system according to another aspect of the present invention,

A chatbot system, comprising: a communicator, a memory, and a processor, wherein the communicator includes a user terminal, a chatbot management server, wherein the chatbot management server stores and manages a plurality of routing information that is an access path to a plurality of external chatbot servers, Wherein the memory is configured to store information on a knowledge base or machine learning basis and to store a set of codes for answering a user ' s query, the code comprising: An operation for receiving an answer corresponding to the natural language query term received from the chatbot management server through the communicator, and an operation for receiving, from the communicator, To the external chat server corresponding to the external chat server To execute an operation for providing the response via an operation of the communication to obtain the answers to salmon query term to the user terminal is used to control the processor.

The processor may further perform an operation of receiving an answer rate from the chattube management server together with the routing information through the communicator and an operation of determining whether the response rate is equal to or greater than a preset threshold value, Executes an operation to acquire an answer to the natural language query through the external chatbot server when the response rate is equal to or greater than the preset threshold value.

According to the present invention, a method for structuring natural language conversation processing between chatbot servers is performed when information is exchanged between artificial intelligent multi-chatbot servers, and when a chatbot server itself has necessary information, a target chatbot server to be queried is found by artificial intelligence, Can be easily exchanged between heterogeneous platforms by acquiring them through the natural language processing method between the chatbots server.

In addition, the users of the chatbot server need not to communicate with many chatbots server by area, and the chatbot server which received the first query searches the other chatbot server with artificial intelligence as needed to generate an answer, And it is easy to connect the chatbot server between the heterogeneous platforms, so that collective intelligence that connects the artificial intelligent chatbot server can be constructed.

1 is a schematic block diagram of a chatbot system according to an embodiment of the present invention.
FIG. 2 is a detailed block diagram of the chatbot server shown in FIG. 1. FIG.
3 is a detailed block diagram of the chatbot management server shown in FIG.
4 is a diagram schematically showing field information constituting the routing table shown in FIG.
5 is a schematic flowchart of a method of providing a chat service in accordance with an embodiment of the present invention.
FIG. 6 is a flowchart of a chat service method performed by a chat server according to an embodiment of the present invention.
FIG. 7 is a view schematically showing a display of a user terminal in a chatbot system according to an embodiment of the present invention, wherein (a) shows only a chat window with the chatbot server, (b) And a chat window with the server as a window.
8 is a schematic block diagram of a chatbot server according to another embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly explain the present invention, parts not related to the description are omitted, and like parts are denoted by similar reference numerals throughout the specification

Throughout the specification, when an element is referred to as " comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.

Also, the terms " part, " " module, " and " module ", etc. in the specification mean a unit for processing at least one function or operation and may be implemented by hardware or software or a combination of hardware and software have.

Hereinafter, a chatbot system according to an embodiment of the present invention will be described with reference to the drawings.

1 is a schematic block diagram of a chatbot system according to an embodiment of the present invention.

1, a chatbot system 10 according to an embodiment of the present invention includes a user terminal 100, a chatbot server 200, a chatbot management server 300, and an external chatbot server 400. 1, the present invention is not limited to FIG. 1, and the present invention is not limited to FIG. 1, and may be configured differently from FIG. 1 according to various embodiments of the present invention. .

The user terminal 100 is connected to the chatbot server 200 via the network 510. Herein, the network 510 mainly refers to a connection structure capable of wirelessly exchanging information between respective nodes such as terminals and servers as a wireless network. One example of such a network is a 3rd Generation Partnership Project (3GPP) , Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, 3G, 4G, 5G, and the like.

The user terminal 100 receives a query of the user and transmits the query to the chat server 200 via the network 510. The chat client 200 receives a response to the query from the chat server 200, Or the like.

The user terminal 100 may comprise a mobile terminal capable of wireless communication, and according to various embodiments of the present invention, the user terminal 100 may be various types of devices. For example, the user terminal 100 may be a portable terminal that can access the remote chat server 200 through the network 510. As an example of the portable terminal, there is a wireless communication device with guaranteed portability and mobility. The wireless communication device includes a 3G, 4G, 5G terminal, an LTE terminal, a smartphone, a smart pad, a tablet PC, a personal communication system (PCS) Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication) -2000, CDMA (Code Division Multiple Access) (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet) terminals, and the like. One example of the user terminal 100 is a notebook, a netbook, an UltraBook, a subnotebook, a desk note, an Ultra-Mobile PC (UMPC), a desktop ), And the like. Here, the user terminal 100 is not limited to the form shown in FIG. 1 or the ones illustrated above.

The chatbot server 200 searches for answers corresponding to natural language query terms received from the user terminal 100 via the network 510. When a corresponding answer is searched, the chatbot server 200 searches the user terminal 100 for answers to the query, Lt; / RTI >

However, when the chatbot server 200 can not retrieve the answer corresponding to the natural language query received from the user terminal 100, the chatbot server 200 transmits information on the natural language query to the chatbot management server 300, And transmits the natural language query to the external chat server 400 corresponding to the routing information and provides the response provided from the external chat server 400 to the user terminal 100. [ Herein, the information about the natural language query may include the entity name identified through the Named Entity Recognition technology and intentions classified through the Intent Classification technology for the natural language query. At this time, the object name identification technique is a technology for recognizing and extracting words corresponding to a person, a company, a place, a time, a unit, etc., which are defined in advance, and an object classification technique is applied to various types of natural language sentences The intention is classified into the intention and then the intention corresponding to the input natural language query word is found, which is well known, and a detailed description thereof will be omitted here.

There are various ways in which the chatbot server 200 searches for an answer corresponding to a natural language query. For example, when the chatbot server 200 is a knowledge base-based chatbot server, a pattern matching method for searching for an answer through pattern matching such as Artificial Intelligence Markup Language (AIML), an ontology for searching a word- You can search for answers in a way. Alternatively, if the chatbot server 200 is a machine learning-based chatbot server, learning data consisting of a large number of keys and values is learned to find an optimal machine learning function model, and then an API (Application Programming Interface ), It is possible to search the input value for the question, the output value for the answer, and the answer accuracy (%). When more than one question intention is obtained as a result of the query term analysis received from the user terminal 100, the chattob server 200 may search for answers found in the chattob server 200 itself, The reply provided from the external chat server 400 according to the information may be processed to generate a composite answer, and the generated composite answer may be provided to the user terminal 100 as a final answer.

The chatbot server 200 and the external chatbot servers 400 connected to use the chatbot service according to the embodiment of the present invention send the chatbot management server 300 to the chatbot management server 300 with their information . The information registered at this time may include an identifier for identifying the chat server and routing information for accessing the chat server, for example, a URL (Uniform Resource Locator) as an access address.

The chatbot management server 300 searches the chatbot server 200 for the information about the natural language query, that is, the information about the object name and the intention, that is, the transmitted information, that is, the object name and the intention, And provides the information to the chatbot server 200. At this time, the routing information provided from the chatbot management server 300 to the chat server 200 may be one or more. For example, the chatbot management server 300 can provide one or more routing information for one natural language query transmitted from the chatbot server 200, or if two or more natural language queries are transmitted from the chatbot server 200 Two or more pieces of routing information may be provided to the chatbot server 200. [

The external chatbot server 400 is composed of one or more chatbot servers and is connected to the chatbot server 200 through the network 520. The network 520 may be a wireless network, such as the network 510, or may be a wired network, such as the Internet.

The external chatbot server 400 has a configuration similar to that of the chatbot server 200 and searches for an answer to a natural language query input from the outside and provides the retrieved answer to the chatbot server 200. [

The external chatbot server 400 may be a chatbot server having a knowledge base on a topic similar to or different from the chatbot server 200. [ Most of the external chatbot servers 400 can be registered in the chatbot management server 300. Since the external chatbot server 400 is used when the search for the query term fails in the chatbot server 200, There is little possibility of being used in the case of the external chatbot server 400 providing information.

The external chatbot server 400 may also receive the routing information through the chattob management server 300 with respect to the query term when an answer search for the input natural language query fails, When the natural language query is received from the chatbot server 200 instead of receiving the query, the searcher may not perform the continuous search through the chatbot management server 300 even if an answer search for the received query term fails. That is, the external chatbot server 400 provides a failure response to the chatbot server 200 when an answer search fails for the natural language query received from the chatbot server 200.

Hereinafter, the operation of each component of the chatbot system 10 shown in FIG. 1 will be described in more detail.

2 is a detailed block diagram of the chatbot server 200 shown in FIG.

2, the chatbot server 200 includes a query receiver 210, a query analyzer 220, an information database 230, an answer generator 240, a routing information acquirer 260, An answer acquisition unit 260, an answer processing unit 270, and an answer provider 280.

The query term receiving unit 210 receives a natural language query from the user terminal 100 via the network 510.

The query analyzing unit 220 analyzes the natural language query received through the query receiving unit 210 and grasps the contents of the received natural language query. These include analysis based on object name identification technology, analysis based on intent classification, and so on. That is, it is determined whether a natural language query received through the query language receiving unit 210 receives a query having an intention classification based on the intention classification, and an object name corresponding to the received query.

The information database 230 stores information for providing answers to user's questions. The information database 230 may be configured to correspond to a method in which the chatbot server 200 searches for and provides an answer corresponding to a natural language query. For example, in the case where the chatbot server 200 is a knowledge base-based chatbot server, the information database 230 may store knowledge in a DB or memory so as to be able to search for an AIML-based pattern matching method and an ontology method, . Accordingly, the chatbot server 200 can search the knowledge graph configured in the information database 230 and find and provide answers corresponding to the natural language query. Or the chatbot server 200 is a machine learning-based chatbot server, the information database 230 comprises an optimal machine learning function model that has learned learning data composed of a large number of keys and values. Accordingly, the chatbot server 200 can receive an answer and answer accuracy, that is, reliability (%) corresponding to the natural language query, as an output value after performing an API query with the natural language query as an input value in the information database 230. The information database 230 according to an embodiment of the present invention may include one or both of a knowledge base for providing answers on the basis of the knowledge base or a machine learning model for providing answers on a machine learning basis.

The answer generation unit 240 searches for answers corresponding to the query input through the information database 230 as the analysis result by the query language analysis unit 220. When a plurality of query terms are analyzed by the query term analyzing unit 220, an answer search for each of a plurality of query terms can be performed.

The answer generator 240 searches the information database 230 for an answer that is matched to the query and extracts the answer as an answer to the query.

However, if an answer matching the query term is not found, the answer generator 240 transmits the entity name and intention to the routing information acquisition unit 260 for the natural language query whose answer is not found. Here, when the answer generating unit 240 performs an answer search for a plurality of query terms, the answer generating unit 240 transmits the answer query to the routing information acquiring unit 250 only for a query term for which an answer search has failed among a plurality of query terms.

Thereafter, the reply generator 240 transmits the routing information acquired by the routing information acquisition unit 250 to the response acquiring unit 260 together with the natural language query, and acquires the natural language query acquired through the response acquiring unit 260 And transmits the answer to the answer processing unit 270. [ At this time, if there is a search result from the information database 230 among a plurality of query terms, the answer generation unit 240 transmits the corresponding answer to the answer processing unit 270.

The routing information acquisition unit 250 transmits the entity name and intention of the natural language query transmitted from the response generation unit 240 to the chatbot management server 300, receives the routing information provided from the chatbot management server 300, To the generation unit (230). At this time, a plurality of pieces of routing information can be received from the chatbot management server 300, and all of the plurality of pieces of routing information thus received are transmitted to the response generation unit 230.

The response acquiring unit 260 receives the routing information and the natural language query transmitted from the response generating unit 240 and transmits the natural language query to the external chat server 400 corresponding to the received routing information through the network 520 A response to the natural language query is received from the external chat server 400, and the response is transmitted to the response generating unit 230. In this case, when receiving a plurality of pieces of routing information corresponding to one query word from the response generation unit 240, the response acquisition unit 260 acquires a plurality of routing information from the plurality of external chatbot servers 400 corresponding to the plurality of received routing information Transmits the query term, and receives a response to the query term, respectively, and transmits the response to the response generation unit 240. When a plurality of query terms and a plurality of routing information corresponding to each of the query terms are transmitted from the response generation unit 240, the response acquisition unit 260 accesses the corresponding external chat server 400 for each routing information, And transmits a plurality of answers respectively corresponding to the plurality of query terms to the response generation unit 240. The response generation unit 240 receives the plurality of answers corresponding to the plurality of query terms.

The answer processing unit 270 receives at least one answer transmitted from the answer generating unit 240, processes it into a reply corresponding to the natural language query, and transmits the answer to the answer providing unit 280. [ In this case, when there is one natural language query term retrieved from the answer generator 240, the answer processor 270 may provide the answer delivered from the answer generator 240 to the answer provider 280 as it is. However, if there are two or more natural language query terms retrieved by the answer generator 240, the two or more answers transmitted from the answer generator 240 are processed to correspond to natural language queries having two or more intentions, And transmits it to the reply providing unit 280. [

The answer providing unit 280 receives the answer transmitted from the answer processing unit 270 and provides the response to the user terminal 100 through the network 510. [

Those skilled in the art will appreciate that the query input unit 210, the query language analysis unit 220, the information database 230, the response generation unit 240, the routing information acquisition unit 260, the response acquisition unit 260, 270, and the answer provider 280 may be separately implemented, or one or more of them may be integrally implemented.

Next, the chatbot management server 300 shown in FIG. 1 will be described.

3 is a detailed block diagram of the chatbot management server 300 shown in FIG.

3, the chatbot management server 300 includes a routing table 310, a receiving unit 320, a chatbot searching unit 330, and a transmitting unit 340.

The routing table 310 is a table for storing and managing the information of the chatbots registered in the chatbot management server 300, in particular, the routing information. As shown in FIG. 4, the routing table 310 includes a chatbot identifier ID, a chat bot name, a manufacturer identifier (ID), a chat-bot access address (URL), an access range, The response rate, and the like. Here, the access range indicates disclosure or authentication, and the status indicates activation or deactivation of the chatbot server. Also, the response rate indicates the reliability of response when chatbot servers with knowledge base based on artificial intelligence knowledge base provide answers to questions from outside. This response rate may vary depending on the degree of learning in the AI knowledge base. For example, the response rate can also be high if the learning completeness of the artificial intelligence knowledge base is high. It will be appreciated by those skilled in the art that the configuration of the routing table 310 shown in FIG. 4 is only an embodiment of the present invention, so that various embodiments can be configured differently.

The receiving unit 320 receives the entity name and the intention of the natural language query transmitted from the chat server 200 and transmits the received entity name and intention to the chat search unit 330. The chatbot searching unit 330 searches the external chatbot server 400 corresponding to the query term in the routing table 310 using the entity name and the intention received from the receiving unit 320 and accesses the retrieved external chatbot server 400 Which is an access address, to the transmitting unit 340. The transmitting unit 340 receives the routing information, At this time, the chatbot searching unit 330 can perform searching in the routing table 310 using the object name and the intention received from the receiving unit 320. The chatbot searching unit 340 may determine an access range or state of the external chatbot server 400 retrieved from the routing table 310 and extract routing information only when routing information can be provided. For example, when the external chatbot server 400 is open or activated, the routing information can be extracted. The chatbot searching unit 320 also extracts not only the routing information from the routing table 310 but also other information about the searched external chatbot server 400 such as a chatbot ID and a chatbot name, .

The transmitting unit 330 provides the chatting server 200 with the routing information transmitted from the chatting search unit 330.

Hereinafter, a method for providing a chat service in accordance with an embodiment of the present invention will be described with reference to the drawings.

5 is a schematic flowchart of a method of providing a chat service in accordance with an embodiment of the present invention. The contents of the user terminal 100, the chat server 200, the chat server management server 300 and the external chat server 400 are described with reference to FIGS. 1 to 4, The present invention can also be applied to a method of providing information.

Referring to FIG. 5, the chatbot server 200 receives a natural language query from the user terminal 100 through the network 510 (S100).

Then, the chatbot server 200 searches for the corresponding answer through the information database 230 using the received natural language query (S110).

Then, it is determined whether an answer corresponding to the query term is searched (S120). If an answer corresponding to the query term is not found, the name and intention of the natural query term are transmitted to the chatbot management server 300 (S130).

The chatbot management server 300 searches the routing table 310 for the external chat server 400 corresponding to the object name and the intention of the natural language query transmitted from the chat server 200 in operation S140.

Thereafter, the chatbot management server 300 provides routing information about the searched external chatbot server 400 to the chatbot server 200 (S150).

Next, the chatbot server 200 accesses the external chat server 400 corresponding to the routing information using the routing information provided from the chatbot management server 300 (S160).

The external chatbot server 400 searches for an answer corresponding to the natural language query transmitted from the chatbot server 200 in step S170 and provides the retrieved answer to the chatbot server 200 as a reply corresponding to the natural language query in step S180, .

Accordingly, the chatbot server 200 provides the answer provided from the external chat server 400 to the user terminal 100 as a response to the natural language query received in the step S100 (S190).

Meanwhile, if an answer corresponding to the natural language query is found in step S120, the retrieved answer is provided to the user terminal 100 (S190).

Also, the external chat server 400 corresponding to the natural language query may not be searched in step S140, or an answer corresponding to the natural language query may not be searched in step S170. However, in the embodiment of the present invention, It is assumed that the external chat server 400 corresponding to the natural language query is searched in step S140 and an answer corresponding to the natural language query is searched in step S170. Alternatively, if the external chat server 400 corresponding to the natural language query is not found in step S140, or if an answer corresponding to the natural language query is not found in step S170, It can respond that the natural language query received from the terminal 100 can not provide an answer.

Next, the chatbot service method will be described in detail.

FIG. 6 is a flowchart of a chat service method performed by the chat server 200 according to an embodiment of the present invention.

Referring to FIG. 6, the chatbot server 200 receives a natural language query from the user terminal 100 through the network 510 (S200).

Thereafter, an answer corresponding to the received query word is retrieved using the information database 230 (S210). The answer may be a knowledge base based answer or a machine learning based answer, as described above.

If an answer corresponding to the query term is not found (S220), the name and intention of the natural language query word are transmitted to the chatbot management server 300 (S230).

Thereafter, the routing information and the response rate provided from the chatbot management server 300 are received (S240), and it is determined whether the received response rate is equal to or greater than a preset threshold value (S250).

If the response rate is equal to or greater than a preset threshold value, the connection is made to the external chat server 400 corresponding to the received routing information in step S260, and the natural language query received in step S200 is transmitted To the external chat server 400 (S270).

Thereafter, when an answer is received from the external chatbot server 400 (S280), the received response is provided to the user terminal 100 (S290).

However, if an answer corresponding to the natural language query is retrieved in step S220, the retrieved answer is provided to the user terminal 100 (S290).

If it is determined in step S250 that the response rate is smaller than the preset threshold, the integrated chat server 400 retrieved by the chattop management server 300 can not trust the integrated response. Accordingly, in step S100, And provides a detection failure response to the query term to the user terminal S100 (S300).

As described above, according to the embodiment of the present invention, even when the chat server 200 receiving the natural language query directly from the user terminal 100 can not retrieve the answer corresponding to the natural language query, the external chat server The user terminal 100 may receive a response to the natural language query from the external chat server 400 through the natural language processing method and the natural language processing method. At this time, an external chat server 400 capable of providing a response to a natural language query through the chatbot management server 300 managing the multi-chat server in order to find the external chat server 400 can be easily It is possible to find and answer the natural language query from the user terminal 100, so that the reliability of the chat service can be improved.

Therefore, in the embodiment of the present invention, when a user queries a knowledge domain that can not be covered by knowledge of a single chatbot server, routing information of a target chatbot server to be linked through information of the acquired multi- And can inquire the external chatbot server by the artificial intelligence itself, so that collective intelligence can be constructed.

In the chatbot system 10 according to the embodiment of the present invention, the user terminal 100 transmits only the natural language query to the chatbot server 200 and receives a response from the chatbot server 200. However, the technical scope of the present invention is not limited thereto, and when the chatter server 200 fails to search for an answer to the natural language query, the user accesses the external chat server 400 and finds an answer to the natural language query, The user terminal 100 can provide the process information to the user terminal 100 so that the user can know the process information.

For example, referring to Figure 7 (a), the display 110 of the user terminal 100 is shown. When the user accesses the chat server 200 through the user terminal 100 and inputs a natural language query through the chat window 111, the chat server 200 receives the natural language query through the chat service method described in FIG. The user can check the answer to the natural language query by searching the response to the input natural language query and providing the response to the user terminal 100 through the chat window 111. [ At this time, only the natural language query input by the user and the answers to the natural language query are displayed in the chat window 111 of the user terminal 100, and the user can access the natural chat query server 400 through the external chat server 400, I do not know how to get the answer by the material of.

7B, when a user accesses the chat server 200 through the user terminal 100 and inputs a natural language query through the chat window 111, and when the chat server 200 receives a natural language query, In the case where the answer to the query is unsuccessful, the process of receiving the routing information from the chatbot management server 300 and accessing the external chatbot server 400 and receiving answers through the material of the natural language query is performed in the chat window 111 And may be displayed by creating a window 112 of the window. That is, the chatbot server 200 generates a chat window for a query with the external chatbot server 400 in the chat window 111 as a chat window 112, and then sends a natural language query to the external chatbot server 400 through the chat window And the response provided from the external chat server 400 is also displayed through the window 112 so that the chat server 200 obtains an answer to the natural language query through the external chat server 400 and transmits the answer to the user terminal 100 So that the user can be informed of the process of providing it.

Alternatively, the window 112 may be generated in the display 110 as a chat window (not shown) separate from the chat window 111 without being generated in the chat window 111 to display the chat service process to the user It is possible.

FIG. 8 is a schematic block diagram of a chatbot server 600 according to another embodiment of the present invention.

8, a chatbot server 600 according to another embodiment of the present invention includes a communicator 610, an input device 620, a memory 630, a processor 640, and a bus 650.

The communicator 610 performs signal transmission with external components. The communicator 610 communicates with the user terminal 100 through the network 510 and directly communicates with the chatbot management server 300 and further communicates with the external chatbot server 400 ).

The input device 620 receives an input from the outside. These inputs may be data inputs or command inputs.

The memory 630 stores information for providing answers to user questions on a knowledge base or machine learning basis.

Thus, the input device 620 can receive inputs to add or change information in the memory 630, either knowledge-based or machine-based.

The memory 630 is also configured to store a set of codes, which code is used to control the processor 640 to perform the following operations. This operation includes an operation of receiving a query word from the user terminal 100 via the communicator 610, an operation of searching for an answer corresponding to the received query term, an operation of determining whether an answer corresponding to the query term is retrieved, An operation of transmitting the object name and the intention to the chatbot management server 300 through the communicator 610, the operation of receiving the routing information provided from the chatbot management server 300 through the communicator 610, An operation of connecting to the external chat server 400 via the communication unit 610 and an operation of transmitting a query word to the external chat server 400 through the communicator 610 and a response from the external chat server 400 through the communicator 610 And providing an answer to the query term via the communicator 610 to the user terminal 100.

Alternatively, the above operations may include receiving an answer rate from the chattvine management server 300 via the communicator 610, comparing the received response rate with a preset threshold value, comparing the received response rate with a preset threshold value And providing an answer detection failure response to the user terminal 100 via the communicator 610 if the answer is less than the predetermined threshold.

The operation of accessing the external chat server 400 through the communicator 610 according to the received routing information may be performed when the received response rate is equal to or greater than a preset threshold value.

Optionally, the above-described operations further include adding or modifying information in a memory 630 based on knowledge base or machine learning based on information input via input unit 620. [

The memory 630 may include a read only memory (ROM), a random access memory (RAM), and a non-volatile random access memory (NVRAM). The processor 640 may also be referred to as a controller, a microcontroller, a microprocessor, a microcomputer, or the like. In addition, the processor 540 may be implemented by hardware or firmware, software, or a combination thereof.

The bus 650 is configured to combine all the components of the chatbot server 600, i.e., the communicator 610, the input device 620, the memory 630, and the processor 640.

The embodiments of the present invention described above are not implemented only by the apparatus and method, but may be implemented through a program for realizing the function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, It belongs to the scope of right.

Claims (15)

As a chatbot system,
Storing and managing a plurality of routing information which is an access path to a plurality of external chatbot servers, searching an external chatbot server corresponding to an object name and intent for a natural language query received from the outside, and routing information for the external chatbot server Provided chatbot management server and
Storing information in a knowledge base or a machine learning basis to answer a user's query, searching answers corresponding to a natural language query received from a user terminal, and providing the retrieved answers to the user terminal as a response to the natural language query And if the answer corresponding to the natural language query is not found, receiving the routing information corresponding to the object name and the intention of the natural language query through the chattube management server and transmitting the natural language query to the external chat server corresponding to the provided routing information A chatbot server for receiving a response from the external chatbot server and providing the response to the user terminal,
. The method according to claim 1,
The chatbot server,
An information database for storing information based on knowledge base or machine learning;
An answer generator for searching for an answer corresponding to the natural language query using the information database;
An analyzing unit analyzing the natural language query to identify an entity name recognized for the natural language query and an intention to be classified for the natural language query;
A routing information acquisition unit for transmitting the entity name and intention of the natural language query term recognized by the analysis unit to the chattube management server and receiving routing information corresponding to the natural language query word from the chattube management server;
An answer acquiring unit for acquiring an answer corresponding to the natural language query through the external chat server after connecting to an external chat server corresponding to the routing information; And
And an answer processing unit for constructing an answer to the natural language query from an answer retrieved by the answer generating unit or an answer obtained by the answer obtaining unit,
. 3. The method of claim 2,
Wherein the response generation unit transmits an entity name and an intention for the natural language query term to the routing information acquisition unit to acquire routing information corresponding to the natural language query term when an answer search for the natural language query term fails using the information database,
Chambot system. The method according to claim 1,
The chatbot management server comprises:
A routing table for storing and managing a plurality of routing information as access paths to the plurality of external chatbots; And
A chatbot search unit for searching the external chatbot server corresponding to the natural language query word from the routing table based on the entity name and intention of the natural language query received from the chatbot server and providing routing information for the searched external chatbot server to the chatbot server,
. 5. The method of claim 4,
Wherein the routing table further includes a response rate for the plurality of external chatbot servers,
Wherein the chatbot searching unit further transmits an answer rate to an external chatbot server corresponding to the natural language query word to the chatbot server,
Chambot system. 6. The method of claim 5,
Wherein the chatbot server is connected to an external chat server corresponding to the routing information provided together with the response rate when the response rate provided from the chattube management server is equal to or greater than a preset threshold value and receives a response to the natural language query,
Chambot system. 7. The method according to any one of claims 1 to 6,
Wherein the chatbot server generates a query term for each of the plurality of intentions when the natural language query term received from the user terminal is a query having a plurality of intentions and then transmits an answer retrieved through the chatbot server for each query word, Processing the obtained answer to reconstruct a complex response, and providing the response to the user terminal in response to the natural language query,
Chambot system. 7. The method according to any one of claims 1 to 6,
Wherein the chatbot server displays, through a display of the user terminal, a process of obtaining an answer to the natural language query through a chat with the external chatbot server,
Chambot system. 9. The method of claim 8,
Wherein the step of acquiring an answer to the natural language query through the chat with the external chatbot server comprises the step of displaying a message in a part of the chat window displayed to provide the answer to the natural language query with the user terminal Being displayed through the window,
Chambot system. Retrieving an answer through an information database for a natural language query received from a user terminal, the information database storing information on a knowledge base or machine learning basis;
Receiving routing information through the chatbot management server when an answer search through the information database fails for the natural language query, the chatbot management server stores a plurality of routing information, which is an access path to a plurality of external chatbot servers, Management -;
Receiving a response to the natural language query through an external chat server corresponding to the routing information provided from the chatbot management server; And
A step of processing the answers found in the step of retrieving the answers or the answers provided in the step of receiving the answers as an answer to the natural language query and providing the processed result to the user terminal
And providing the service to the service provider. 11. The method of claim 10,
Wherein the chatbot management server further stores and manages a response rate for each of the plurality of external chatbot servers,
Receiving the routing information and the response rate from the chattube management server together with the routing information,
Between providing the routing information and receiving the answer,
Further comprising the step of determining whether the response rate is equal to or greater than a preset threshold value,
Performing the step of receiving the answer when the answer rate is equal to or greater than the predetermined threshold value in the step of determining whether the answer rate is equal to or greater than a preset threshold value,
How to provide chatbot service. 12. The method of claim 11,
And providing an answer detection failure to the user terminal in response to the query if the answer rate is less than the preset threshold value.
How to provide chatbot service. 12. The method of claim 11,
Wherein the chatbot server transmits an entity name and an intention for the natural language query to the chatbot management server in the step of receiving the routing information,
Wherein the chatbot management server searches an external chat server based on an object name and an intention of the natural language query,
How to provide chatbot service. As a chatbot system,
A communication device, a memory, and a processor,
The communicator includes a user terminal, a chatbot management server, wherein the chatbot management server stores and manages a plurality of routing information, which is an access path to a plurality of external chatbot servers, and communicates with an external chatbot server,
Wherein the memory is configured to store information on a knowledge base or machine learning basis to answer a user ' s query and to store a set of codes,
The code includes:
Retrieving an answer corresponding to a natural language query received from the user terminal through the communicator;
Receiving routing information corresponding to an entity name and intention for the natural language query from the chattube management server through the communicator;
Accessing an external chat server corresponding to the obtained routing information through the communicator to obtain an answer to the natural language query; And
And providing the response to the user terminal through the communicator
Lt; RTI ID = 0.0 > a < / RTI > processor,
Chambot system. 15. The method of claim 14,
The processor comprising:
Receiving an answer rate from the chattube management server together with the routing information through the communicator; And
Determining whether the response rate is equal to or greater than a preset threshold value
Lt; / RTI >
The processor comprising:
And if the answer rate is equal to or greater than the predetermined threshold value, acquiring an answer to the natural language query through the external chatbot server,
Chambot system.

Images