KR102148439B1 - Server, method, and system for providing chatbot service with adaptive development environment based on block unit combined api function and scenario - Google Patents

Abstract

Provided is a method for providing a chatbot service providing a block-based adaptive development environment combined with API functions and scenarios. The method includes the steps of: uploading, into at least one kind of store platform, at least one unit block combined with at least one scenario and an application programing interface (API) for performing the at least one scenario; if a user terminal makes a purchase of any one of the at least one unit block or at least one combination through the at least one kind of store platform, selecting the purchased one or the at least one combination; and if the user terminal places the at least one unit block selected on the user terminal onto a scenario builder through a drag and drop action, setting a chatbot, which is created as at least one combination of at least one function of the at least one unit block combined on the scenario builder forms one scenario, to be operated on a platform designated by the user terminal. According to the present invention, an adaptive and flexible platform can be provided since scenarios and functions are all included in a block.

Description

A chatbot service providing server, method, and system that provides a block-based adaptive development environment that combines AI functions and scenarios {SERVER, METHOD, AND SYSTEM FOR PROVIDING CHATBOT SERVICE WITH ADAPTIVE DEVELOPMENT ENVIRONMENT BASED ON BLOCK UNIT COMBINED API FUNCTION AND SCENARIO}

The present invention relates to a chatbot service providing server that provides a block-based adaptive development environment in which API functions and scenarios are combined. Flexible development capable of creating a chatbot by simply combining unit blocks by dragging and dropping on a scenario builder Provide a platform.

Natural language-based conversational AI or chatbots have a structure in which users can inquire directly to find information and acquire information proactively. Chatbots are emerging as the next generation social platform thanks to the proactive information acquisition structure and the expansion of various communication channels such as social network services and generalized smartphones. Chatbots currently commercially available are mostly rule-based systems, and responses are made based on pre-defined words and patterns.Therefore, if the patterns are not predefined and learned by the system, they cannot respond to user queries. This happens. In order to develop chatbots that are suitable for the scenarios they want, even though chatbots with the same platform and mechanism exist, companies or operators that provide chatbots cannot utilize them and have to re-develop them from scratch, so chatbot development always starts from the origin.

At this time, for the chatbot developer, a method for producing a chatbot with one block without separately building a number of separate parameters according to each parameter has been researched and developed. In this regard, Korean Patent Registration No. 10-1944353 (2019) Announcement on April 17th), when at least one sentence is input from the producer terminal, parameter information, which is attribute information for each word included in at least one sentence, is extracted for each word, and two or more parameters selected by the producer terminal Grouping of information is performed, and the chatbot produced by the producer terminal is driven by the user terminal accessing the chatbot service server, and the chatbot is one of two or more parameter information grouped in advance from the sentences of the chat message input by the user terminal. When any one parameter information is extracted, a configuration for executing a preset command by referring to the extracted parameter information is disclosed.

However, even if the above-described configuration is used, it is only possible to create a scenario when creating a scenario to be serviced by the chatbot, and it is only a configuration that simply aggregates the functions of each parameter. Due to the lack of a platform, adaptive development of chatbots goes back to the origin. In addition, even if the chatbot was developed by investing a large amount of money and manpower, the scenario of the chatbot was created based on the developer's thoughts, so if there is a problem or something to be added, the scenario must be changed again. However, maintenance costs and manpower input are inevitable, and losses are incurred due to failure to provide an adaptive chatbot. Accordingly, there is a need for a chatbot development platform that can adaptively develop and maintain not only system administrators and developers, but also individual companies or business owners according to business types and circumstances.

According to an embodiment of the present invention, by creating a chatbot using at least one unit block uploaded by a user's drag and drop on a scenario builder, not only the developer but also the general public can create a desired chatbot with a combination of blocks combined with the scenario. It is a structure in which not only system administrators but also general people can freely create and sell blocks, so anyone can generate revenue, and because the block itself contains both scenarios and functions, it is possible to provide an adaptive and flexible platform. , It is possible to provide a chatbot service providing method that provides a block-based adaptive development environment in which API functions and scenarios are combined. However, the technical problem to be achieved by the present embodiment is not limited to the technical problem as described above, and other technical problems may exist.

As a technical means for achieving the above-described technical problem, an embodiment of the present invention is a combination of at least one scenario and an API (Application Programming Interface) for executing at least one scenario to at least one kind of store platform. Uploading at least one unit block, when purchasing any one or at least one combination of at least one unit block from a user terminal through at least one type of store platform, the purchased one or at least one combination At least one of at least one function of the at least one unit block combined on the scenario builder when the selected step, and when at least one unit block selected from the user terminal is uploaded on the scenario builder by a drag-and-drop method from the user terminal. It includes the step of setting the chatbot to be executed on the platform designated by the user terminal as the combination forms one scenario.

Here, when the user terminal selects any one unit block or a plurality of unit blocks, the chatbot service providing server generates a chatbot in which the selected unit block and the natural language processing module are combined and transmits it to the user terminal. I can.

Another embodiment of the present invention is an upload unit for uploading at least one unit block to which at least one scenario and an API (Application Programming Interface) for executing at least one scenario are combined to at least one kind of store platform, at least Any one or at least one combination of at least one unit block in which one scenario and an API for executing at least one scenario are combined is purchased from a user terminal through at least one kind of store platform, any one purchased Alternatively, when at least one unit block selected from the user terminal is placed on the selection unit for selecting at least one combination, and at least one unit block selected from the user terminal on the scenario builder, at least one unit block combined on the scenario builder It includes a setting unit for setting the chatbot on which one function is implemented to run on the platform designated by the user terminal.

Another embodiment of the present invention provides at least one unit block in which at least one scenario and an API (Application Programming Interface) for executing at least one scenario are combined, and when the unit block is selected, on the scenario builder A chatbot service providing server that generates a chatbot by uploading the selected unit block, and a combination of at least one unit block or at least one unit block are selected and purchased from at least one type of store platform and placed on the scenario builder, It includes a user terminal that drives or shares a chatbot generated based on at least one unit block.

According to any one of the above-described problem solving means of the present invention, by creating a chatbot using at least one unit block uploaded by a user's drag and drop on the scenario builder, not only the developer but also the general public can use the block to which the scenario is combined. Anyone can create and sell blocks freely as well as system administrators as well as system administrators can freely create and sell blocks, and since the block itself contains both scenarios and functions, it is adaptive and flexible. Platform can be provided.

1 is a view for explaining a chatbot service providing system that provides a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a chatbot service providing server included in the system of FIG. 1.
3 is a view for explaining an embodiment in which a chatbot service providing a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention is implemented.
FIG. 4 is a diagram illustrating a process of transmitting and receiving data between components included in a chatbot service providing system providing a block-based adaptive development environment in which the API functions and scenarios of FIG. 1 are combined according to an embodiment of the present invention. to be.
5 is a flowchart illustrating a method for providing a chatbot service providing a block-based adaptive development environment in which API functions and scenarios are combined according to an 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 of ordinary skill in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and similar reference numerals are assigned to similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only "directly connected" but also "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included, and one or more other features, not excluding other components, unless specifically stated to the contrary. It is to be understood that it does not preclude the presence or addition of any number, step, action, component, part, or combination thereof.

The terms "about", "substantially" and the like, as used throughout the specification, are used in or close to the numerical value when manufacturing and material tolerances specific to the stated meaning are presented, and are used in the sense of the present invention. To assist, accurate or absolute figures are used to prevent unfair use of the stated disclosure by unscrupulous infringers. As used throughout the specification of the present invention, the term "step (to)" or "step of" does not mean "step for".

In the present specification, the term "unit" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Further, one unit may be realized using two or more hardware, or two or more units may be realized using one hardware.

In this specification, some of the operations or functions described as being performed by the terminal, device, or device may be performed instead in a server connected to the terminal, device, or device. Likewise, some of the operations or functions described as being performed by the server may also be performed by a terminal, device, or device connected to the server.

In this specification, some of the operations or functions described as mapping or matching with the terminal means mapping or matching the unique number of the terminal or the identification information of the individual, which is the identification information of the terminal. Can be interpreted as.

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

1 is a view for explaining a chatbot service providing system that provides a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention. Referring to FIG. 1, a chatbot service providing system 1 providing a block-based adaptive development environment in which API functions and scenarios are combined, at least one user terminal 100, a chatbot service providing server 300, at least one May include a shopping mall server 400 of. However, since the chatbot service providing system 1 that provides a block-based adaptive development environment in which the API functions and scenarios of FIG. 1 are combined is only an embodiment of the present invention, the present invention is limited to interpretation through FIG. It does not become.

In this case, each component of FIG. 1 is generally connected through a network 200. For example, as shown in FIG. 1, at least one user terminal 100 may be connected to the chatbot service providing server 300 through the network 200. In addition, the chatbot service providing server 300 may be connected to at least one user terminal 100 and at least one shopping mall server 400 through the network 200. In addition, at least one shopping mall server 400 may be connected to the chatbot service providing server 300 through the network 200.

Here, the network refers to a connection structure in which information exchange is possible between respective nodes such as a plurality of terminals and servers, and examples of such networks include RF, 3rd Generation Partnership Project (3GPP) network, and Long Term (LTE). Evolution) network, 5GPP (5th Generation Partnership Project) network, WIMAX (World Interoperability for Microwave Access) network, Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network) , PAN (Personal Area Network), Bluetooth (Bluetooth) network, Wi-Fi, NFC network, satellite broadcasting network, analog broadcasting network, DMB (Digital Multimedia Broadcasting) network, Visible Light Communication (VLC), LiFi and the like are included, but are not limited thereto.

In the following, the term “at least one” is defined as a term including the singular number and the plural number, and even if the term “at least one” does not exist, each component may exist in the singular or plural, and may mean the singular or plural. It will be self-evident. In addition, it will be possible to change according to the embodiment that each component is provided in a singular or plural.

At least one user terminal 100 is a general person or seller who creates or develops a chatbot using a web page, app page, program or application related to a chatbot service that provides a block-based adaptive development environment in which API functions and scenarios are combined. , It may be a terminal such as a developer or a company. At this time, the at least one user terminal 100 may be a terminal of a user who creates, edits, or transforms his own chatbot through an interface for selecting a desired unit block when creating or developing a chatbot. In addition, at least one user terminal 100 transmits product data posted on the pages of the plurality of shopping mall servers 400 to the chatbot service providing server 300, and answering questions recommended from the chatbot service providing server 300 It may be a terminal that receives the data set and selects, transforms, or edits any one of the Q&A data sets. In addition, the at least one user terminal 100 may be a terminal that generates a chatbot by including the selected Q&A data set in a scenario.

Here, the at least one user terminal 100 may be implemented as a computer capable of accessing a remote server or terminal through a network. Here, the computer is, for example, a laptop with navigation, a web browser, a desktop, a laptop, a VR HMD (for example, HTC VIVE, Oculus Rift, GearVR, DayDream, PSVR, etc.) ), etc. In this case, the at least one user terminal 100 may be implemented as a terminal capable of accessing a remote server or terminal through a network. Here, VR HMD is for PC (for example, HTC VIVE, Oculus Rift, FOVE, Deepon, etc.), for mobile (for example, GearVR, DayDream, Storm Horse Racing, Google Cardboard, etc.) and console (PSVR) and It includes all standalone models (eg, Deepon, PICO, etc.) implemented independently. At least one user terminal 100, for example, as a wireless communication device that is guaranteed portability and mobility, navigation, PCS (Personal Communication System), GSM (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)-2000, W-CDMA(W-Code Division Multiple Access), Wibro(Wireless Broadband Internet) ) All kinds of handheld-based wireless communication devices such as terminals, smartphones, smart pads, and tablet PCs, wearable devices, and Bluetooth (BLE, Bluetooth Low Energy) ), NFC (Near-Field Communication), RFID (Radio-Frequency IDentification), Ultrasound (Ultrasonic), infrared, Wi-Fi, Li-Fi (LiFi) can include various devices equipped with communication modules such as.

The chatbot service providing server 300 may be a server that provides a chatbot service web page, an app page, a program, or an application that provides a block-based adaptive development environment in which API functions and scenarios are combined. In addition, the chatbot service providing server 300 may be a server that provides an interface through which at least one user terminal 100 can produce or develop a chatbot according to its intention and scenario. To this end, the chatbot service providing server 300 combines not only a scenario, but also an API (Application Programming Interface) capable of executing it, and provides it to at least one user terminal 100 as a unit block, and the user terminal 100 It may be a server that allows you to create a chatbot with a desired function just by selecting a unit block, that is, without knowledge about the program, without coding, or starting from the origin. To this end, the chatbot service providing server 300 combines at least one type of scenario and an interface made to control functions provided by an operating system or a programming language together in a block so that it can be used in an application program. It can be the server that built the database. And, the chatbot service providing server 300, when inputting at least one shopping mall page in which a product is registered in the user terminal 100, collects and analyzes product data uploaded to at least one shopping mall page, and queries suitable for product data. It may be a server that recommends a response data set. In addition, when the recommended Q&A data set is selected from the user terminal 100, the chatbot service providing server 300 inserts a unit block using the selected Q&A data set as a scenario into the chatbot to respond to customer inquiries. It may be a server that creates a provided chatbot. Here, in order to extract the Q&A data set, the chatbot service providing server 300 may be a server including an engine that analyzes text or images.

Here, the chatbot service providing server 300 may be implemented as a computer that can access a remote server or terminal through a network. Here, the computer may include, for example, a navigation system, a notebook equipped with a web browser, a desktop, a laptop, and the like.

The at least one shopping mall server 400 may be a server of a shopping mall using a chatbot service related web page, an app page, a program, or an application that provides a block-based adaptive development environment in which API functions and scenarios are combined. In this case, the at least one shopping mall server 400 may be a server that manages a shopping mall page of the user terminal 100 and user information accessible to the shopping mall page. In addition, at least one shopping mall server 400 may be a server that allows when accessing the account of the user terminal 100 and accessing the shopping mall page from the chatbot service providing server 300, and answering questions to the shopping mall page. It may be a server that inputs and stores data registered in the chatbot service providing server 300 in the frame.

Here, the at least one shopping mall server 400 may be implemented as a computer that can access a remote server or terminal through a network. Here, the computer may include, for example, a navigation system, a notebook equipped with a web browser, a desktop, a laptop, and the like.

FIG. 2 is a block diagram illustrating a chatbot service providing server included in the system of FIG. 1, and FIG. 3 is a block diagram for providing a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention. A diagram for explaining an embodiment in which a chatbot service is implemented.

Referring to FIG. 2, the chatbot service providing server 300 may include an upload unit 310, a selection unit 320, and a setting unit 330.

The chatbot service providing server 300 or another server (not shown) that operates in conjunction with the chatbot service providing server 300 according to an embodiment of the present invention is used as at least one user terminal 100 and at least one shopping mall server 400 for API functions and scenarios. When transmitting a chatbot service application, program, app page, web page, etc. that provide a block-based adaptive development environment combined with, at least one user terminal 100 and at least one shopping mall server 400 Chatbot service applications, programs, app pages, web pages, etc. that provide a block-based adaptive development environment in which functions and scenarios are combined can be installed or opened. In addition, a service program may be driven in at least one user terminal 100 and at least one shopping mall server 400 by using a script executed in a web browser. Here, the web browser is a program that enables you to use the web (WWW: world wide web) service, and refers to a program that receives and displays hypertext described in HTML (hyper text mark-up language). For example, Netscape , Explorer, chrome, etc. In addition, the application means an application on the terminal, and includes, for example, an app that is executed on a mobile terminal (smartphone).

Hereinafter, before describing the chatbot service according to an embodiment of the present invention with reference to FIG. 2, the concept of the chatbot will be briefly described.

Chatbot is a compound word of chat and robot, and refers to an interactive system that interacts with people based on artificial intelligence, a technology that combines the artificial intelligence of a robot with an interactive interface. With the advancement of artificial intelligence technology for chatbots, machine learning or deep learning technology is applied to chatbots to understand users' questions, understand their intentions, and provide appropriate answers, making it possible to perform more complex tasks. In addition, as the spread of smartphones and mobile messenger users increase rapidly, most smartphone users use messenger apps, so if you provide chatbot services using messenger apps, you can provide chatbot services to users in a familiar environment without installing new apps. Became possible. Accordingly, existing messenger companies are developing a messenger platform with artificial intelligence technology as well as opening their own messenger platform APIs to preoccupy the chatbot market.

Chatbots can be classified into rule-based chatbots and machine learning-based chatbots according to their operation method, one-time Q&A chatbots and continuous conversational chatbots according to information exchange method, and search model chatbots and generation model chatbots according to the answer generation method. The rule-based chatbot operates based on a predefined rule, and includes an interpretation rule that interprets the user's input, a reaction rule that responds to the user's input, or a rule that generates a response. Rule-based chatbot development does not require a large amount of data and is relatively easy to implement.For rule-based chatbots, high-quality conversation services are possible if rules are well defined, but a lot of time and manpower are required to well define many rules. do.

Machine learning-based chatbots use machine learning algorithms to understand natural language and generate or select appropriate answers. The NLP (Natural Language Process) module decomposes the natural language input by the user by using various linguistic resources, and the NLU (Natural Language Understanding) module understands the meaning of the decomposed natural language and answers refined information. It is passed to the final classification model or generation model for generation. Furthermore, by using the result of the NLU module, the machine learning algorithm selects a previously defined answer or the NLG (Natural Language Generation) module generates a new answer. Machine learning-based chatbots can communicate at a high level, but require a high level of natural language processing technology, and the NLP module and NLU module require many linguistic resources and data (corporate dictionary, morpheme dictionary, semantic translation dictionary).

In addition, machine learning modules made to perform a specific task often need to be newly developed depending on the application field, and most are difficult to recycle. One-time Q&A chatbots do not require a conversation management function and operate in a similar form to a search engine that presents the most relevant answers to questions. Most of the FAQ chatbots that provide fixed answers to frequently asked questions are designed in a one-time Q&A format. Continuous conversational chatbots have a conversation start and end, collect information from users to achieve a specific purpose, and end the conversation when a final response is made based on the collected information. Since it is impossible for chatbots to respond to various topics, most of them have a structured conversation format (scenario) depending on the topic of a specific conversation. A chatbot system that operates according to a predetermined scenario is called a frame-based system. The frame-based discourse system constructs a slot, which is a set of information for answering a user's question, generates a question based on the slot, and develops the discourse according to the state in which the requested information is collected.

Search-based chatbots work by selecting and presenting the most appropriate answer among answers stored in the existing knowledge base according to the user's question. Generation-based chatbots generate and provide new answers that are not previously defined according to user questions. The generative model chatbot tends to make a grammatical error or generate an unusual answer because it has a large degree of freedom to generate an answer, and a very large amount of learning data is required to create a generative model chatbot. Currently, most commercially available chatbots are search-based chatbots or hybrid methods that combine a search model and a generation model, and recently, many researchers are conducting research to create a chatbot by applying a deep learning algorithm (Encoder-Decoder, Seq2Seq).

An embodiment of the present invention will be described with reference to FIG. 2 based on the above-described concept. Hereinafter, contents not described in FIG. 2 will be omitted from overlapping parts that are the same, similar, or premised on the above-described contents.

The upload unit 310 may upload at least one unit block to which at least one scenario and an application programming interface (API) for executing the at least one scenario are combined to at least one type of store platform. In this case, the at least one unit block may be a block corresponding to a function or design of at least one block constituting the chatbot. This may be thousands to tens of thousands, but the number of types is not limited. In this case, the block corresponding to the function may be a block including a map view, a delivery inquiry, a payment inquiry, a company extension number confirmation, and a Q&A Excel data push, but the function is not limited to those listed. Further, the block corresponding to the design may be a block corresponding to the skin of the chatbot, but is not downplayed because it is not listed.

The selection unit 320 selects from the user terminal 100 any one or at least one combination of at least one scenario and at least one unit block in which an API (Application Programming Interface) for executing at least one scenario is combined I can receive it. At this time, the scenario is a service unit that the user can experience within the chatbot. Assuming that there is a chatbot that provides services related to finance, for example, the service unit that this chatbot can provide, that is, the scenario is deposit, savings, loan, It can be pension, bancassurance, etc. The unit block is responsible for the function of this scenario, and the scenario is a combination of these. Unit blocks of A-B-C can be combined to create a scenario for D, or unit blocks of A-C-B can be combined to create a scenario for E. A scenario is a scenario that can be performed in various ways depending on the number of combined order or combination method, and it is a unit block. In the scenario builder, the smallest unit that responds to the user's intent is called a unit block, and one scenario is formed by gathering various blocks. That is, the chatbot operator can systematically manage a number of unit blocks by grouping them by desired service in this scenario unit.

In this case, as described above, the scenario is the number of each case that the user can query when using the chatbot and the answer to it in the form of an algorithm.As described above, in the scenario, the developer or the producer It is impossible to grasp, the situation of querying for each customer varies, and there are various variables, and the desired scenario is different depending on what kind of chatbot each user (client, seller, developer) creates, so any one chatbot was created. It was almost impossible to reuse it. Accordingly, an embodiment of the present invention provides a unit block by combining each unit scenario and an API that is a function for performing it, thereby providing a platform for creating a chatbot simply by combining or combining each unit block. Can provide. In this case, the API may be a set of subfunctions, protocols, and tools for creating an application, and may include a communication method between various components. This is because a pre-programmed module is provided so that the scenario is actually driven on a smartphone, etc., so if an API is provided together, it becomes possible to create a chatbot without writing a program according to the desired scenario.

At this time, the user terminal 100 outputs a UI/UX that allows the unit block to be simply uploaded by dragging and dropping the unit block on the scenario builder. The unit blocks uploaded by drag and drop by the user terminal 100 are The service intended by the user is provided by itself or in combination with other unit blocks. In this way, even if you are not a programmer or a general person, you can easily create a scenario with a combination of unit blocks and create a chatbot that implements it. For example, assuming that the user is a seller who sells goods through e-commerce, the functions he needs, for example, assuming that he wants to replace the customer center staff with a chatbot, is a question-and-answer function for the delivery process, and guidance. It can be created by including a viewing function, a shipping tracking query answering function, and a payment inquiry query answering function. Of course, the functions are not limited to those listed.

When the at least one unit block selected by the user terminal 100 is uploaded on the scenario builder by a drag-and-drop method from the user terminal 100, the setting unit 330 A chatbot in which at least one function is implemented may be set to run on a platform designated by the user terminal 100. Here, the function corresponding to the at least one unit block may include at least one function usable in the chatbot, and the at least one function may include at least one function required for e-commerce transactions.

In this case, when the user terminal 100 selects any one unit block or a plurality of unit blocks from the user terminal 100, the chatbot service providing server 300 provides a chatbot in which the selected unit block and the natural language processing module are combined. It can be generated and transmitted to the user terminal 100. In addition, the chatbot service providing server 300 may provide a sharing platform in which the selected unit block is shared for a fee or for free when a unit block is selected in the user terminal 100. The chatbot service providing server 300 collects product data of products registered on at least one shopping mall page from the user terminal 100, performs pre-processing, analyzes the product data, and stores a previously stored Q&A data set at the user terminal ( 100), and when the recommended Q&A data set is selected in the user terminal 100, a chatbot may be generated to correspond to the selected Q&A data set.

In addition, the chatbot service providing server 300 may pre-process, interpret, and output a result corresponding to the intention of the natural language included in at least one scenario. The task mainly performed in the natural language process of the chatbot service providing server 300 is morpheme analysis of natural language. The morpheme analysis (POS-tagging) refers to the work of dividing the primitive corpus into morpheme units and substituting part-of-speech information on each morpheme. Morphology, root, prefix/suffix, part-of-speech (POS, part-ofspeech), etc. The structure of the attribute can be grasped, so that the customer or the client using the chatbot responds to the query according to the scenario. And, in the natural language understanding step of the chatbot service providing server 300, after decomposing the input of the user (chatbot user), the process of determining an appropriate syntax and semantic system for the decomposed objects can be performed. have. At this time, in the natural language understanding step of the chatbot service providing server 300, processing such as general entity name recognition, detailed entity name recognition, vocabulary meaning analysis, syntax analysis, intention analysis, and the like may be performed. When extracting the entity name, in order to operate the chatbot in a specific domain or task, a specific entity name recognizer (Specific-NER) can be further used, and for entity name recognition or semantic determination, for example, Hidden Markov model (Hidden Markov model). model), Support Vector Machines, Conditional Random Fields, etc., as well as deep learning, i.e., FFNN (Feed-Forward Neural Network), CNN (Convolutional Neural Network), LSTM You can also use (Long Short Term Memory). Finally, in the natural language generation step of the chatbot service providing server 300, a natural language sentence corresponding to the grammar and dictionary is generated from the knowledge expression inside the computer for communication between humans and computers. The process can be carried out.

The process of natural language generation can be divided into three parts: first, the component part of the structure representing the information to be conveyed, which determines what to say, and second, the dialogue structure to determine the order of the sentences and the rules for sentences. It may include a part to be applied, and thirdly, a part to apply information on words and a sentence-based rule in order to generate an actual sentence. And, in this case, when the natural language processing function of the chatbot service providing server 300 is combined with the scenario builder, it can be generated as a search-type chatbot and a generation-type chatbot according to the response generation method of the chatbot, which is a variant of RNN among the deep learning architecture. Generating chatbots that automatically generate answers using Dual-encoder, Seq2seq, etc. may be created. In the structure of seq2seq, the encoder receives a query as an input and is encoded in the hidden layer, and the decoder receives the answer as an input and learns about the question and answer. In addition, word-embedding for learning the meaning of words may be further added.

When the user terminal 100 selects any one unit block or a plurality of unit blocks from the user terminal 100, the chatbot service providing server 300 generates a chatbot in which the selected unit block and natural language processing function are combined. It can be transmitted to the user terminal 100. For example, suppose a user is a seller of men's sneakers, and you create a chatbot to guide and answer questions for sales. When a customer enters a purchase budget, the chatbot should recommend a suitable product. For example, if the customer tells us that they are looking for "men's sneakers for less than KRW 100,000", it asks for dimensions, color, style, etc. You can find a sales list and recommend it. To do this, a product database corresponding to price, color, dimension, and style must be established, and the unit block of the chatbot must be defined to perform a function of extracting a product list corresponding to information received from customers. Accordingly, a chatbot that searches for a hotel, a chatbot that recommends cosmetics, a chatbot that recommends clothes, and a chatbot that analyzes customer receipts to determine the details of regular payments can be implemented, and each unit block can be integrated into one chatbot. It can also be implemented by putting it in, and you can create your own chatbot simply by editing the unit blocks freely.

In addition, the chatbot service providing server 300 generates a chatbot with a unit block selected by the user terminal 100, but may further provide items that should be premised by default. For example, when creating or developing a chatbot, the most avoidable thing is the situation where the conversation ends with "I didn't understand". In scenarios, such as when the customer just answers the question "What model do you want?" Even with undefined or unpredictable answers, you should continue to ask appropriate questions and continue the conversation. In addition, even if there is no item you are looking for, you must show a similar product and induce other choices, and if the situation where the chatbot fails to answer appropriately is repeated, the customer will say "ha, frustrated" or utter a swearword or sigh. In this case, rather than continuing the conversation, the chatbot should immediately connect with an ARS agent to prevent the customer from leaving the shopping mall page, and these unit blocks can be provided arbitrarily or by default.

In addition, the chatbot service providing server 300 may respond based on predefined words and patterns, as well as provide a gradual learning style even if it is not a pattern that is predefined and learned in the system. This is a method to improve the performance of the system by determining the intention based on the similarity in the system even if the customer's query is not defined and the intention of the question cannot be determined. In addition, by newly self-learning the corresponding query and intention, it performs wider query processing and quick response processing to future inquiries. First of all, the intent that the chatbot did not understand is first extracted and determined as the most similar intent to the corresponding combination. For example, the query "Can I put the sweater in cold water and run it in the washing machine?" If the intent is not understood by the chatbot, that is, the intent is not understood because it is not in the scenario or there is no information, the entities such as "sweater", "cold water", "detergent", "washing machine", and "turn" are extracted, and the above It maps to the intention of "How to wash the sweater" containing the combination, and responds to the customer with information on how to wash the sweater from the database. In addition, if the chatbot learns a query that does not understand the intention, and is configured to respond to similar inquiries by itself, the chatbot evolves itself through the process, and the customer's query is expected and learned in advance. The process is unnecessary, and responses can be generated centered on entities included in the query. Therefore, as more customers use it, performance improvement can be expected.

The chatbot service providing server 300 may provide a sharing platform in which, when a unit block is produced in the user terminal 100, the produced unit block is shared for a fee or for free. For example, suppose a user (seller) who sells a female beach dress creates a unit block of A, B, C, and D, and creates a chatbot with A, B, C, and D unit blocks. At this time, because A, B, C, and D contain information to be provided to customers when selling women's beach dresses, or functions to answer or request a customer's desired query, when other clothing sellers use it, It is not necessary to include only A+B+C+D, and E may be included or any one of A to D may be transformed. Accordingly, rather than sharing the chatbot itself, each unit block, which is a raw material for making it, can be shared separately, thereby making it easier for other users to create a chatbot, and reducing the trouble of writing a program from the beginning.

The chatbot service providing server 300 collects product data of products registered on at least one shopping mall page from the user terminal 100, performs pre-processing, analyzes the product data, and stores a previously stored Q&A data set at the user terminal ( 100), and when the recommended Q&A data set is selected in the user terminal 100, a chatbot may be generated to correspond to the selected Q&A data set. At this time, the recommendation module 350 includes a text analysis engine, an OCR text engine, a similar image search engine, and a category analysis engine for recommending a previously stored Q&A data set to the user terminal 100 by analyzing product data. I can.

In addition, the chatbot service providing server 300 according to an embodiment of the present invention, a hierarchical recurrent encoder-decoder (HRED), which is a hierarchical recurrent neural network that can consider the context of an article, is a beam search in sentence generation. Search) can be applied to improve the model. Here, HRED is an extension of the cyclic neural network, and may be composed of three cyclic neural network modules Encoder RNN, Decoder RNN, and Context RNN. Unlike the basic circulatory neural network that generates sentences only with the information of the encoder and decoder, contextual information is added to improve the completeness in sentence generation.The words are encoded as vectors in the encoder RNN, and the values are input as input values of the context RNN. , When the vector value is returned, it enters the Decoder RNN and finally outputs the sentence. Beam search is used as a search method in the learning process, and the circulatory neural network uses the best first search. Since the highest priority search is to use only the best probability value, the beam search is based on the above search method, but the search is conducted by specifying the range. In the beam search method, first, when the beam range is N in the learning process, the top N probability values are set in the current state, and after setting, they are input values of the next state, and N result values are created. Therefore, the total number of cases becomes the power of N, and among them, only the upper N values are left and deleted. Through beam search, it is possible to increase the accuracy of the entire sentence and improve the performance of the generated sentence with a plurality of results obtained as one input value. Of course, it will be apparent that a method for generating an accurate answer is not limited to the above-described method, and that various methods may be used according to embodiments.

Hereinafter, the operation process according to the configuration of the chatbot service providing server of FIG. 2 will be described in detail with reference to FIG. 3 as an example. However, it will be apparent that the embodiment is only any one of various embodiments of the present invention, and is not limited thereto.

Referring to FIG. 3A, (a) when the chatbot service providing server 300 starts to create a chatbot in the user terminal 100, it provides at least one unit block and selects the unit block selected by the user terminal 100 to the user terminal ( 100) is mapped and saved. And, if the user is a seller and wants to register an appropriate Q&A or F&Q at the same time on the shopping mall page where the product he sells is posted, or if he/she wants to receive an appropriate Q&A and put it into the chatbot, (b) the user terminal 100 , The chatbot service providing server 300 provides information on a shopping mall page on which the product it sells is posted. At this time, the chatbot service providing server 300 accesses the shopping mall page to extract product information from product data, and if it is text, it goes through pre-processing and analysis as it is, and if it is not text, it checks whether the text exists in the product image, and If there is no text to be read in the image, a similar image search is used, and if there is no text, the most similar Q&A data set is extracted using an engine that analyzes a product name by category, and recommends it to the user terminal 100. At this time, in the case of selecting any one Q&A data set from the user terminal 100, (c) the chatbot service providing server 300 includes a unit block capable of answering based on the Q&A data set, Another unit block selected by the terminal 100 is combined to create a chatbot. And, (d) when editing a Q&A in the user terminal 100 or a query that is not in the scenario in the customer terminal (not shown), the chatbot service providing server 300 can achieve similar intentions as much as possible with the above-described incremental learning technique. It extracts and answers entities to fit, and re-learns using configurations that are not in the scenario so that the scenario can gradually evolve.

3B to 3E are diagrams illustrating a process of creating a chatbot after constructing a scenario using at least one block according to an embodiment of the present invention. Referring to FIG. 3B, a box present in My Menu is an icon corresponding to one unit block. At this time, when shopping for i-blocks (unit blocks) is selected, a platform for purchasing i-blocks, which is connected to the store and performs a function or skin role, is displayed as shown in FIG. 3C. At this time, when dragging the i-block existing in the My Menu frame of FIG. 3B to the Apply My Menu frame, the selected i-block, such as product inquiry and installation inquiry, is applied to the chatbot as shown in the chatbot screen on the right. The applied menu automatically creates a scenario as shown in FIG. 3D in the chatbot interface through scenario creation.

At this time, when creating a scenario, it is self-evident that the iBlock is not limited to perform any one function or skin, as described above, and since it can be continuously updated or uploaded, it is not limited to any one function. Does not. Scenarios are displayed in the scenario builder as a scenario map as shown in FIG. 3D. In the case of order payment, payment inquiry, delivery period, order cancellation, exchange & refund, the chatbot operates according to the scenario as shown in FIG. 3E. Is done.

For details that are not described about the method of providing a chatbot service that provides a block-based adaptive development environment in which the API functions and scenarios of FIGS. 2 and 3 are combined, refer to FIG. 1 above, and block-based adaptation in which API functions and scenarios are combined. A method of providing a chatbot service providing an appropriate development environment is the same as the description or can be easily inferred from the description, so a description thereof will be omitted.

FIG. 4 is a diagram illustrating a process of transmitting and receiving data between components included in a chatbot service providing system providing a block-based adaptive development environment in which the API functions and scenarios of FIG. 1 are combined according to an embodiment of the present invention. to be. Hereinafter, an example of a process in which data is transmitted/received between each component will be described through FIG. 4, but the present application is not limitedly interpreted as such an embodiment, and according to various embodiments described above, It is apparent to those skilled in the art that the process of transmitting and receiving data may be changed.

Referring to FIG. 4, the chatbot service providing server 300 selects any one or at least one combination of unit blocks from the user terminal 100 and drags and drops it onto the scenario builder (S4100, S4200), and the selected unit The block is placed on the scenario builder and NLP is combined to create a chatbot (S4300).

Here, the chatbot service providing server 300 transmits and drives the chatbot generated to the user terminal 100 (S4400, S4500), and the chatbot service providing server 300 is a platform designated by the user terminal 100 or at least It is shared by one shopping mall server 400 (S4600).

The order between the above-described steps S4100 to S4600 is only an example and is not limited thereto. That is, the order of the above-described steps S4100 to S4600 may be mutually changed, and some of the steps may be executed or deleted at the same time.

For details that are not described about the method of providing a chatbot service that provides a block-based adaptive development environment in which the API functions and scenarios of FIG. 4 are combined, the block-based adaptation in which API functions and scenarios are combined through FIGS. 1 to 3 above. A method of providing a chatbot service providing an appropriate development environment is the same as the description or can be easily inferred from the description, so a description thereof will be omitted.

5 is a flowchart illustrating a method for providing a chatbot service providing a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention. Referring to FIG. 5, the chatbot service providing server uploads at least one unit block in which at least one scenario and an application programming interface (API) for executing at least one scenario are combined to at least one kind of store platform. (S5100).

In addition, the chatbot service providing server provides at least one scenario and at least one unit block in which an API for executing at least one scenario is combined, or a combination of at least one type of the user terminal through at least one type of store platform. In the case of purchasing at, one or at least one of the purchased combinations is selected (S5200).

Finally, the chatbot service providing server, when at least one unit block selected from the user terminal is uploaded on the scenario builder by drag-and-drop method from the user terminal, at least one function of at least one unit block combined on the scenario builder. The implemented chatbot is set to run on the platform designated by the user terminal (S5300).

For details that are not described about the method of providing a chatbot service that provides a block-based adaptive development environment in which the API functions and scenarios of FIG. 5 are combined, the block-based adaptation in which the API functions and scenarios are combined through FIGS. 1 to 4 above. A method of providing a chatbot service providing an appropriate development environment is the same as the description or can be easily inferred from the description, so a description thereof will be omitted.

The chatbot service providing method for providing a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment described with reference to FIG. 5 is a computer-executable command such as an application or program module executed by a computer. It may be implemented in the form of a recording medium including a. Computer-readable media can be any available media that can be accessed by a computer, and includes both volatile and nonvolatile media, removable and non-removable media. Further, the computer-readable medium may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The above-described method for providing a chatbot service providing a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention is an application basically installed in a terminal (this is a platform or operating system basically installed in the terminal, etc.). It can be executed by an application (that is, a program) directly installed on the master terminal through an application providing server such as an application store server, an application, or a web server related to the service. It can also be implemented. In this sense, the method for providing a chatbot service that provides a block-based adaptive development environment in which API functions and scenarios are combined according to an embodiment of the present invention described above is basically an application installed on a terminal or directly installed by a user (i.e. Program) and recorded on a computer-readable recording medium such as a terminal.

The above description of the present invention is for illustrative purposes only, and those of ordinary skill in the art to which the present invention pertains will be able to understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative and non-limiting in all respects. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

Claims (7)

In the chatbot service providing method executed in the chatbot service providing server,
Uploading at least one unit block to which at least one scenario and an application programming interface (API) for executing the at least one scenario are combined to at least one type of store platform;
When a user terminal purchases any one or a combination of at least one of the at least one unit block through the at least one type of store platform, receiving the purchased one or at least one combination;
When at least one unit block selected from the user terminal is uploaded on the scenario builder by a drag-and-drop method from the user terminal, at least one combination of at least one function of at least one unit block combined on the scenario builder is Setting a chatbot implemented in one scenario to run on a platform designated by the user terminal; And
When the user terminal selects any one unit block or a plurality of unit blocks among the at least one unit block, generating a chatbot in which the selected unit block and the natural language processing module are combined and transmitting the chatbot to the user terminal; But,
The natural language processing module,
Constituent parts of the structure representing the information to be conveyed that determine what to say, the part that applies the rules for the dialogue structure and sentences to determine the order of the sentences, and the information about words and the syntactic rules to generate the actual sentences. Generate natural language by dividing it into applied parts,
The at least one unit block is a block corresponding to a function or a design of at least one block constituting the chatbot,
The block corresponding to the function is a block including map view, delivery inquiry, payment inquiry, company extension number confirmation, and Q&A Excel data push, and the block corresponding to the design is a block corresponding to the skin of the chatbot,
In the above scenario, the number of each case that a user can query when using a chatbot and an answer to it are produced in the form of an algorithm,
The chatbot service providing server,
When a unit block is selected in the user terminal, a sharing platform in which the selected unit block is shared for a fee or for free is provided.
A method of providing a chatbot service that provides a block-based adaptive development environment in which API functions and scenarios are combined.
delete The method of claim 1,
The function corresponding to the at least one unit block includes at least one function usable in the chatbot,
The at least one function includes at least one function required for e-commerce, a chatbot service providing method for providing a block-based adaptive development environment in which API functions and scenarios are combined.
delete The method of claim 1,
The chatbot service providing server,
After collecting product data of products registered on at least one shopping mall page from the user terminal, performing pre-processing, analyzing the product data, and recommending a previously stored Q&A data set to the user terminal, and the recommended Q&A data When a set is selected in the user terminal, a chatbot is created to correspond to the selected Q&A data set,
A method of providing a chatbot service that provides a block-based adaptive development environment in which API functions and scenarios are combined, which uses HRED (Hierarchical Recurrent Encoder-Decoder), a hierarchical recurrent neural network for the context of an article.
An upload unit for uploading at least one unit block to which at least one scenario and an application programming interface (API) for executing the at least one scenario are combined to at least one type of store platform;
When the user terminal purchases any one or at least one combination of at least one unit block in which the at least one scenario and the API for executing the at least one scenario are combined through the at least one kind of store platform, A selection unit for selecting one or at least one of the purchased combinations; And
When at least one unit block selected from the user terminal is uploaded on the scenario builder by a drag-and-drop method from the user terminal, the chatbot implementing at least one function of the at least one unit block combined on the scenario builder Including; a setting unit configured to run on the platform designated by the user terminal;
When the user terminal selects any one unit block or a plurality of unit blocks among the at least one unit block, a chatbot in which the selected unit block and the natural language processing module are combined is generated and transmitted to the user terminal,
The at least one unit block is a block corresponding to a function or a design of at least one block constituting the chatbot,
The block corresponding to the function is a block including map view, delivery inquiry, payment inquiry, company extension number confirmation, and Q&A Excel data push, and the block corresponding to the design is a block corresponding to the skin of the chatbot,
In the above scenario, the number of each case that a user can query when using a chatbot and an answer to it are produced in the form of an algorithm,
When a unit block is selected in the user terminal, a sharing platform in which the selected unit block is shared for a fee or for free is provided.
Chatbot service providing server that provides a block-based adaptive development environment in which API functions and scenarios are combined.
Provides at least one unit block in which at least one scenario and an API (Application Programming Interface) for executing the at least one scenario are combined, and when the unit block is selected, the selected unit block is uploaded on the scenario builder to start the chatbot. A generated chatbot service providing server; And
Select and purchase any one of the at least one unit block or a combination of at least one unit block from at least one type of store platform, upload it on the scenario builder, and generate a chatbot based on the at least one unit block. Including; driving or sharing user terminal;
When the user terminal selects any one unit block or a plurality of unit blocks among the at least one unit block, a chatbot in which the selected unit block and the natural language processing module are combined is generated and transmitted to the user terminal,
The at least one unit block is a block corresponding to a function or a design of at least one block constituting the chatbot,
The block corresponding to the function is a block including map view, delivery inquiry, payment inquiry, company extension number confirmation, and Q&A Excel data push, and the block corresponding to the design is a block corresponding to the skin of the chatbot,
In the above scenario, the number of each case that a user can query when using a chatbot and an answer to it are produced in the form of an algorithm,
When a unit block is selected in the user terminal, a sharing platform in which the selected unit block is shared for a fee or for free is provided.
Chatbot service providing system that provides a block-based adaptive development environment in which API functions and scenarios are combined.

Images