KR20190042337A - Interactive chatbot operation method and system based on natural language processing for activation of messenger group chat room - Google Patents

Abstract

Disclosed are a method and a system for operating an interactive chatbot based on natural language processing for activation of a messenger group chat room. The method is based on a messenger server operable in a smartphone or a desktop environment, and a bot server to provide an interactive chatbot instance based on natural language processing. The messenger server provides a group chat room including functions such as real-time chatting between a plurality of messenger clients and a schedule, a notification, and a vote in a chat room via an application. A messenger client can create a group chat room, and then generate an interactive chatbot instance to invite the interactive chatbot instance to the group chat room as a member. The invited interactive chatbot analyzes previous conversation content as a member of the group chat room to recommend a subject of common interest of members or analyzes inputted chatting of the members to manage a chat room service such as a schedule, a notification or a vote to contribute to activation of group chat rooms.

Description

{Interactive chatbot operation method and system based on natural language processing for activation of messenger group chat room}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a chat robot, that is, a chatbot technology, and more particularly, to a method for operating an interactive chatbot based on natural language processing technology in a messenger group chat room.

A variety of applications are available and used on smartphones and desktops. Among them, a messenger application providing a real-time chat service between users is very useful as a communication means. Recently, messenger applications are being used more than phones and characters.

Recent instant messaging applications offer chat rooms for one-to-one contact and social networking, as well as group chat rooms, which allow multiple application users to communicate simultaneously in a group. These group chat rooms are operated not only for friendship, but also for various purposes such as family gathering, study group, and business consultation. Especially, as the spread of smart phones has widespread recently, and the popularization of various smart devices such as tablet PCs has been widespread, the use of messenger applications has increased rapidly.

However, as the group chat rooms are widely used, the life cycle of group chat rooms tends to become shorter and shorter. Group chat rooms can be created easily and easily through messenger applications. This is why group chat rooms are created quickly and are quickly deactivated. Although this may not be the case for a group chat room purpose, in the case of a group chat room for socializing purposes, such a short life cycle may adversely affect friendship among members of the chat room.

A major reason for the short life cycle of a group chat room is that a group chat room is a real-time chat room on-line, so if members of a group chat room do not have a common topic of interest or a topic or conversation that will cause continuous conversation, And can be deactivated. In the group chat room, conversation is gradually reduced among the members, and conversation is interrupted at some moment, and this situation can be continued for a certain period of time. When this situation is reached, there may be contradictory situations in which the members of the same group are in the same chat room, but the relationship between the members is farther away.

Recently, attempts have been made to apply virtual robots that chat like humans, such as chatbots, to instant messenger applications. As an example, Patent Application No. 10-2015-0061679 (hereinafter referred to as " Prior Art 1 ") entitled " Method and apparatus for providing chat service using client bots " A method and apparatus for registering a client bot as a friend of a group member with respect to the group and helping contents related to the group member through the client bot quickly and conveniently. Chat is a one-to-one chat only. It does not provide any guidance on how to chat with a large number of members (participants) in a group chat room where a large number of users are participating, Since it operates on the basis of keywords and rules, the user always has to input a chat containing the command first, Because it works in "passive" do not have the ability to lead a conversation with leadership.

Another example related to the chatbot is a patent application No. 10-2014-0032082 (hereinafter referred to as "Prior Art 2") entitled "Mobile Business Intermediary System Using Program" The present invention proposes a system in which a bot server is provided between a mobile terminal providing a messenger application for a system for internal business and a corporate business system, and a business intermediary is provided to a terminal user through a chat application. Prior Art 2 As in the prior art 1, the bots and users are always allowed to chat only in a 1: 1 relationship, and the bots are fixed Since it operates on a keyword basis, the bots always have a limit to operate manually.

1. Patent Application No. 10-2015-0061679 (entitled " Method and Apparatus for Providing Chat Service Using Client Bots " 2. Patent Application No. 10-2014-0032082 (entitled " Mobile Business Intermediary System Using a Program "

Accordingly, in order to solve the above-mentioned problems and to secure prior patents, it is an object of the present invention to provide a group chat room in which at least three clients can simultaneously chat in real time in a messenger service operated on a smart phone or a desktop Inviting an interactive chatbot so that the interactive chatbot can be operated in the group chat room, and a system therefor.

It is another object of the present invention to provide an interactive chatbot operating method in a group chat room, which enables a chatbot to be operated in a bot server, and to allow an instant messenger client to create and destroy an instance through a messenger server, and a system therefor.

Another object of the present invention is to provide an interactive chatting system capable of promptly inputting a chat message in a group chat room according to a predetermined chatbot setting without a user's input of a sentence (message) A method for operating a chatbot and a system therefor are provided.

Yet another object of the present invention is to provide an interactive chatbot operating method capable of chatting with a messenger client based on natural language processing, analyzing input chat of a messenger client and analyzing the context of sentences from the chat client, and a system therefor .

The present invention is not limited to the above-described embodiments, and various modifications may be made without departing from the spirit and scope of the invention.

According to embodiments of the present invention for realizing the above objects, there is provided a method for operating an interactive chatbot based on a natural language processing so that an interactive chatbot instance is operated to contribute to activation of a group chat room. The method for operating the natural language processing interactive chatbot comprises the steps of: generating an interactive chatbot instance based on natural language processing through an instant messenger client of one of the members of the group chat room created through the messenger server through the bot server; Adding the interactive chatbot instance to the group chat room; and adding the interactive chatbot instance to the group chat room to allow the interactive chatbot instance to search the chat room created by the members through their respective messenger clients in the group chat room through the messenger server (I) suggesting a subject of common interest of the group chat room as a member of the group chat room according to an operating environment set by the interactive chatbot instance after learning, (ii) Schedule registration and notification of the above chat room, announcement notification, And providing at least one of the services for managing the voting and the voting management.

In the exemplary embodiments, the interactive chatbot operating method may further include setting an operation environment necessary for performing the function of the interactive chatbot in which the one of the messenger clients is generated.

In exemplary embodiments, the setting items of the operating environment of the interactive chatbot instance include (i) a name of the interactive chatbot instance used by the members to call the interactive chatbot instance in the group chat room , (ii) an initial common interest topic to be recommended to the group chat room when there is no initial chat content, (iii) a previous chat conversation content learning period of the group chat room, (iv) A common topic recommendation period for recommending a common interest topic, and (v) a notification period for notifying members of the schedule when the schedule, notice, and vote of the group chat room are registered, can do.

In the exemplary embodiments, the group chat room provides a chat function in which the messenger clients of three or more members can participate in real-time chatting, and can be operated on a smart phone or a desktop.

In exemplary embodiments, the common interest topic of the group chat room may be extracted based on the interactive chatbot instance being machine-learned of the previous conversation contents in the learning step.

In exemplary embodiments, a prediction model called Latent Dirichlet Allocation (LDA) may be used for extracting a common interest topic of the group chat room.

In exemplary embodiments, the interactive chattop operation method may be configured such that the interactive chattop instance is associated with a video, an article, and an article associated with the common interest topics via a web crawling module Searching for at least one of the web pages, and informing the group chatting room through a chat.

In an exemplary embodiment, the interactive chattop operation method may further include a step of, even if members of the group chat room do not input a chat message for a predetermined period or longer, And then actively initiating the conversation by chatting first.

In the exemplary embodiments, the learning may be performed by the interactive chatbot instance using a deep-learning-based natural language processing technique and a natural language processing knowledge database. The interactive chatbot instance may include a text sent by the members of the group chat room through the messenger clients Into words and morpheme units, and analyzing the context of the sentence and storing the analyzed context in the context database.

In the exemplary embodiments, the learning step may further include generating the sentence using the natural language processing technique based on the sentence stored in the context database of the interactive chatbot instance, and providing the sentence through chat .

In exemplary embodiments, the interactive chatbot instance may utilize a deep-running based Word embedding model or a Sequence-to-Sequence model to analyze the context of sentences entered in the group chat room.

According to embodiments of the present invention, an interactive chatbot-based natural language processing system is provided for allowing an interactive chatbot instance to be operated so as to contribute to activation of a group chat room. The operating system of the natural language processing based interactive chatbot includes a bot server, a messenger server, and a messenger client. The bot server generates an interactive chatbot instance based on the natural language processing in response to the request, and provides the chat chat room with the operation environment of the chatbot instance according to the provided setting value to the group chat room requested to generate the interactive chatbot instance. The messenger server generates a group chat room in response to a request from three or more users' messenger clients, and provides a real-time chat service to messenger clients and chat box instances participating as members in the same group chat room. Wherein the messenger client is provided in a user terminal and provides a real-time chat service by a user through the messenger server, requests the bot server to create the interactive chatbot instance and add the chat chat instance to the group chat room, Can be provided. When the chatbot instance is added to the group chat room, the chatbot instance is searched through the chat server through the messenger server to search for conversation contents created by the members in the group chat room. (I) a common interest topic of the group chatting room is proposed as a member of the group chatting room, (ii) a schedule registration and notification of the group chatting room, a notice notification And / or voting management, thereby contributing to the activation of the group chat room.

In exemplary embodiments, the bot server may include a common topic extraction module, a user sentence analysis module, and a natural language sentence generation module. The common subject extracting module can find common topics of interest of the members of the group chat room through machine learning on the conversation sentences entered in the group chat room and store them in a common interest subject database. The user sentence analysis module may divide the sentence into words and morpheme units inputted in the group chatting room using the natural language processing technique and the natural language processing knowledge database to analyze the context of the sentence and store the analyzed context in the context database . The natural language sentence generation module may generate a new sentence matching the context by utilizing the natural language processing technique based on the analyzed sentence stored in the context database.

In exemplary embodiments, the bot server may be configured to send a video, an internet article, and a web page related to the common interest theme through a crawling technique on the web, based on a common interest theme of the group chat room stored in the common interest subject database And a web crawling module for finding at least one of the pages and informing the group chatting room through a chat using the interactive chatbot instance.

In an exemplary embodiment, the bot server may include a chatbot information database for storing information of each generated chatbot instance, a knowledge base for the natural chat language processing of the interactive chatbot instance, And a context database for storing sentence semantics of the analyzed natural language.

In exemplary embodiments, when the common subject extraction module finds the common interest topic, a common interest topic of the conversation contents among the members of the group chat room is analyzed using a prediction model called Latent Dirichlet Allocation (LDA) Can be found.

In exemplary embodiments, the user sentence analysis module may use either Deep Learning based Word embedding model or Sequence-to-Sequence model when analyzing the context of a sentence.

In exemplary embodiments, the setting items of the operating environment of the interactive chatbot instance include (i) a name of the interactive chatbot instance used by the members to call the interactive chatbot instance in the group chat room , (ii) an initial common interest topic to be recommended to the group chat room when there is no initial chat content, (iii) a previous chat conversation content learning period of the group chat room, (iv) A common topic recommendation period for recommending a common interest topic, and (v) a notification period for notifying members of the schedule when the schedule, notice, and vote of the group chat room are registered, can do.

In the exemplary embodiments, the interactive chatbot instance may be configured such that, even if members of the group chat room do not input chat messages for a predetermined period or longer, members of the group chat room are actively chatted May be provided.

The interactive chatbot operation for activating the IM chat room according to the present invention having the above-described configuration has the following effects.

First, a topic of common interest of the members through analysis of conversation contents of the group chat room can be periodically provided through the interactive chatbot. Interactive chatbots based on natural language processing can be invited to IM chat room to actively participate in conversation and maintain conversation and connection. Thus, the group chat room can be continuously activated.

Secondly, since the interactive chatbot registers and manages the service based on the user sentence and periodically informs the service of the schedule, announcement, and voting in the group chat room, it provides the convenience to the members of the group chat room .

Third, since the interactive chatbot is based on natural language processing, it is possible to process a range of a more flexible range when analyzing an input sentence. This allows the user to use sentences that communicate with the actual person rather than a keyword-only sentence. Also, since the interactive chatbot can input the chat first to the user for the recommendation of the common topic, the registered schedule, the announcement, the vote, etc., the user can feel the interactive chatbot as a member of the group chat room.

Fourthly, the messenger server can provide various services through a method such as crowdsourcing using all the collected chat conversation contents and data on various interest topics extracted therefrom.

1 is a block diagram schematically illustrating the configuration of a system for implementing an interactive chatbot operation method based on natural language processing according to exemplary embodiments of the present invention.
2 is a flowchart illustrating an interactive chatbot creation and operation procedure according to a preferred embodiment of the present invention.
3 is a flowchart of a common topic recommendation service of an interactive chatbot according to a preferred embodiment of the present invention.
4 is a flowchart illustrating a chat room registration and notification service of an interactive chatbot according to a preferred embodiment of the present invention.
5 is a flowchart illustrating a chatting room announcement service of an interactive chatbot according to a preferred embodiment of the present invention.
FIG. 6 is a flowchart illustrating a chat room voting notification service of an interactive chatbot according to a preferred embodiment of the present invention.
7 is a diagram illustrating a chatting room service provided in a group chatting room according to a preferred embodiment of the present invention.
Figure 8 is a diagram illustrating preferences for an interactive chatbot instance, in accordance with a preferred embodiment of the present invention.
9 is a view of a chat room for an example of a common topic recommendation service of an interactive chatbot performed according to a preferred embodiment of the present invention.
FIG. 10 is a view showing a chat room for a case in which an interactive chatbot performs a schedule registration and notification service in a chat room according to a preferred embodiment of the present invention.
FIG. 11 is a diagram illustrating a chat room for an example of an interactive chatbot performed by a public registration and notification service in a chat room, according to a preferred embodiment of the present invention.
12 is a view showing a chat room for an example of an interactive chatbot performing a voting registration and notification service in a chat room according to a preferred embodiment of the present invention.

For the embodiments of the invention disclosed herein, specific structural and functional descriptions are set forth for the purpose of describing an embodiment of the invention only, and it is to be understood that the embodiments of the invention may be practiced in various forms, The present invention should not be construed as limited to the embodiments described in Figs.

The terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

For the embodiments of the invention disclosed herein, specific structural and functional descriptions are set forth for the purpose of describing an embodiment of the invention only, and it is to be understood that the embodiments of the invention may be practiced in various forms, And should not be construed as limited to the embodiments described in the foregoing description. That is, the present invention is capable of various modifications and various forms, and specific embodiments are illustrated in the drawings and described in detail in the following description. It is to be understood, however, that the invention is not intended to be limited to the particular forms disclosed, but on the contrary, is intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

Hereinafter, a method for operating an interactive chatbot based on natural language processing for activating a messenger chat room according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings. The same reference numerals are used for the same constituent elements in the drawings and redundant explanations for the same constituent elements are omitted.

FIG. 1 illustrates a schematic configuration of an overall system 50 for implementing a natural language processing based interactive chatbot operating method according to a preferred embodiment of the present invention.

According to an exemplary embodiment, the system 50 may include a bot server 100, a messenger server 200, and a group chat room 300. The bot server 100 is a server for providing an interactive chatbot service based on natural language processing. The messenger server 200 is a server system installed in a user terminal such as a smart phone or a desktop to provide a messenger service. The group chat room 300 refers to a chat room created from the instant messenger client 320. The communication between the user terminal on which the group chat room 300 is executed and the messenger server 200, the communication between the bot server 100 and the chat terminal instance executed in the user terminal, and the communication between the bot server 100 and the messenger server 200 Although the wireless communication method is shown in the drawing, the present invention is not limited to this, and communication between the bot server 100 and the messenger server 200 may be performed using a wired communication method.

According to an exemplary embodiment, the bot server 100 may include an interactive chatbot system 110 and a bot server database 120.

The interactive chatbot system 110 may include a common topic extraction module 111, a user sentence analysis module 112, a natural language sentence generation module 113, and a web crawling module 114. These modules 111, 112, 113 and 114 may be program modules implemented by a program.

According to an exemplary embodiment, the bot server database 120 may include a common interest subject database 121, a context database 122, a natural language processing knowledge database 123, and a chatbot information database 124. The common interest subject database 121 may store common interests of the messenger clients 320. The context database 122 may store sentence semantics of the analyzed natural language. The natural language processing knowledge database 123 is a knowledge base in which the interactive chatbot instance 310 stores and manages knowledge information for natural language processing. The chatbot information database 124 may store information of each generated chatbot instance. Although not shown, the bot server database 120 may further include a rule database in which rules regarding the operation of the interactive chatbot instance 310 are defined. The rule information to be stored in the rule database may be incorporated and managed as a part of the chatbot information database 124, for example.

The common topic extraction module 111 of the interactive chatbot system 110 may be a module for performing a machine learning method for finding a common topic of interest for the input conversation sentences. In order to find a common topic of interest, the module 111 inputs the previous conversation history of the group chat room to which the created chatbox instance 310 belongs, for example, a Latent Dirichlet Allocation (LDA) You can use a model to find common topics. Furthermore, the found common theme can be stored in the common interest subject database 121. [ If there is no previous conversation history in the group chat room in which they are associated, they may have a conversation collection period for a period of time to find common interest topics. The conversation collection can be based on meeting a certain amount of conversation over a period of time. For example, over a period of one week, more than 100 conversations must be entered to base a recommendation on a common interest topic. During the period of collecting the conversation, you will not recommend a common topic.

The user sentence analysis module 112 may be a module for analyzing the meaning of the sentence inputted into the chat room. For semantic analysis, the user sentence analysis module 120 may use a natural language processing technique such as a deep learning based word embedding or a sequence-to-sequence model, and a natural language processing knowledge database 123. In order to analyze the sentence, the user analyzes the sentence by word and morpheme unit based on the natural language processing technique, analyzes the result, and grasps the meaning of the user's sentence using the analyzed result and the knowledge base 123 for natural language processing. Also, the analyzed context is stored in the context database 122.

The natural language sentence generation module 113 may use the analyzed context data stored in the context database 122. [ The natural language sentence generation module 113 generates a new sentence corresponding to the sentence context by using a natural language processing technique like the user sentence analysis module 112 and generates a new sentence corresponding to the sentence context, To the group chatting room 300. [0050]

The web crawling module 114 finds at least one of a video, an internet article, and a web page related to a common interest topic through a crawling technique on the web based on a common interest topic of the group chat room stored in the common interest subject database 121 And outputting the found result to the group chat room 300 through chat using the interactive chatbot instance 310. [

The bot server 100 uses the modules 111 to 114 of the interactive chatbot system 110 and databases 121 to 124 to provide a common topic recommendation service, chat room schedule registration and notification service, Services, and chat room voting management services.

According to an exemplary embodiment, the messenger server 200 may include a chat room management module 210 that provides a service for chat room management, such as a service for creating and deleting chat rooms. In addition, the messenger server 200 may include a chat information database 220 and a user information database 230, which are necessary for chat room management.

The messenger server 200 can manage the plurality of messenger clients 320 through the chat room management module 210 by creating the group chat room 300 or deleting the generated chat room. In the user terminal in which the messenger is installed, the messenger clients 320 can chat with each other via the group chat room 300. [ The chat room management module 210 stores the information of the generated group chat room 300 and the information of the messenger clients 320 in the chat room information database 220 and the user information database 230 Can be stored. The contents of the chat between the chat room members (participants) in the group chat room 300 can be stored in the chat room information DB 220. [

The group chat room 300 may be created through the chat room management module 210 of the messenger server 200 at the request of the messenger client 320 of three or more members. In the group chat room 300, in addition to the messenger client 320 of three or more members, the interactive chat box instance 310 can participate as a conversation member. Each member of the group chat room 300 can create the interactive chatto instance 310 through the bot server 100 and invite the generated chatto instance 310 to the group chat room 300. [

According to an exemplary embodiment, the generated group chat room 300 can basically provide a real-time chat service among the members. In addition, a schedule registration and management service, a public service, and / or a voting service in the group chat room may be provided as shown in the screen illustrated in Fig.

According to the exemplary embodiment, in the case of such schedule, notification, and / or voting service, the members of the group chat room 300 may directly register through the chat room menu as shown in FIG. According to another exemplary embodiment, members of the group chat room 300 chat as illustrated in FIGS. 10, 11, and 12, and the chatbox instance 310 recognizes the chat contents and automatically registers a schedule It is possible.

FIG. 2 is a flowchart schematically illustrating a procedure for generating and operating an interactive chatbot according to an exemplary embodiment.

Referring to FIG. 2 together with FIG. 1, when the group chat room 300 has already been created, the instant messenger client 320 of the members of the group chat room 300 transmits, The interactive chatbot instance 310 can be created and invited to the group chat room 300. That is, the messenger client 320 can request the generation of the chatbox instance 310 to the bot server 100, and the bot server 100 generates the interactive chatbox instance 310 in response to the request, (Step 400). ≪ RTI ID = 0.0 >

When inviting the chatbot instance 310 to the group chat room 300, the basic environment setting related to the operation of the chatbot instance 310 can be performed (step 401). 8, the items that can be set as the basic environment of the chatbox instance 310 include (i) the chatbox instance 310 used for calling the interactive chatbox instance 310 in the group chat room 300, (Ii) an initial common interest topic of the group chat room 300 (which may be limited only when there is no initial chat room in the group chat room 300), (iii) learning the conversation contents of the group chat room 300 (Iv) a common topic recommendation period (e.g., 6 hours, 12 hours, 1 day, 3 days, 3 days) for informing or recommending a common interest topic every time there is no conversation in the group chat room 300 for a predetermined period, (V) a notice period for notifying members of the schedule, announcements, and polls when their schedules, announcements, and votes are registered in the group chat room 300, For, one month, one week, three days , Time such as one day before, 12 hours before and 3 hours before, 1 hour before) may comprise at least a portion of or the like.

The messenger client 320 of the member requesting the generation of the interactive chatbot instance 310 may add the chambot instance 310 with the basic configuration as a member of the group chat room 300 (step 402). The addition of the chatbot instance 310 searches for and adds the chatbot instance 310 provided by the bot server 100 by pressing a user addition button in the same manner as a method in which a member invites an external user to the group chat room 300 Lt; / RTI >

The chatbot instance 310 added to the group chat room 300 for the first time can receive all the chat conversations between the members in the group chat room 300 through the messenger server 200 (step 403) . That is, the conversation contents exchanged among the members in the group chat room 300 are stored in the chat room information DB 220. When the chatbox instance 310 newly joins the group chat room 300, the chatbox instance 310 can request the messenger server 200 to search for the existing chat content and read it.

The chatbot instance 310 receives the read dialogue content, and extracts a common interest topic of the member using the common theme extraction module 111 (step 404). The common topic extraction module 111 can mechanically learn the conversation sentences of the group chat room 300 provided through the interactive chat box instance 310 to find the common interest topic of the members of the group chat room 300. [ When extracting a common interest topic, you can create a dialogue content learning model and use it. For example, a Latent Dirichlet Allocation (LDA) model can be used as a dialogue content learning model. And store the found common interest subjects in the common interest subject database 121. [

After the extracted common interest topic is stored in the common interest subject database 121, the chatbot instance 310 continuously performs the chatbot service according to the environment setting set before being rejected by the member of the group chat room 300 (Steps 405 and 406).

Next, FIG. 3 is a flowchart showing an execution procedure of the common topic recommendation service of the interactive chatbot according to one embodiment.

Referring to FIG. 3 together with FIG. 1, the interactive chatbot instance 310 participating in the group chat room 300 can check whether there is an interactive input for a set period in the group chat room 300 (step 500).

If there is no dialogue during the set period, the chatbot instance 310 calls the common topic extraction module 111 (step 501), and the called common topic extraction module 111 searches for the current topic in the participating group chat room 300 Can be retrieved and read from the chatting room information database 220 (step 502).

The common subject extracting module 111 may extract common topics of interest of the members of the group chat room 300 using the read conversation contents data (step 503).

The common subject extraction module 111 may store the extracted common interest subjects in the common interest subject database 121 (step 506).

Thereafter, the web crawling module 113 may be used to crawl content on the web, for example, a video, an article, or a web page related to a common interest topic (step 507). The retrieved result can be output to the group chat room 300 through the interactive chatbot instance 310 at a set interval. That is, the web crawling module 113 delivers the crawling result to the chatbox instance 310, and the chatbox instance 310 can be uploaded to the corresponding chat room 300 (step 508).

Next, FIG. 9 shows a screen example of the group chat room 300 in which the common topic recommendation service is performed by the interactive chatbot according to the exemplary embodiment.

FIG. 9A illustrates an example of a screen where a messenger client 320 of the group chat room 300 has a chat room 300 that is created in the conventional manner. That is, the left screen shows the chat characters in the group chat room 300 where the members talk about the experience and the price of the game. At the top of the chat room screen, the date of chatting is written as May 14, 2017, indicating that a chat between the instant messenger clients 320 was made on that day.

9B shows an interactive chatbot instance 310 in the group chatting room 300 after the last chat in FIG. 9A, for example, three days later and no chat until May 17, A screen for presenting a common topic to a chat room is extracted. That is, the chatbot instance 310 extracts a common interest topic of the existing group chat room 300 using the common topic extraction module 111 according to the dialogue input confirmation cycle, extracts it using the web crawling module 113 It shows you how to chat with the contents on the subject. The screen illustrated in FIG. 9 (B) shows that the chatbot instance 310 recognizes that the special price information and the discount information for a specific game are a common theme based on chat characters between the members of the group chat room 300, Show the situation.

This allows other messenger clients 320 to continue conversations on the topic provided by the interactive chatbot instance 110. [ If there is no conversation among the members of the group chat room 300 during a period set in the group chat room 300 to which the chat chat group 300 belongs, the interactive chat box instance 110 provides contents about a topic And interact with the messenger clients 320 of the members. Accordingly, the interactive chatbot instance 110 is not limited to a passive conversation participant in the group chat room 300, but can actively induce and lead a conversation among the members while presenting a conversation topic among the members.

4 is a flowchart illustrating a chat room schedule registration and notification service of an interactive chatbot according to an exemplary embodiment.

4, when a member inputs chat contents through the respective messenger clients 320 in the group chat room 300, the chat box instance 310 belonging to the group chat room 300 receives chat messages (Step 600).

The chatbot instance 310 may analyze the detected chat message to determine which member has entered the name (e.g., 'chatbot') of the configured chatbot instance 310 (step 601).

If the name of the chatbot is input, the chatbot instance 310 calls the user sentence analysis module 112, and the chatbot instance 310 can transmit the chat message input by the member to the user sentence analysis module 112 (Step 602).

In response to this call, the user sentence analysis module 112 may analyze the input sentence of the member and extract the context of the input sentence (step 603).

 Based on the extracted context, the user sentence analysis module 112 can check whether the input sentence is a sentence related to a certain registration (step 604). The user sentence analysis module 112 includes the name of the chatbot in which the sentence is set, and can check whether the input sentence includes the vocabulary related to the time by confirming it in the natural language processing knowledge database 123. [ If so, the user sentence analysis module 112 can determine that the input sentence is a sentence related to a certain registration.

If the input sentence is determined to be a sentence related to the schedule registration, the user sentence analysis module 112 can check the context related to the space time in the analyzed sentence context. In addition, the context associated with the identified space-time can be stored in the context database 122 (step 605).

Then, the chatbox instantener 310 can register a schedule in the group chat room 300 based on the received space-time context. In addition, the member can be notified via chat in accordance with the basic notification cycle set in the chatbox instance 310 (step 606).

FIG. 10 shows an example of a screen in which the interactive chatbox system 110 provides a chat room schedule registration and notification service in the group chat room 300.

FIG. 10 is a view illustrating a chat window for the interactive chatbot system 110 according to an embodiment of the present invention. 10A illustrates a case where one of the members of the group chat room 300 to which the interactive chatbot instance 110 belongs requests the interactive chatbot instance 110 to register a schedule. For example, a chatbot name ("Chatbot") is called in response to a sentence input by the requesting messenger client 320, and then time information and space information related to the schedule ("This Sunday", " Nearby Bornea House Wedding Hall ") and request a schedule registration with the contents of the schedule (" Wedding Ceremony "). Accordingly, the interactive chatbot instance 310 belonging to the group chat room 300 analyzes the input sentence through the sentence analysis module 112, receives the analysis result, determines whether it is a sentence about the schedule registration can do.

To this end, the interactive chatbot instance 310 first confirms that the configured name of the interactive chatbot 110 has been called. If so, determine if it is about the schedule. Whether the content is related to the schedule can be judged using the natural language processing knowledge base 123 and the analyzed context. If it is determined that the content is related to the schedule, the schedule registration can be completed using the context for the time and space information related to the schedule registration.

FIG. 10B shows an example of performing a schedule notification in which the interactive chatbot instance 310 of the group chat room 300 is registered, for example, one day before the registered schedule. The interactive chatbot instance 310 is configured to send a schedule registered through chat in accordance with a schedule notification period set as one month before, one week before, three days before, one day before, 12 hours before, three hours before, one hour before, ) To the members of the organization.

Next, FIG. 5 is a flowchart of the chat room announcement notification service of the interactive chatbot system 110 according to the exemplary embodiment.

Referring to FIG. 5, the chat room announcement notification has steps similar to the chat room schedule registration and notification service. Steps 700 to 702 are substantially the same as steps 600 to 602. That is, a process of detecting the chat input of the member and confirming whether or not the name of the chatbot is inputted, and a process of calling the user sentence analysis module 112 are performed.

Then, the user sentence analysis module 112 analyzes the input sentence and extracts the context therefrom (step 703).

Then, based on the extracted context, it can be confirmed whether the input sentence is a sentence related to the known registration (step 704). That is, as in the case of the schedule registration, if the input sentence includes the name of the set chatbot, and it is determined that the input related vocabulary is included, it can be regarded as a sentence related to the known registration. The vocabulary associated with the announcement can be checked in the natural language processing knowledge database 123.

If it is confirmed that the sentence associated with the notification registration is inputted, the sentence can be registered as a notice (step 705).

Then, the chatbot instance 310 continuously informs the group chat room 300 of the registered notice according to the set period (step 706).

FIG. 11 illustrates a chat window related to the interactive chatbot system 110, in which the interactive chatbot system 110 performs a service for registering notices and providing notifications in the group chat room 300, as described above.

11A shows a case where one of the members of the group chat room 300 to which the interactive chatbot instance 310 belongs requests a public registration to the interactive chatbot instance 310. FIG. ("Chatbot") of the chatbot instance 310 is displayed on the input sentence of one of the members of the group chat room 300 and the contents for registering in the announcement Payment of tuition "), and can request a public registration. Accordingly, the interactive chatbot instance 310 belonging to the group chat room 300 analyzes the context of the input sentence through the user sentence analysis module 112, and judges whether the sentence is related to the known registration according to the analysis result .

FIG. 11B shows an example in which the interactive chatbot instance 310 of the group chat room 300 performs the notification announcement in which the interactive chatbot instance 310 is registered, for example, one day before the registered notice. The interactive chatbot instance 310 continuously informs the members of the registered chat through the chat according to the notification period set as one month before, one week before, three days before, one day before, 12 hours before, three hours before, one hour before .

Next, FIG. 6 is a flowchart of the interactive chatbot providing chat room ballot notification service according to a preferred embodiment of the present invention.

Referring to FIG. 6, the voting service for the chat room members has steps similar to the chat room registration and notification service. Steps 800 to 802 are substantially the same as steps 600 to 602. That is, a chattering detection of members in the group chat room 300, a confirmation of whether or not a chatbot name is input in the chat content, and a process of calling the user sentence analysis module 112 are performed .

In response to this call, the user sentence analysis module 112 may analyze the input sentence and extract the context for it (step 803).

Based on the context extracted by the user sentence analysis module 112, the user sentence analysis module 112 can check whether the input sentence is a sentence related to the ballot (step 804). If the inputted sentence includes the name of the set chatbot, it can be verified in the natural language processing knowledge database 123 to confirm whether or not the vocabulary associated with the vote is included. If it is included, the input sentence can be judged to be related to the ballot.

If the input sentence is judged to be a sentence related to the ballot, the chatbot instance 310 can check whether there is a person who does not vote until that time (step 805). If there is a person who has not voted by then, it is confirmed who he is (step 806).

Then, the chatbot instance 310 can notify the member who has not voted using the natural language sentence generation module 113 (step 807). This notification can be made in accordance with a preset cycle or at a predetermined time after confirmation. This notice can increase turnout.

FIG. 12 is a view illustrating a chat window for an interactive chatbot system 110 according to an embodiment of the present invention. 12A shows an example in which one of the members of the group chat room 300 to which the interactive chatbot instance 310 belongs requests the interactive chatbot instance 310 to register a vote. In the input sentence of the messenger client 320, a chatbot name (" Chatbot ") is called, and then a content for registering in a vote ("can go to the baseball field this Saturday") is input, "O / X"), enter the number of days to vote ("4 days"), and request a vote registration. Accordingly, the interactive chatbot instance 310 can analyze the context through the sentence analysis module 112 similar to the known registration and notification service, and determine whether the sentence is related to the registration of the vote.

FIG. 12B shows an example of requesting to vote for the messenger clients 320 who have not voted after 4 days according to the registered voting progress time. The interactive chatbot instance 310 may generate a message in the group chat room 300 to request a vote for a member who has not yet voted. The member will then complete the ballot within the time limit.

The present invention can be applied to a messenger application and used to activate a messenger service.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims. It can be understood that it is possible.

100: Bot server 110: Interactive chatbot system
120: Bot server database 200: Messenger server
300: Group chat room instance 310: Chatbox instance
320: Messenger client

Claims (19)

Generating an interactive chatbot instance based on a natural language processing through an instant messenger client of one of members of a group chat room created through a messenger server;
Adding the generated interactive chatbot instance to the group chat room;
The interactive chatbot instance added to the group chat room is searched through the messenger server for members' conversation contents created in the group chat room through their respective messenger clients and learned using the common topic extraction module; And
(I) suggesting a common interest topic of the group chatting room as a member of the group chatting room according to an operating environment in which the interactive chatbot instance is set, (ii) scheduling and notifying of the group chatting room, And performing at least one of providing at least one of a service, a service, and a voting management,
Wherein the interactive chatbot instance is operated so as to contribute to the activation of the group chat room. The interactive chatbot operating method according to claim 1, further comprising setting an operation environment necessary for performing the function of the interactive chatbot in which the one of the messenger clients is generated. The method of claim 2, wherein the setting items of the operating environment of the interactive chatbot instance include: (i) a name of the interactive chatbot instance used by the members to call the interactive chatbot instance in the group chat room, (ii) an initial common interest topic to be recommended to the group chat room when there is no initial chat content; (iii) a previous chat conversation content learning period of the group chat room; (iv) A common topic recommendation period for recommending a topic, and (v) at least a part of a notification period for checking the arrival of the due date when the schedule, notice, and vote of the group chat room are registered and informing the members in advance How to operate an interactive chatbot as a feature. The interactive chatbot operating method according to claim 1, wherein the group chat room provides a chat function for real-time chatting by messenger clients of three or more members, and can be operated on a smart phone or a desktop . The interactive chatbot operating method according to claim 1, wherein a common interest topic of the group chat room is extracted based on the machine learning of the previous chat content in the learning step of the interactive chatbot instance. 6. The interactive chatbot operating method according to claim 5, wherein a prediction model called Latent Dirichlet Allocation (LDA) is used for extracting a common interest topic of the group chat room. The interactive chatbot instance of claim 6, wherein the interactive chatbot instance searches at least one of a video, an article, and a web page related to the common interest topics through a web crawling technique through a web crawling module, The method comprising the steps of: communicating with the chatbot; The interactive chatbot system according to claim 1, wherein, even if the members of the group chat room do not input a chat message for a predetermined period of time, the interactive chatbot instance actively starts chatting to the members of the group chat room first according to the set operation environment Further comprising the steps of: 2. The method according to claim 1, wherein the step of learning comprises: using the deep learning-based natural language processing technology and the natural language processing knowledge database, the interactive chatbot instance is used to translate a sentence input by the members of the group chat room through the messenger clients into words And analyzing the context of the sentence and storing the sentence in the context database. [12] The method of claim 9, wherein the learning further comprises generating a sentence using the natural language processing technique based on a sentence stored in the context database by the interactive chatbot instance, How to interact with chatbot. The interactive chatbot operation method according to claim 9, wherein the interactive chatbot instance uses a deep-running-based Word embedding model or a sequence-to-sequence model to analyze a context of a sentence input to the group chat room . A chat server for generating an interactive chatbot instance based on natural language processing in response to a request, setting an operation environment of the chatbot instance according to a provided setting value, and providing the interactive chatbot instance to a group chat room requested to generate the interactive chatbot instance;
A messenger server that creates a group chat room in response to a request of messenger clients of three or more users, provides a messenger client who participates in the same group chat room as a member, and a chat chat instance to a real-time chat service; And
A chat server installed in the user terminal and providing a real-time chat service by the user through the messenger server, requesting the bot server to create the interactive chatbot instance and adding the chat chat instance to the group chat room, A messenger client that can provide,
Wherein the chatbot instance is added to the group chat room, the chat room instance is searched through the messenger server to learn conversation contents created by the members in the group chat room and learned through the bot server, and after the learning, (I) proposing a common interest topic of the group chat room, and (ii) providing at least one of a schedule registration and notification of the group chat room, a notification notice, and a vote management service Wherein the chat room is capable of contributing to activation of the group chat room. The system according to claim 12, wherein the bot server includes: a common subject extracting module for finding common topics of interest of the members of the group chat room through machine learning on the chat sentences entered in the group chat room and storing them in a common interest subject database; A user sentence analysis module for analyzing the context of the sentence by dividing the sentence into words and morpheme units inputted in the group chatting room using the natural language processing technique and the natural language processing knowledge database and storing the analyzed context in the context database; And a natural language sentence generation module for generating a new sentence corresponding to the context using the natural language processing technique based on the analyzed sentence stored in the context database. [14] The system of claim 13, wherein the bot server further comprises: a crawling method for browsing images, internet articles, and web pages related to the common interest topic based on a common interest topic of the group chat room stored in the common interest subject database Further comprising a web crawling module for finding at least one of the chat rooms and informing the group chat room through a chat using the interactive chat box instance. 14. The system of claim 13, wherein the bot server comprises: a chatbot information database for storing information of each generated chatbot instance; a knowledge base for natural language processing; And a bot server database including a common interest subject database and a context database for storing sentence meaning of the analyzed natural language. 14. The system of claim 13, wherein the common subject extraction module analyzes common topics of conversation contents among the members of the group chat room using a prediction model called Latent Dirichlet Allocation (LDA) when finding the common interest topic Interactive chatbot operation system. 14. The interactive chatbot operating system of claim 13, wherein the user sentence analysis module uses one of a deep-based Word embedding model and a sequence-to-sequence model when analyzing the context of the sentence. The method of claim 12, wherein the setting items of the interactive environment of the interactive chatbot instance include (i) a name of the interactive chatbot instance used by the members to call the interactive chatbot instance in the group chat room, (ii) an initial common interest topic to be recommended to the group chat room when there is no initial chat content; (iii) a previous chat conversation content learning period of the group chat room; (iv) A common topic recommendation period for recommending a topic, and (v) at least a part of a notification period for checking the arrival of the due date when the schedule, notice, and vote of the group chat room are registered and informing the members in advance Interactive chatbot operation system. 13. The interactive chatbot service system according to claim 12, wherein, in the interactive chatbot instance, even if the members of the group chat room do not input chat messages for a predetermined period or longer, Wherein the natural language processing-based interactive chatbot has a function of providing a natural language processing-based interactive chatbot.

Images