KR20200095795A - Program of processing a message that provides feedback expression items - Google Patents

Abstract

The present invention relates to a program for message processing providing a feedback expression item. The program includes the following functions of: providing a message interface to a user terminal and processing a message transceived through the message interface, while providing a text, an expression image item or a feedback item combined with a text and at least one expression image item to the user terminal through the interface; storing and managing the text, the expression image item and the feedback item, which are stored in a database, by matching the same in accordance with corresponding classification information; and extracting and analyzing message information transceived through the message interface and information of users using a message service providing device, deriving elements, which are going to be processed with learning, from the extracted and analyzed text information, expression image item information and user information, analyzing an initial feedback item model by repetitively learning the derived elements in accordance with attribute information by element, creating and recommending a feedback item, which is suitable for a dialogue context, to a user based on the analyzed initial feedback item model, and newly creating and recommending a feedback item, which reflects the accuracy of the feedback item recommended to the user, to the user by analyzing the accuracy. Therefore, the program is capable of guaranteeing the properness of feedback item recommendation considering a situation or context of a dialogue.

Description

Program for processing a message that provides feedback expression items

The present invention relates to a program for message processing. More specifically, the present invention relates to a program for message processing that provides a feedback item for expressing an opinion corresponding to a message of a counterpart.

In a message service such as a chat service, various types of expression methods such as emoticons and emojis are used to express the emotions of users.

Emoticons are symbols used to express the emotions of service users in the chat service, and are a combination of emotions and icons, and in the early days, emoticons were made with a simple combination of various symbols and letters on the keyboard. , In recent years, it has developed into a form representing the user's emotions through various types of character images.

In addition, emoji is a method of outputting emoji embedded in a font form of a platform system, unlike an emoticon method based on a character combination, and a method of outputting an emotional expression image corresponding to an input emoji code is illustrated.

As service users' demand for such emojis and emoticons increases, it has become common to designate various types of image items in a message service and provide them to service users.

In the existing chat service, the user entered the conversation content and opened the emoji and emoticon selection window and selected and entered the appropriate one. However, in order to make the process more convenient, the text entered by the user is automatically converted to emoji or emoticon and displayed. The technology was also developed.

However, since such a conventional technology simply converts and displays a specific keyword into a pre-matched emoticon when a specific keyword is input, it is completely irrelevant to the content, context, and emotional state of the conversation, and in an inappropriate manner in the light of the conversation situation. There are also many things displayed.

Accordingly, there is a need to develop technologies such as recommending more appropriate emojis and emoticons in consideration of the user's emotional state and conversation context, beyond simple keyword matching technology to convert emoji and emoticons.

The present invention was conceived to solve the above-described problems, and it is possible to recommend emojis and emoticons in messages exchanged with each other by entering a conversation between a user and a chat partner, but recommended emojis and emoticons An object of the present invention is to provide a program for message processing that recommends a feedback item in consideration of the emotional state of the message and the dialogue context of the message.

A program according to an embodiment of the present invention for solving the above-described problems provides a message interface to a user terminal, processes messages transmitted and received through the message interface, and provides text, expression image items through the interface. Or providing a feedback item in which text and one or more expression image items are combined to the user terminal; A function of matching, storing and managing text, expression image items, and feedback items stored in the database according to the classification information; It extracts and analyzes the message information transmitted and received through the message interface and the information of users using the message service providing device, and derives elements to be learned and processed from the extracted and analyzed text information, expression image item information, and user information. Iteratively learns elements according to attribute information for each element to analyze the initial model of the feedback item, creates a feedback item that fits the dialogue context based on the analyzed initial model of the feedback item, and recommends it to the user, and the feedback item recommended to the user It includes a function of analyzing the accuracy of and creating a new feedback item reflecting the accuracy and recommending it to the user.

According to an embodiment of the present invention, various types of feedback items are recommended in consideration of the conversation situation and context, and the appropriateness of the feedback item recommendation is improved by considering the context or context of the conversation, compared to a known algorithm that automatically recommends automatically through keyword matching. There is a collateral effect.

In addition, by considering the emotion and the dialogue context through the dialogue content, emotions that do not appear on the surface but are contained within the immediately preceding utterance can be considered, and emotions that cannot be expressed by the utterance itself can be supplemented.

1 is a diagram schematically showing an entire system according to an embodiment of the present invention.
2 is a block diagram showing a service providing apparatus according to an embodiment of the present invention in more detail.
3 is a more detailed block diagram of a feedback item recommendation unit according to an embodiment of the present invention.
4 is a flowchart illustrating a method of recommending a feedback item according to an embodiment of the present invention.
5 to 28 are diagrams for explaining a message interface provided according to an embodiment of the present invention.

The following merely illustrates the principles of the present invention. Therefore, although those skilled in the art can implement the principles of the present invention and invent various devices included in the concept and scope of the present invention, although not clearly described or illustrated herein. In addition, it is understood that all conditional terms and examples listed in this specification are, in principle, expressly intended only for the purpose of making the concept of the present invention understood, and are not limited to the embodiments and states specifically listed as such. Should be.

In addition, it is to be understood that all detailed descriptions that list the principles, aspects and embodiments of the invention as well as specific embodiments are intended to include structural and functional equivalents of these matters. It should also be understood that these equivalents include all currently invented equivalents as well as equivalents to be developed in the future, ie, all devices invented to perform the same function regardless of structure.

Thus, for example, the block diagrams herein are to be understood as representing a conceptual perspective of exemplary circuits embodying the principles of the invention. Similarly, all flow diagrams, state diagrams, pseudo-codes, etc. are understood to represent various processes performed by a computer or processor, whether or not the computer or processor is clearly depicted, which may be substantially represented on a computer-readable medium. Should be.

Also, the clear use of terms presented in terms of processors, controls, or similar concepts should not be interpreted exclusively by reference to hardware capable of executing software, and without limitation, digital signal processor (DSP) hardware, ROM for storing software. It should be understood to implicitly include (ROM), RAM and non-volatile memory. Other hardware for governors may also be included.

The above objects, features, and advantages will become more apparent through the following detailed description in connection with the accompanying drawings, and accordingly, those skilled in the art to which the present invention pertains can easily implement the technical spirit of the present invention. There will be. In addition, in the description of the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

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

1 is a diagram schematically showing an entire system according to an embodiment of the present invention.

Referring to FIG. 1, a system according to an embodiment of the present invention may include a plurality of user terminals 100 and a service providing apparatus 200.

More specifically, the user terminals 100 and the service providing apparatus 200 may be wired or wirelessly connected through a network, and each user terminal 100 and the service providing apparatus 200 may be connected to the Internet for communication between networks. Data can be transmitted and received through network, LAN, WAN, Public Switched Telephone Network (PSTN), Public Switched Data Network (PSDN), cable TV network, WIFI, mobile communication network, and other wireless communication networks. In addition, each user terminal 100 and the service providing apparatus 200 may include respective communication modules for communicating with a protocol corresponding to each communication network.

In addition, the user terminal 100 described in the present specification includes a mobile phone, a smart phone, a laptop computer, a terminal for digital broadcasting, a personal digital assistant (PDA), a portable multimedia player (PMP), and a navigation system. Although it may be included, the present invention is not limited thereto, and may be various devices capable of user input and information display.

In such a system, the user terminal 100 may be connected to the service providing device 200 to receive a message interface service with the user terminal.

For this processing, the service providing device 200 provides a message interface to the user terminal 100, transfers a message received through the message interface to the user terminal, or transmits a message received from the user terminal to the user terminal 100. ) Can be performed.

For example, the service providing device 200 may be a chat server device, the user terminal 100 may be a portable terminal with a chatting application installed, and the service providing device 200 may be used between each user terminal or between user terminals. A chat interface within a group created by this is provided, and messages transmitted and received through the chat interface may be relayed and stored.

The message processed by the service providing device 200 may include various message content, the content may include data in an email, short or long text message format, and the service providing device 200 may include the message content. The data in the e-mail, short or long text message format can be transmitted through a mobile communication network or a dedicated Internet network.

Meanwhile, the service providing device 200 includes a communication protocol conversion, a file server, a collect server, a push gateway server, a transmission server, an account server, an administration server, a reception server, a conversion server, a messaging server, It may be implemented as a server device that performs an overall role such as a push receiving server and a push sending server function.

In addition, communication networks to which the user terminal 100 and the service providing device 200 are connected may be different networks, and accordingly, the service providing device 200 may provide data conversion processing according to each network protocol. In addition, the service providing device 200 may encrypt and manage all or part of the message information as necessary.

In addition, the service providing apparatus 200 according to an embodiment of the present invention may provide various additional information and functions processed according to a message service to the user terminal 100, provide service subscription and affiliate processing, or provide a financial server It is possible to further perform billing processing and the like for providing an alliance service with a plurality of service providers by communicating with.

In addition, a separate application may be installed in the user terminal 100, and a general portable terminal including a separate control unit (not shown), a storage unit (not shown), an output unit (not shown), and a communication unit (not shown) It can basically include the composition of.

Through such a system, when the service providing device 200 provides a relay service of message content transmitted and received through a message interface, text, an expression image item or text and one or more expression image items are combined in the message interface. Feedback items may be included.

The intention expression image item may include an image representing the user's emotion, mood, state, and other intentions to be expressed, and the image includes, for example, an emoji emoji format matching specific expression information, a text emoticon format Alternatively, it may be composed of data in a separately designated picture file format or a file format that allows movement. In addition, the expression information may include keywords indicating emotions, moods, states, or other intentions, and data indicating additional descriptions, and one or more keywords and data indicating additional descriptions may be matched to one image.

The feedback item is a text or expression image item representing a user's emotion, mood, state, and other intentions to express, or in a format in which text and image are combined, within the device based on the context of the conversation between the user and the chat partner. It can be automatically formed through learning.

Accordingly, the service providing device 200 may fuse the expression information with the text and image, and store and manage the fused information as a feedback item. The service providing device 200 generates a list of feedback items and It may be provided through a message interface output from the terminal 100.

The user terminal 100 may input a message through the message interface, or select and input a desired text or expression image item or a feedback item in which text and expression image items are combined from the feedback item list. A message including a feedback item may be delivered to another user terminal or may be registered as a post in a group chat room.

Further, the service providing device 200 may generate and recommend a feedback item by performing learning corresponding to the feedback items input from the feedback item list input from the message interface of the user terminal 100, and the generated feedback By analyzing the accuracy of the item, a feedback item reflecting the accuracy can be generated, and this can be newly recommended to the user.

FIG. 2 is a block diagram showing a service providing apparatus according to an embodiment of the present invention in more detail, and FIG. 3 is a block diagram showing a feedback item recommendation unit according to an embodiment of the present invention.

2 and 3, the service providing apparatus 200 includes a control unit 210, a communication unit 220, a message interface unit 225, a feedback item recommendation unit 230, a database 240, and a visual item management unit ( 250) may be included.

The controller 210 controls the overall operation of each module to provide operation processing according to the service providing method according to an embodiment of the present invention, and may include one or more microprocessors for this.

The communication unit 220 may include one or more modules that enable wired or wireless communication between networks in which the user terminal 100 or the service providing device 200 is located. For example, the communication unit 220 may include at least one of a mobile communication module, a wired Internet module, a wireless Internet module, and a short-range communication module.

In addition, the message interface unit 225 may register user information in response to users using a service according to an exemplary embodiment of the present invention, and relay, store, and manage transmitted/received messages. The user information may include user identification information for transmitting and receiving messages, and the user information may be newly registered, updated, and managed according to a user information registration or update request received from the user terminal 100. In addition, all or part of the user information may be encrypted and stored.

On the other hand, the database 240 may store a program for the operation of each component, and may store input/output data (eg, text, expression image item, feedback item, etc.), and perform learning processing. It is also possible to store and manage databases classified separately for this purpose.

The feedback item management unit 250 may match and store and manage feedback items stored through the database 240 according to their classification information. The classification information may include attribute information for each element of an image item, and the attribute information for each element includes, for example, information on type of expression of intention, information on image type, information of image matching the expression of intention, and the number of times of use by multiple users. Information, image information matched with user information, and feedback processing information may be exemplified.

In addition, the feedback item management unit 250 receives a text message and an emoji, an emoticon message, or a message in which the text and an emoji and an emoticon are combined through the message interface unit 225, and receives the Applicable to classification information.

The feedback item recommendation unit 230 includes a preprocessor 10 and a feedback item generation learning engine 20.

The preprocessor 10 may include at least a message box extraction and analysis module 11, a user information extraction and analysis module 12, and an element derivation module 13.

The message box information extraction analysis module 11 extracts and analyzes information contained in a message box of a user who uses a chat service through the user terminal 100. At this time, the message box information may include text information and expression image item information input by the user, and the message box information extraction analysis module 11 may perform a conversation context analysis of the entire message box when analyzing the message box information. In addition, the dialogue context allows the user to directly designate a range and analyze it, or the service providing device 200 automatically recognizes and analyzes the range.

In addition, the message box information extraction and analysis module 11 can analyze the meaning of "SAD", that is, the emotion of "SAD", when extracting the text information "SAD" from the message box, and the message box When the intention expression image item information of the crying face emoticon" is extracted, the meaning of the "crying face emoticon", that is, the emotion of the "crying face emoticon" may be analyzed.

The user information extraction analysis module 12 performs a function of extracting and analyzing user information using the chat service through the user terminal 100, and the user information includes at least user profile picture information, user gender information, and user status notification. Information, etc. may be included.

In addition, the user information extraction analysis module 12 can analyze the extroverted characteristics of the user or the semantic information contained in the picture when extracting the user profile picture information, and extract the user status notification word information to determine the meaning of the user's emotional state. Can be analyzed.

For example, if the user's profile picture information is picture information taken with a large number of people at a festival site, the user information extraction and analysis module 12 analyzes it and can derive an analysis result indicating that the user has extroverted characteristics, When the user status notification word information is stored as "depressed", the user information extraction analysis module 12 may derive an analysis result that the user's emotional state is a sad state.

The element derivation module 13 is a learning engine 20 that generates a feedback item from text information, expression image item information, and user information extracted and analyzed by the message box information extraction analysis module 11 and the user information extraction analysis module 12. ) Can derive elements to be input.

For example, when text information "SAD" is extracted from the message box information extraction and analysis module 11, and the analysis of the emotion of "SAD" results in "sad emotion", the element derivation module 13 is " It is possible to derive a plurality of related expression image items corresponding to "sad emotion". In addition, when the message box information extraction analysis module 11 extracts the expression image item information "crying face emoticon" and the analysis of the emotion possessed by the "crying face emoticon" results in "sad emotion" The derivation module 13 may derive a plurality of text and related expression image items corresponding to "sad emotion" and having a high correlation with the "crying face emoticon".

In addition, as a result of extracting the user's profile picture information from the user information extraction and analysis module 12, the element derivation module 13 extracts the user's profile picture information, and when an analysis result indicating that he likes music and has extroverted characteristics is derived, the element derivation module 13 is This high text and/or pseudo-expression image item, that is, a plurality of text information and pseudo-expression image item information related to music and having extroverted elements can be derived.

At this time, the plurality of expression image items derived from the above-described element derivation module 13 are previously learned and stored in the database 240 with text information, expression image item information, and user information extracted and analyzed by the preprocessor 10. It can be derived by matching and matching with existing text and multiple expression image items, and this process can be learned through machine learning.

The feedback item generation learning engine 20 receives preprocessed information from the preprocessor 10 through the database 240, performs learning necessary for generating a feedback item, and recommends the learned information to a user.

At this time, the feedback item generation learning engine 20 may be machine learning processing, and one or more processing methods may be used as the machine learning algorithm in the machine learning processing, and further, input variables and target variables may be combined with a plurality of machine learning algorithms. The optimal machine learning algorithm is determined by applying to a combination of a plurality of machine learning algorithms and pre-learning, taking into account at least one of the accuracy of each algorithm, calculation time versus accuracy, performance ratio, and fitness of the model according to the learned result. I can.

The machine learning algorithm is, for example, Ridge Regression, Logistic Regression, Generalized Linear Model, Random Forest, Gradient Boosting Model, and Neural A network (Neural Network) or the like may be used.

Ridge regression is a basic linear model and provides additional parameters to handle extreme values or outliers. Logistic regression analysis is a basic linear model, and it is effective to use when the target variable, that is, the target to be predicted, shows a binomial distribution. The generalized linear model is a basic linear model, and it is effective to use when the target variable shows the Poisson distribution. As a random forest ensemble model, a model consisting of a number of decision trees, each decision tree is individually developed in consideration of the association between input variables and target variables, and has more flexible characteristics than the upper linear model. The change stimulus model is also called the Generalized Boosting Model or GBM, and is an ensemble model that has a similar tendency to the Random Forest, but there is a difference that each decision tree is improved by considering the accuracy of the decision tree that has already been made. It is also considered.

Neural networks are very flexible models that can predict virtually any target and can encompass both linear and non-linear patterns by adjusting parameters. Therefore, it is necessary to fine-tune parameter tuning. In particular, neural networks are statistical learning algorithms inspired by biological neural networks in machine learning and cognitive science, and deep learning methods can be implemented by complexly configuring such neural networks.

For example, the neural network may include an overall Fully Connected (FC), a Convolutional Neural Network (CNN) that is optimized with Max Pooling and convolution, and Recurrent Neural Networks (RNN) useful for time-series data.

The feedback item generation learning engine 20 that is learned through such a machine learning algorithm includes a derivation element learning module 21, a feedback item generation and recommendation module 22, and an accuracy analysis and reflection feedback item recommendation module 23. Can be.

The derived element learning module 21 is a module that learns information extracted, analyzed, and derived from the preprocessor 10 according to the classification information, and the classification information may include attribute information for each element of an image item, Element-specific attribute information includes, for example, text information, expression type information, image type information, image information matching the expression, number of times of use by multiple users, image information matching user information, and feedback processing information. It can be illustrated.

That is, the derivation element learning module 21 may analyze the initial feedback item model by repeatedly learning information extracted, analyzed, and derived from the preprocessor 10. At this time, the initial model of the feedback item may include at least the user's expected response, the user-specific expression image item preference, the relationship between the user and the chat partner, the current or this week's conversation trend, and the user-specific recommended response, etc., and according to newly learned information. The feedback item initial model may also be updated to reflect this.

For example, in the derivation factor learning module 21, after the user and the chat partner have a conversation about food, the user inputs "I am so sad" and "what?" from the chat partner. When the answer is received, the derivation element learning module 21 is an element that the user can write as an answer to the chat partner, that is, "There is nothing to eat in my fridge.", "Because I'm starving." Answers in the form of, etc. can be extracted.

In addition, the derivation factor learning module 21 is, for example, text information most input by a number of users of the chat service today, "omg", "Are you kidding me?" In this case, it can be learned and analyzed to extract today's conversation trend as an initial model of the feedback item, and the relationship between friends, family, boss, and lover can be analyzed by learning and analyzing the conversation contents of the user who uses the chat service and the chat partner. It can be identified, and an answer suitable for the relationship can be extracted as an initial model of the feedback item.

The feedback item generation and recommendation module 22 generates a feedback item suitable for the dialogue context based on the initial model of the feedback item learned and analyzed from the derivation element learning module 21 and recommends it to the user. The feedback item may be a text or a pseudo-expression image item, or may be configured in a form in which text and a pseudo-expression image item are combined.

The feedback item creation and recommendation module 22, for example, after the user and the chat partner have a conversation about food, the user inputs "I am so sad", and the chat partner "what?" When the answer is received, a specific feedback item that is an initial model of the feedback item learned and analyzed from the derivation element learning module 21, that is, "There is nothing to eat in my fridge.", "Because I'm starving." An answer may be recommended to a user by generating a feedback item in the form of a text including, for example, a feedback item in the form of an expression image item, or a feedback item in the form of a combination of a text and an expression image item.

In addition, the feedback item generation and recommendation module 22 is provided to the user when the conversation trend learned and analyzed from the derivation element learning module 21 is "hahahaha", "Are you kidding me?", and "omg". A feedback item in the form of a text or expression image item including ", "Are you kidding me?", and "omg", or a feedback item in the form of a combination of text and expression image items may be generated and recommended to users. .

Depending on the design, the feedback item generation and recommendation module 22 may recommend a feedback item set by a specific user, that is, a public figure or a celebrity, when recommending a feedback item to a user. At this time, a specific user may be designated and set by the administrator of the service providing device 200, and the recommended feedback item set by a specific user is composed of a feedback item preferred by a specific user that is automatically learned and analyzed from the derivation element learning module 21. Or, it may be composed of a feedback item directly written and created by a specific user.

Accuracy analysis and reflection feedback item recommendation module 23 is a module that analyzes the accuracy of the feedback item generated and recommended from the feedback item generation and recommendation module 22, and recommends a feedback item reflecting the feedback item to the user. It may represent a result of calculating the feedback item generated and recommended by the recommendation module 22.

In addition, the accuracy analysis and reflection feedback item recommendation module 23 uses the calculated result value of the feedback item to highlight meaning, emotion intensity, importance, mood, situation, environment, etc. It is possible to recommend a feedback item to a user by expressing differentially with various expressions and effect functions.

For example, the accuracy analysis module 23 may display the calculation result of the feedback item in percentage (%), and the percentage may vary depending on the accuracy, and the feedback item having the highest percentage is It may be highlighted or displayed relatively larger than other feedback items. In addition, the accuracy analysis module 23 may display a speech balloon in a dark color when the emotional state of the user or the chat partner is sad, and may display a feedback item explaining this.

The accuracy analysis and reflection feedback item recommendation module 23 may be automatically executed or may be designed to perform analysis application when a user selects a specific icon, for example, an icon of an analysis execution button.

4 is a flowchart illustrating a method of recommending a feedback item according to an exemplary embodiment of the present invention, and FIGS. 5 to 19 are diagrams illustrating a message interface provided according to an exemplary embodiment of the present invention.

Referring to FIG. 4, the service providing apparatus 200 first extracts text information and expression image item information included in message box information, as shown in FIG. 5, and analyzes them. In this case, the text information and the pseudo-expression image item information may be extracted as text by analyzing information of the text and the pseudo-expression image item, or a numerical value may be extracted. In addition, when analyzing the message box information, as shown in FIGS. 6 and 8, the user directly designates and analyzes the range, or the device automatically recognizes the range and analyzes the message box information. In addition, the message box information extraction and analysis module 11 can analyze the meaning of "SAD", that is, the emotion of "SAD", when extracting the text information "SAD" from the message box, and the message box When the intention expression image item information of the crying face emoticon" is extracted, the meaning of the "crying face emoticon", that is, the emotion of the "crying face emoticon" may be analyzed (S101).

Next, the service providing device 200 extracts user profile picture information, user gender information, user status notification word information, etc. included in the user information, and analyzes it. At this time, the service providing device 200 may analyze the extroverted characteristics of the user or the semantic information contained in the photo when extracting the user profile picture information, and analyze the meaning of the user's emotional state by extracting the user state notification word information. have. For example, if the user's profile picture information is picture information taken with a large number of people at a festival site, the service providing device 200 can analyze it and derive an analysis result indicating that the user has extroverted characteristics, and user status If the notification word information is stored as "depressed", an analysis result may be derived that the user's emotional state is sad (S102).

Next, the service providing apparatus 200 derives an element to be input to the feedback item generation learning engine from the message box information and the information extracted and analyzed from the user information. For example, the service providing apparatus 200 extracts text information "SAD" from the message box information as shown in FIG. 9, and when the result of the analysis of the emotion of "SAD" is "sad emotion", the "sad emotion" It is possible to derive a plurality of related text and expression image items corresponding to ". In addition, when the result of the analysis of the emotion of the "crying face emoticon", which is the expression image item information, is extracted from the message box information, the result is "sad emotion", it corresponds to "sad emotion". , It is possible to derive a plurality of text items and related expression image items that are highly related to the “crying face emoticon”. In addition, the service providing device 200 extracts the user's profile picture information from the user information, and when an analysis result indicating that the user likes music and has extroverted characteristics is derived, a text or expression having high similarity to the characteristics of the user's profile picture information Items, that is, a plurality of text information and expression item information related to music and having an extroverted element may be derived (S103).

Next, the service providing apparatus 200 analyzes the initial feedback item model by repeatedly learning the element-derived information through machine learning. At this time, the initial model of the feedback item may include at least the user's expected response, the user-specific expression image item preference, the relationship between the user and the chat partner, the current or this week's conversation trend, and the user-specific recommended response, etc., and according to newly learned information. The feedback item initial model may also be updated to reflect this. For example, after a user and a chat partner have a conversation about food, the user types "I am so sad" and the chat partner says "what?" When receiving an answer of, the service providing device 200 provides elements that the user can write as an answer to the chat partner, that is, "There is nothing to eat in my fridge.", "Because I'm starving, which are initial models of feedback items." ." Answers in the form of, etc. can be extracted. In addition, the service providing device 200 includes, for example, "omg" and "Are you kidding me?" as text information most frequently input by a plurality of users using a chat service today. In this case, it can be learned and analyzed to extract today's conversation trend as an initial model of the feedback item, and the relationship between friends, family, boss, and lover can be analyzed by learning and analyzing the conversation contents of the user who uses the chat service and the chat partner. After grasping, the answer suitable for the relationship may be extracted and analyzed as a feedback item initial model (S104).

After the initial model of the feedback item is extracted and analyzed, the service providing device 200 performs a function of learning necessary for generating a feedback item and recommending a feedback item to a user based on the learned information. In this case, the learning information may be machine learning processed, and one or more processing methods may be used as a machine learning algorithm in the machine learning processing. Furthermore, the input variable and the target variable are applied to a combination of a plurality of machine learning algorithms and a plurality of machine learning algorithms to pre-learn, and the accuracy of each algorithm according to the learned result, calculation time to accuracy, performance ratio, and model fitness. An optimal machine learning algorithm may be determined in consideration of at least one of them. That is, the service providing apparatus 200 may learn the extracted, analyzed, and derived information according to the classification information, analyze an initial model of a feedback item, and recommend a feedback item that is an answer suitable for a conversation context to a user. For example, the service providing device 200 may recommend a feedback item suitable for a conversation context to a user as shown in FIGS. 10 to 15. To take a more detailed example, a user and a chat partner have a conversation about food. Later, the user enters "I am so sad" and the chat partner says "what?" When receiving an answer of, the service providing device 200 provides elements that the user can write as an answer to the chat partner, that is, "There is nothing to eat in my fridge.", "Because I'm starving, which are initial models of feedback items." ." An answer may be recommended to a user by generating a feedback item in the form of a text including, for example, a feedback item in the form of an expression image item, or a feedback item in the form of a combination of a text and an expression image item. In addition, as shown in FIGS. 16 and 17, the service providing device 200 includes “omg” and “Are you kidding me?” as text information most frequently input by a plurality of users of the chat service today. If yes, it can be learned and analyzed to extract today's conversation trend as an initial model of a feedback item, and a feedback item in the form of a text or dictation image item including "Are you kidding me?" or "omg" to the user Alternatively, a feedback item in the form of combining text and expression image items may be generated and recommended to the user. In addition, as shown in FIGS. 18 and 19, the service providing device 200 learns and analyzes the conversation contents of the user who uses the chat service and the chat partner to grasp the relationship between friends, family, boss, and lover, and answers appropriate for the relationship. Is extracted as an initial model of the feedback item and can be recommended to the user.It analyzes and extracts the conversation type (persona) of the chat partner according to the text information and the expression image item information entered by the chat partner and displays it to the user. May be. Depending on the design, the service providing apparatus 200 may recommend a feedback item set by a specific user, that is, a public figure or a celebrity, when recommending a feedback item to a user as shown in FIGS. 20 and 21. At this time, a specific user may be designated and set by the administrator of the service providing device 200, and the set recommended feedback item of a specific user is composed of a feedback item preferred by a specific user automatically learned and analyzed from the device, or created by a specific user. And the generated feedback item (S105).

Next, the service providing apparatus 200 analyzes the accuracy of the generated and recommended feedback items, differentially expresses the feedback items based on the analysis results, and recommends them to the user. At this time, the feedback item may be displayed in a state including the calculated result value, and the calculated result value of the feedback item is highlighted, shaded, and gradated according to meaning, emotion, importance, mood, situation, environment, etc. , Animating, various expressions and effects functions, etc., can be differentially expressed to recommend feedback items to users. For example, the service providing device 200 may display the calculation result of the feedback item as a percentage (%), as shown in FIG. 22, and the percentage may vary depending on the accuracy, and the highest percentage The feedback item to have may be highlighted or displayed relatively larger than other feedback items. In addition, when it is determined that the emotional state of the user or the chat partner is a happy state based on the text information and the expression image item information input from the user or the chat partner as shown in FIGS. 23 to 27, the speech balloon Is displayed in a light color, and if it is determined that it is in a sad state, a speech bubble is displayed in a dark color, and a feedback item explaining this may be displayed at the bottom of the chat window, or the emotional state of the user or the chat partner is step-by-step as shown in FIG. Separately, the level of emotion may be displayed under the speech bubble (S106).

Meanwhile, the above-described method according to various embodiments of the present invention is implemented in the form of installation data to be executed on a terminal device and stored in various non-transitory computer readable mediums in each server or device. Can be provided. Accordingly, the user terminal 100 may access the server or device and download the installation data.

The non-transitory readable medium refers to a medium that stores data semi-permanently and can be read by a device, rather than a medium that stores data for a short moment, such as a register, cache, or memory. Specifically, the various applications or programs described above may be provided by being stored in a non-transitory readable medium such as a CD, DVD, hard disk, Blu-ray disk, USB, memory card, ROM, or the like.

In addition, although the preferred embodiments of the present invention have been illustrated and described above, the present invention is not limited to the specific embodiments described above, and the technical field to which the present invention pertains without departing from the gist of the present invention claimed in the claims. In addition, various modifications can be implemented by those skilled in the art, and these modifications should not be individually understood from the technical idea or prospect of the present invention.

Claims (6)

In a program for providing a message service,
Provides a message interface to a user terminal, and processes messages transmitted and received through the message interface, and a feedback item in which text, an expression image item, or text and one or more expression image items are combined, is sent to the user terminal through the interface. The functions it provides;
A function of matching, storing and managing text, expression image items, and feedback items stored in the database according to the classification information;
It extracts and analyzes message information transmitted and received through the message interface and information of users using the message service providing device, and derives elements to be learned and processed from the extracted and analyzed text information, expression image item information, and user information. Iteratively learns elements according to the attribute information of each element to analyze the initial model of the feedback item, creates a feedback item that fits the dialogue context based on the analyzed initial model of the feedback item, and recommends it to the user, and the feedback item recommended to the user It includes a function that analyzes the accuracy of the product, creates a new feedback item reflecting the accuracy, and recommends it to the user.
A program for message processing that provides feedback expression items.
The method of claim 1,
The feedback item recommendation function,
A function of extracting and analyzing message box information and user information, and deriving an element to be learned from among information extracted and analyzed from the message box information and user information; And
The element derived from the preprocessor is learned through a machine learning algorithm to generate a feedback item initial model, and based on the feedback item initial model, a feedback item suitable for a conversation context is generated and recommended to the user. It characterized in that it further comprises a function of performing the accuracy analysis and recommending a feedback item displaying the calculated result value to the user.
A program for message processing that provides feedback expression items.
The method of claim 1,
The feedback item,
A text or expression image item representing the user's emotions, moods, states, and other intentions to express, or a format in which text and images are combined, through learning within the device based on the dialogue context between the user and the chat partner. Characterized in that it is automatically formed
A program for message processing that provides feedback expression items.
The method of claim 1,
The initial model of the feedback item,
Characterized in that it includes at least one of a user expected answer, a user-specific expression image item preference, a relationship between a user and a chat partner, a conversation trend of today or this week, and a recommended answer for each user.
A program for message processing that provides feedback expression items.
The method of claim 1,
The attribute information for each element,
Characterized in that it comprises at least one of text information, expression type information, image type information, image information matching the expression, number of times of use of a plurality of users, image information matching user information, and feedback processing information.
A program for message processing that provides feedback expression items.
According to claim 2,
The calculated result value of the feedback item is
It is characterized in that it is differentially expressed by highlighting, shading, gradation display, animation and special effect functions according to at least one of emphasis on meaning, intensity of emotion, importance, mood, situation, and environment.
A program for message processing that provides feedback expression items.

Images