KR20240040859A - Method for providing chatbot service for mentality supporting service - Google Patents

Abstract

The method of providing a psychological support chatbot service according to the present invention includes the steps of receiving voice data, which is the contents of a consultation between a client and a counselor, from at least a counselor terminal (hereinafter referred to as 'user terminal') among the client terminal and the counselor terminal; converting the voice data into text data; transmitting the text data to the user terminal; Upon receiving a phrase label manipulated by the user from the user terminal, labeling the phrase label to the corresponding conversation; Receiving a counseling sentence transmitted from the client terminal, parsing the counseling sentence, and dividing the counseling sentence into a plurality of words; and extracting at least one specific word representing the user's emotion from among the plurality of words.

Description

Method of providing psychological support chatbot service {METHOD FOR PROVIDING CHATBOT SERVICE FOR MENTALITY SUPPORTING SERVICE}

The present invention relates to a method of providing a service, and more specifically, to a method of providing a psychological support chatbot service.

In modern society, people are under a lot of stress. In addition, many people suffer from mental suffering or are at risk for their mental health. However, a vague fear of psychotherapy makes people reluctant to seek treatment. Additionally, the initial treatment period may be missed due to the burden of the time or cost of treatment. In addition, it is difficult to find psychological treatment methods to treat these people.

Accordingly, psychotherapy through chatbots is being used. Chatbot refers to communication software based on artificial intelligence (AI) that provides appropriate answers to questions or various related information through text conversations with people. Chatbots perform functions such as friends, personal assistants, counselors, and lovers. There are instances where it is used as a companion for elderly people living alone and prevents dementia through conversation.

If you are experiencing mental stress, it is advisable to receive counseling from a professional counselor, but it is not easy due to various factors such as time and cost. Additionally, social prejudice is a factor that cannot be ignored. Accordingly, ways to provide psychological counseling services for the user's psychological state, that is, the user's emotions, through conversation in the form of a chatbot or app are being studied.

Embodiments of the present invention seek to provide a method of providing a psychological support chatbot service for psychological counseling.

A method of providing a psychological support chatbot service according to an embodiment of the present invention is to obtain the contents of the consultation between the client and the counselor from at least the counselor terminal 70 (hereinafter referred to as 'user terminal') among the client terminal 60 and the counselor terminal 70. Receiving voice data; converting the voice data into text data; transmitting the text data to the user terminal; And when receiving a phrase label manipulated by a user from the user terminal, labeling the phrase label to the corresponding conversation; receiving a counseling sentence for artificial intelligence counseling from the client terminal 60, and Parsing the counseling sentence and dividing it into a plurality of words; extracting at least one specific word representing a user's emotion from among the plurality of words; calculating an emotional level for the at least one specific word and generating a counseling progress sentence according to the emotional level; and transmitting the consultation progress sentence to the user terminal.

Additionally, analyzing the text data with reference to the phrase label; And it may further include updating the summary using the analyzed counseling content analysis.

Additionally, processing the text data into big data to generate a big data processed product; and deriving counseling assistance content from the big data processed product.

In addition, the step of transmitting the summary and the counseling assistance content to the counselor terminal 70 may be further included.

In addition, the step of extracting the specific word includes extracting the at least one specific word that matches a preset emotional word among the plurality of words, storing the at least one specific word in a database for counseling, and performing the counseling. The step of generating a progress sentence calculates the emotion size according to the emotion importance set for the at least one specific word, and the step of generating the counseling progress sentence determines that if the emotion size is greater than or equal to a set standard size, the user emotion is negative. The counseling progress sentence is judged as an emotion, and the counseling progress sentence asking about the motive for causing the negative emotion is generated, and the counseling progress sentence asking about the motive for causing the user emotion is the cause and reason of the occurrence of the negative emotion, and the negative emotion. It may be a sentence that asks at least one of the episodes about.

The psychological counseling agent chatbot server according to an embodiment of the present invention is a server 10 that communicates data with at least the counselor terminal 70 (hereinafter referred to as 'user terminal') among the client terminal 60 and the counselor terminal 70, a server communication unit 115 that receives voice data, which is the content of the consultation between the client and the counselor, from the user terminal; A text converter 130 that converts the voice data into text data; A counseling app for installation is transmitted to the user terminal to be installed, the text data is displayed on the installed counseling app, and when a phrase label manipulated by the user is received from the user terminal, the phrase label is added to the conversation. Labeling app management unit 125; a data analysis unit 135 that analyzes the text data with reference to the phrase label and updates a summary using the analyzed counseling content analysis; And a big data processing unit 140 that processes the text data into big data to generate a big data product and derives counseling assistance content from the big data product, and the data analysis unit 135 is configured to analyze the client. Receive a counseling sentence transmitted from the terminal 60, parse the counseling sentence, divide it into a plurality of words, extract at least one specific word representing the user's emotion from the plurality of words, and extract the at least one specific word representing the user's emotion. It may further include a counseling unit 150 that calculates the emotion level for a specific word and generates a counseling progress sentence according to the emotion level.

The method of providing a psychological support chatbot service according to the present invention extracts at least one specific word representing the user's emotion from a counseling sentence transmitted from a client terminal, and generates a counseling progress sentence based on a counseling technique based on at least one specific word information. Therefore, by conducting psychological counseling with the user, appropriate psychological counseling services can be provided for each user.

1 is a block diagram of a counseling system according to an embodiment of the present invention.
FIG. 2 is a block diagram of the terminal of FIG. 1.
Figure 3 is a block diagram of the server of Figure 1.
Figures 4 and 5 show a chat window of a counseling app running on a user terminal.
FIG. 6 is a flowchart for explaining the operation of the counseling system of FIG. 1.
Figure 7 is a system diagram showing a chatbot consultation system according to another embodiment of the present invention.
Figure 8 is a graph showing Russell's emotional space applied to the psychological counseling agent chatbot device according to the present invention.
Figure 9 is a flowchart showing the operation method of the psychological counseling agent chatbot server of Figure 8.
FIG. 10 is a flowchart showing an example of the operation of the psychological counseling agent chatbot device shown in FIG. 9.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. When describing with reference to the drawings, identical or corresponding components will be assigned the same reference numerals and redundant description thereof will be omitted.

Terms such as first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, a first component may be named a second component, and similarly, the second component may also be named a first component without departing from the scope of the present invention. The term and/or includes any of a plurality of related stated items or a combination of a plurality of related stated items. The term 'or' can be interpreted as a logical exclusive sum in the context, but if there is no direct description such as 'otherwise, otherwise, a logical exclusive sum', it is interpreted as a logical sum with the same meaning as 'and/or', that is, a logical sum.

Additionally, singular expressions include plural expressions unless the context clearly dictates otherwise.

When a component is said to be “connected” or “connected” to another component, it is understood that it may be directly connected or connected to the other component, but that other components may exist in between. It should be. On the other hand, when it is mentioned that a component is “directly connected” or “directly connected” to another component, it should be understood that there are no other components in between. This is the same in the top, bottom, left, and right placement relationships. For example, when a component is said to be on top of another component, it includes not only the case where a component is directly on top of another part, but also the case where another component is interposed between them.

In addition, the fact that the first component and the second component on the network are connected or connected means that data can be exchanged between the first component and the second component wired or wirelessly.

The suffixes “module” and “part” for components used in the following description are simply given in consideration of the ease of writing this specification, and do not in themselves give any particularly important meaning or role. Accordingly, the terms “module” and “unit” may be used interchangeably.

When these components are implemented in actual applications, two or more components may be combined into one component, or one component may be subdivided into two or more components, as needed. The same reference numerals are assigned to identical or similar components throughout the drawings, and detailed descriptions of components having the same reference numerals may be omitted and replaced with descriptions of the above-described components.

Additionally, the present invention encompasses all possible combinations of the embodiments shown herein. The various embodiments of the present invention are different from one another but are not mutually exclusive. The specific shapes, structures, functions, and characteristics described herein in one embodiment may be implemented in other embodiments. For example, components mentioned in the first and second embodiments can perform all the functions of the first and second embodiments.

In cases where an embodiment can be implemented differently, a specific process sequence may be performed differently from the described sequence. For example, two processes described in succession may be performed substantially at the same time, or may be performed in an order opposite to that in which they are described.

1 is a block diagram of a counseling system according to an embodiment of the present invention. Figure 2 is a block diagram of the terminal of Figure 1, and Figure 3 is a block diagram of the server of Figure 1. Figures 4 and 5 show a chat window of a counseling app running on a user terminal.

Referring to Figure 1, the communication system of the present invention may include a server 10, a client terminal 60, and a counselor terminal 70.

The client terminal 60 is a terminal used by a user who wants to receive counseling, and the counselor terminal 70 is a terminal used by a person who specializes in helping solve problems or answering questions through counseling.

Among the server 10, the client terminal 60, and the counselor terminal 70, at least the server 10 and the counselor terminal 70 are capable of data communication, and the client terminal 60 and the counselor terminal 70 are at least wired. Audio communication using a telephone network may be possible.

The client terminal 60 and the counselor terminal 70 may be electronic devices such as a wired phone or a PC capable of making phone calls. The client terminal 60 and counselor terminal 70 are electronic devices capable of wireless communication, including a smart phone, portable terminal, mobile terminal, and personal digital assistant (PDA). ), PMP (Portable Multimedia Player) terminal, telematics terminal, navigation terminal, personal computer, laptop computer, Slate PC, Tablet PC, ultrabook ), wearable devices (e.g., including watch type terminal (Smartwatch), glass type terminal (Smart Glass), HMD (Head Mounted Display), etc.), Wibro terminal, IPTV (Internet Protocol Television) terminal , Smart TV, digital broadcasting terminal, AVN (Audio Video Navigation) terminal, A/V (Audio/Video) system, and flexible terminal (Flexible Terminal) may be any one or a combination. In the present invention, it is preferable that at least the counselor terminal 70 is a terminal capable of installing an application program.

When the client terminal 60 is a general wired telephone, the client terminal 60 can communicate with the counselor terminal 70 through a wired telephone network. The counselor terminal 70 can communicate with the server 10 through a data communication network (Internet, etc.). The counselor terminal 70 may transmit audio signals exchanged with and received from the client terminal 60 to the server 10. If the client terminal 60 is a terminal capable of data communication, data communication can be performed with the server 10. Hereinafter, the client terminal 60 will be described assuming that it is a terminal capable of data communication.

Referring to FIG. 2, the terminal may include a terminal control unit 210, a terminal communication unit 220, a terminal input unit 230, an operation unit 240, a terminal output unit 250, and a terminal storage unit 260. . This terminal may be the client terminal 60 and/or the counselor terminal 70, or at least the counselor terminal 70.

The terminal communication unit 220 can communicate with other components of the communication network using wired/wireless communication technology.

Wired communication technologies include Power Line Communication (PLC), USB communication, Ethernet, serial communication, and optical/coaxial cables.

Wireless communication technology may include short-range communication technology, wireless Internet technology, GPS technology, mobile communication technology, and low-power wide area network technology.

Short-range communication technologies include Bluetooth, RFID (Radio Frequency Identification), IrDA (infrared Data Association), UWB (Ultra Wideband), ZigBee, Near Field Communication (NFC), and ultrasonic communication. (Ultra Sound Communication: USC), Visible Light Communication (VLC), etc. can be used.

Wireless Internet technologies include WLAN (Wireless LAN) (Wi-Fi), DLNA (Digital Living Network Alliance), Wibro (Wireless broadband), Wimax (World Interoperability for Microwave Access), and HSDPA (High Speed Downlink Packet Access). You can. Wireless Internet technology may be suitable for short-distance communication.

Mobile communication technologies include GSM (Global System for Mobile communication), CDMA (Code Division Multi Access), CDMA2000 (Code Division Multi Access 2000), EV-DO (Enhanced Voice-Data Optimized or Enhanced Voice-Data Only), and WCDMA (Wideband). CDMA), HSDPA (High Speed Downlink Packet Access), HSUPA (High Speed Uplink Packet Access), LTE (Long Term Evolution), LTEA (Long Term Evolution-Advanced), next-generation mobile communications (5G, etc.) can be used. .

Low Power Wide Area Network (LPWAN) can refer to a low-power wireless wide area network that has a very wide service range (more than 10 km) and provides communication speeds of hundreds of kilobits per second (kbps). Low-power wide area network technology can be used as a dedicated communication technology to provide the Internet of Things (IoT). To realize LPWAN, LoRaWAN (Long Range wireless network), SIGFOX, LTE-MTC (LTE Machine-Type Communications), narrowband Internet of Things (NB-IoT), etc. There may be technology.

The terminal input unit 230 may be equipped with a microphone for inputting audio signals. The terminal input unit 230 may be equipped with a camera for inputting video signals.

A microphone can receive external acoustic signals and process them into electrical voice data. The processed voice data can be transmitted to the outside (server 10, the other party's terminal, etc.) through the terminal communication unit 220.

The camera can process image frames such as still images or videos obtained by an image sensor in video call mode or shooting mode. The processed image frame may be displayed on the display module 255, which will be described later. Image frames processed by the camera may be stored in the terminal storage unit 260 or transmitted externally through the terminal communication unit 220.

The terminal operation unit (user input unit) 240 generates key input data that the user inputs to control the operation of the terminal. The terminal operation unit 240 is composed of a keyboard, mouse, key pad, dome switch, and touch pad (static/electrostatic) that can receive commands or information through a user's push or touch operation. It can be. Additionally, the terminal operation unit 240 may be configured with a jog wheel or a jog method for rotating keys, a method of operating such as a joystick, or a finger mouse. In particular, when the touch pad forms a mutual layer structure with the display module 255, which will be described later, it can be called a touch screen.

The terminal output unit 250 may be equipped with a speaker to output an audio signal. The terminal output unit 250 may be equipped with a display module 255 to output a video signal.

The display module 255 can display and output information processed in the terminal. As described above, when the display module 255 and the touch pad form a layered structure to form a touch screen, the display module 255 may be used as an input device capable of inputting information by the user's touch in addition to an output device. You can. Hereinafter, the display module 255 will be described assuming that it is a touch screen capable of user input.

The terminal storage unit 260 can store a number of application programs running in the terminal, data for operating the terminal, and commands. At least some of these applications may be downloaded from an external server via wireless communication. The application program is stored in the terminal storage unit 260, is installed on the terminal, and can perform its operations or functions under the control of the terminal control unit 210.

The terminal storage unit 260 includes flash memory type, hard disk type, multimedia card micro type, and card type memory (for example, SD or XD memory, etc.) ), magnetic memory, magnetic disk, optical disk, random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and PROM. (Programmable Read-Only Memory) It can be implemented as any one or a combination of some of them. The terminal may operate web storage that performs the storage function of the terminal storage unit 260 on the Internet, or may operate in relation to the web storage.

The terminal control unit 210 can control and process the overall operation of the terminal by controlling the operation of each unit. The terminal control unit 210 can be interpreted as a microcontroller or central processing unit.

Referring to FIG. 3, the server 10 includes a server control unit 110, a server communication unit 115, a server storage unit 120, an app management unit 125, a text conversion unit 130, a data analysis unit 135, It may include a big data processing unit 140 and a counseling unit 150.

The server 10 may communicate with the client terminal 60, the counselor terminal 70, and other components of the communication network through the server communication unit 115. The server communication unit 115 can use various wired/wireless communication technologies. For wired/wireless communication technology, refer to the description of the terminal communication unit 220.

The server communication unit 115 can receive voice data, which is the content of the conversation between the client terminal 60 and the counselor terminal 70, from at least the counselor terminal 70. The server communication unit 115 may receive voice data from the client terminal 60 and the counselor terminal 70, respectively, or may relay voice data between the client terminal 60 and the counselor terminal 70.

The server storage unit 120 can store various data and application programs. The server storage unit 120 may be implemented with various storage media.

The app management unit 125 may cause the counseling app related to the present invention to be downloaded to at least the counselor terminal 70 (hereinafter referred to as 'user terminal') among the client terminal 60 and the counselor terminal 70. The app management unit 125 exchanges data with the counseling app running on the user terminal and can use or manipulate the data.

It is desirable that the user terminal on which the counseling app has been downloaded and installed is running at the time of counseling.

The text conversion unit 130 can convert voice data containing consultation content into text data. The text conversion unit 130 can convert the consultation content into text in real time or convert the recorded consultation content into text after the consultation is completed. Since character conversion in real time is desirable in relation to labeling, which will be described later, the description below assumes character conversion in real time.

The app management unit 125 can transmit the converted text data to the user terminal. The user terminal that has received the text data can display it as text in the chat window (A) of the counseling app running on the display module 255. However, it is not limited to this and can also be displayed on a text app on a mobile phone or smartphone.

Referring to Figures 4 and 5, the audio consultation content between the counselor and the client may be converted into text and displayed in a chat format in the chat window (A).

The chat window (A) contains basic metadata such as the creation date of the voice data containing the consultation content recorded or when the counseling app is run, and the time when each conversation voice was created, and additional information such as the consultation date and the time when each conversation was uttered. Information can be displayed.

If this drawing shows the chat window (A) of the counseling app running on the counselor terminal 70, the left side may be displayed as a conversation box uttered by the client, and the right side may be displayed as a conversation box uttered by the counselor.

The counseling app running on the user terminal can provide a UI (User Interface) that allows the user to label the chat window (A).

As an example of labeling, a counseling app may provide a flagging feature. When the user performs at least one of the following operations, such as touching, pressing and holding, or sliding to the left or right, any conversation in the chat window (A), the consultation app can label the conversation with a flag. there is. In the counseling app, labeling can be canceled when an operation such as a double touch is performed.

As another example of labeling, a counseling app may provide a tagging feature. The user can label (tag) a tag in any dialog box of the chat window (A). Tags may be content created or already created by the user. The previously written content can be further divided into flags, such as dichotomizing it into 'high importance' and 'low importance'. ‘High importance’ tagging can be done by touching a conversation or sliding it to the left, and ‘low importance’ tagging can be done by sliding it to the right.

Users can manipulate dialog boxes containing key information they consider important or essential to be labeled with flags or corresponding tags. Key information may include client information, counseling field, counseling content, suggestion content, client satisfaction, etc. It is advisable to categorize key information and manualize it in advance so that you are aware of counseling fraud. This labeling allows not only the counselor but also the client to manipulate labeling in dialog boxes that they consider to be important information.

The counseling app can display the corresponding dialog box differently when labeling manipulation is performed. For example in Figure 5, a dialog box tagged or flagged as important information may be displayed with a thicker border, and a dialog box tagged as having low importance may have its border displayed as a dotted line. As another example, the background color or text properties (font, color, size, etc.) in a dialog box can be changed.

When labeling manipulation is performed in the counseling app, the identifier of the conversation and the labeled phrase label (flag, tag, etc.) can be transmitted to the server 10.

When the app management unit 125 receives a phrase label, it can refer to the conversation identifier and associate (label) the phrase label with the corresponding conversation.

The data analysis unit 135 can analyze text data. The data analysis unit 135 may refer to phrase labels when analyzing text data. The data analysis unit 135 may determine which conversations in the text data are important and/or unimportant through the phrase labels. The data analysis unit 135 may omit unimportant conversations from the analysis and increase the weight of important conversations.

The data analysis unit 135 can determine the structure of text data and extract the smallest meaningful unit of morpheme through sentence segmentation, analysis, extraction, and original restoration.

The data analysis unit 135 may generate a consultation content analysis product by analyzing text data referenced by the phrase label. Counseling content analysis may be extracted morphemes classified into predefined categories. Classification categories may include client information, counselor information, counseling field, counseling content, major keywords, problem solution plan, and client satisfaction. Some of the information that makes up a category may be associated with a phrase label.

Various techniques may be applied to classification by the data analysis unit 135. For example, analysis methods such as the meaning of words, priority among words, correlation between words, method of interpreting the meaning of sentences, and method of understanding the client's condition are stored in advance, and the data analysis unit (135) ) can use stored analysis methods. Phrase labels can be inserted as words or used to restore their original form. A phrase label may be associated with the corresponding word.

The data analysis unit 135 may generate a summary or update the summary using the analyzed counseling content analysis. The data analysis unit 135 may analyze in real time and continuously update the summary after initially generating it, or may generate the summary after the consultation ends.

The data analysis unit 135 may generate a summary centered on sentences related to key keywords or important labels.

The summary may include additional counseling assistance content, which will be described later. The summary may contain detailed information for the client to carry out the problem solution plan suggested by the counselor to the client. Detailed information may include hospital information, disease-related information, travel information, restaurant information, entertainment information, book-related information, expert information, school information, etc. Details may include what other clients have done.

The big data processing unit 140 may process text data as big data to generate a big data processed product. The big data processing unit 140 may process big data by referring to the analysis of the consultation content and/or the summary that is being created or being updated. Text data used in the big data processing unit 140 may be data for which morphological analysis has been completed.

The big data processing unit 140 collects and stores unstructured data other than text data with the client, counseling data from other clients, etc., preprocesses it, and predicts and classifies it through various analysis techniques (statistical analysis, deep learning, machine learning, etc.) Consultation content analysis, such as , can be generated.

Counseling content analysis may include types of counseling related to similar counseling content as the client, evaluation of the counselor, solutions to major problems, and client satisfaction.

The big data processing unit 140 can derive the auxiliary content of the consultation from the big data processed product and transmit the auxiliary counseling content to the counselor terminal 70.

The counseling department 150 may be responsible for a type of artificial intelligence counseling. The counseling department 150 can provide counseling with the client if the client wishes to have counseling without a counselor. The counseling department 150 can provide chatbot or voice recognition counseling. The counseling unit 150 can recognize the client's voice as text using the function of the text conversion unit 130. The counseling unit 150 can communicate with the client through the data analysis unit 135 and/or the big data processing unit 140 and ensure that appropriate counseling is provided. However, it is not limited to this, and the counseling unit 150 may itself perform machine learning or deep learning to perform counseling.

The more existing counseling data the counseling department 150 has, the more appropriately and accurately it can help solve the client's problems. The counseling unit 150 can provide psychological support tailored to each individual client through various pre-stored counseling data.

The consultation app provided by the app management unit 125 may provide a user interface (UI) necessary for artificial intelligence consultation.

FIG. 6 is a flowchart for explaining the operation of the counseling system of FIG. 1. Refer to Figures 1 to 5. Figure 6 shows a method of providing personalized psychological support services through big data analysis according to an embodiment of the present invention, viewed from the side of the server 10.

Referring to FIG. 6, the server 10 may receive voice data, which is the content of the consultation between the client and the counselor, from the user terminal (S310). The server 10 can convert voice data into text data (S315). The server 10 may transmit text data to the user terminal (S320).

When the server 10 receives a phrase label manipulated by the user from the user terminal (S325), it can label the conversation with the phrase label (S330).

The server 10 can analyze text data by referring to the phrase label (S335). The server 10 may generate a summary or update a previously generated summary using the analyzed counseling content (S340). Some of the content in the summary may be associated with text labels.

The server 10 may process text data as big data to generate a big data processed product (S345). When processing big data, analysis of consultation contents can be used.

The server 10 can derive counseling assistance content from big data processing (S350).

When the consultation ends, the server 10 can generate a final summary (S335).

6 illustrates that the server 10 relays voice between the client terminal 60 and the counselor terminal 70, or at least receives voice data from the counselor terminal 70, but the server 10 is not limited to this. When the server 10 performs artificial intelligence counseling, the server 10 receives voice data from the client terminal 60 (S310), converts it into text data (S315), and, if necessary, sends the text data to the client terminal. It is transmitted to (60) (S320), and when the client manipulates the dialog box of the chat window (A), the phrase label is received from the client terminal (60) (S325), the conversation is labeled (S330), and the text data is analyzed. (S335), the summary is updated (S340), the text data is processed as big data (S345), and the content necessary for counseling is derived (S350) to perform counseling. When the consultation is completed, the server 10 can generate a final summary (S355).

Figure 7 is a system diagram showing a chatbot consultation system according to another embodiment of the present invention. Figure 8 is a graph showing Russell's emotional space applied to the psychological counseling agent chatbot device according to the present invention. Refer to Figures 1 to 6.

Referring to FIG. 7, the chatbot counseling system may include a psychological counseling agent chatbot server 10 and a client terminal 60. The psychological counseling agent chatbot server 10 and the client terminal 60 may correspond to the server 10 and the client terminal 60 of FIGS. 1 to 6, respectively. The psychological counseling agent chatbot server 10 is a big data and artificially learned server that can provide artificial intelligence psychological counseling services. The client terminal 60 can provide psychological counseling directly with the psychological counseling agent chatbot server 10 rather than with a counselor.

The client terminal 60 can run the psychological counseling agent chatbot server 10 and an application that can receive non-face-to-face psychological counseling. The application may correspond to the 'counseling app' of FIGS. 1 to 6. The counseling app not only displays the conversation content converted from voice to text in a dialog box, but also allows the user (client) to directly write a text conversation for chatting and send it to the psychological counseling agent chatbot server (10). Conversations written by the client can be displayed in the chat window (A) of the counseling app.

The client terminal 60 may execute the application and transmit a counseling sentence for psychological counseling to the psychological counseling agent chatbot server 10. The client terminal 60 may receive a counseling progress sentence transmitted from the psychological counseling agent chatbot server 10.

The client terminal 60 can receive psychological counseling regarding current emotions by transmitting and receiving counseling sentences and counseling progress sentences with the psychological counseling agent chatbot server 10.

Counseling with clients of the client terminal 60 may be mainly performed in the counseling unit 150 of the psychological counseling agent chatbot server 10.

The psychological counseling agent chatbot server 10 recognizes whether the current emotion is a negative emotion (e.g., sadness, anger, fear, irritation, loneliness, etc.) through text-based conversation, and reduces the size of the negative emotion through appropriate conversation. It may be an application of chatbot agent technology to do this. Additionally, the psychological counseling agent chatbot server 10 can perform efficient psychological counseling by storing key content related to psychological counseling in a database with the user's consent.

Referring to Figure 8, the graph of Russell's emotion space shows 21 different types of emotions in a two-dimensional emotion space composed of two independent parameters, arousal and pleasure axes. It is mapped. Specifically, in the graph shown in Figure 8, the emotions located in the first and fourth quadrants are positive emotions (e.g., comfort, happiness, joy, excitement, etc.), and the emotions located in the second and third quadrants are positive emotions. Negative emotions (e.g., sad, angry, fearful, irritated, etc.) Therefore, if the user has positive emotions, just congratulate him and join in on his joy. However, if the user has negative emotions, it is necessary to actively lead a conversation to alleviate those emotions.

The server communication unit 115 may receive a counseling sentence transmitted from a user terminal, and may transmit a counseling progress sentence generated by the counseling unit 150 to the user terminal.

Clients can film their face or input their voice using a camera or microphone. The text converter 130 extracts an input sentence from the user's mouth shape through facial recognition, or converts the user's voice into text using speech-to-text (STT) to obtain a consultation sentence. can do.

The text converter 130 can reversely convert text into voice. The text converter 130 can convert the counseling progress sentences into voice data using text-to-speech (TTS) and output it as a voice from the speaker of the client's terminal 60.

The data analysis unit 135 may parse the counseling sentence received from the server communication unit 115 or converted from the server storage unit 120 and divide it into a plurality of words. The data analysis unit 135 may perform morpheme analysis and/or parsing on the input sentence using natural language processing techniques. For morpheme analysis (extraction), refer to the previous content.

Syntactic analysis may be to determine the structure of the sentence by dividing the sentence into a plurality of words and analyzing the relationship between the plurality of divided words.

The data analysis unit 135 may perform syntax analysis on the input sentence according to the hexadecimal principle. For example, if the data analysis unit 135 assumes that the counseling sentence received from the server communication unit 115 is “I am sad,” the data analysis unit 135 may change the counseling sentence to ‘who: me’ and ‘ What: Can be divided into multiple words, such as 'sad'. The data analysis unit 135 can estimate and fill in default values or omitted words such as 'how: emotion' and 'when: present'.

The data analysis unit 135 may extract at least one specific emotion-related word representing the user's emotion from the plurality of divided words. The data analysis unit 135 may extract the at least one specific emotion-related word that matches a preset emotion word from among the plurality of words. For example, the data analysis unit 135 may extract 'sadness', a specific emotion-related word, from the word 'what: sad' among the plurality of words.

The server storage unit 120 may perform a database function. The server storage unit 120 can store various information related to psychological counseling in a database. The server storage unit 120 may store the contents of the client's previous psychological counseling. The contents of the consultation may be stored as the ‘summary’ described above, or may be stored in parallel.

The data analysis unit 135 may store the extracted specific emotion-related word 'sadness' in the 'mood_status' value in the server storage unit 120.

The server storage unit 120 may store the user's consultation content obtained through sentence-based conversation. Due to the nature of psychological counseling, if the user is reminded of a previous cause or event, negative emotions may arise again in the user. Therefore, re-inquiring the user about the cause of the negative emotion, the event that caused the negative emotion, etc., which was previously queried, may cause irritation. Accordingly, after storing the consultation contents in the server storage unit 120, a function to retrieve the stored consultation contents when necessary can be added to prevent negative emotions from occurring again in the user. The contents of the consultation can be extracted from the ‘summary’.

The server storage unit 120 stores a counseling sentence input through the server communication unit 115 or converted by the text conversion unit 130, a plurality of words obtained through the data analysis unit 135, and at least one specific emotion extracted. Various counseling contents obtained through conversation with the user, such as related words and counseling progress sentences generated by the counseling unit 150, can be stored.

The counseling unit 150 may calculate the emotion size for at least one specific emotion-related word extracted from the data analysis unit 135, generate a counseling progress sentence according to the emotion size, and transmit it to the client terminal 60.

The counseling unit 150 may generate sentences for conducting psychological counseling based on various counseling techniques.

Counseling techniques may include various known counseling techniques. For example, counseling techniques may include open questions, restatement and reflection, objectification of emotions, conversation centered on empathic responses, waiting, etc. These counseling techniques enable systematic counseling and can help the user be in a comfortable state.

The consultation progress sentence may be a sentence provided to the user to proceed with the consultation. Specifically, the counseling progress sentences may include various sentences for conducting the consultation, such as question sentences for understanding the user's emotional state, response sentences for responding to the user's input sentences, etc.

The counseling unit 150 may calculate the emotion magnitude according to the emotion importance set for at least one specific emotion-related word. The emotion magnitude may be set in the server storage unit 120 as an emotion value assigned to each word representing a negative emotion.

If the emotion size is greater than or equal to a set standard size, the counseling unit 150 may determine that the user's emotion is a negative emotion and generate the counseling progress sentence asking about the motive for causing the negative emotion. That is, when the counseling unit 150 determines that at least one specific emotion-related word is a negative user emotion, the counseling unit 150 may generate a question sentence related to the at least one specific emotion-related word. At this time, negative emotions can include sadness, anger, fear, surprise, and loneliness.

Specifically, the counseling unit 150 may generate a counseling progress sentence to identify at least one of the cause and reason of negative emotions, the magnitude of negative emotions, and episodes related to negative emotions.

For example, if the mood_status value in the server storage unit 120 is stored as 'sadness', the counseling unit 150 generates a question sentence to determine the cause, reason, and magnitude of the user's sad emotions, etc. can do. At this time, the size of the emotion may be stored in the mood_val value in the counseling server storage unit 120.

If the size of the sad emotion is very large, other emotions such as guilt or episodic memories may be mixed in, so the counseling unit 150 may generate a question sentence to identify the event or episode that caused the sad emotion. You can.

The counseling unit 150 may generate a response sentence to relieve the emotion by quoting expressions such as proverbs, sayings, and proverbs depending on the cause of the emotion. Additionally, the counseling unit 150 may generate a response sentence to recommend a meditation technique to the user or play music to relieve the emotion, depending on the cause of the emotion. The counseling unit 150 can perform these functions with the help of the big data processing unit 140 and the data analysis unit 135. This may apply throughout the specification.

If the emotion size is lower than the reference size, the counseling unit 150 may determine that the user's emotion is not a negative emotion and generate the counseling progress sentence to respond to the sentence. That is, if the counseling unit 150 determines that the emotion is not negative, it can induce the user's emotion to be resolved by generating a response sentence to agree with the user's story.

Meanwhile, the description of the above-mentioned question sentences and response sentences is only an example, and question sentences and response sentences with various contents may be generated depending on the embodiment.

When the consultation ends, the counseling unit 150 may generate a question sentence to inquire whether the content of the consultation with the user, that is, the consultation sentence and the consultation progress sentence, are saved. Accordingly, if the user wishes to save the consultation contents, the consultation contents may be stored in the server storage unit 120.

Figure 9 is a flowchart showing the operation method of the psychological counseling agent chatbot server of Figure 8. The psychological counseling agent chatbot server 10 may perform a method of providing a psychological support chatbot service.

Referring to FIG. 9, the data analysis unit 135 included in the psychological counseling agent chatbot server 10 may parse the counseling sentence received from the server communication unit 115 and divide it into a plurality of words (S410). . The server communication unit 115 may receive the counseling sentence transmitted from the user terminal and transmit it to the data analysis unit 135.

The data analysis unit 135 may parse the counseling sentence and divide it into a plurality of words. Syntactic analysis may be to determine the structure of the sentence by dividing the counseling sentence into a plurality of words and analyzing the relationship between the plurality of divided words.

The data analysis unit 135 may extract at least one specific emotion-related word representing the user's emotion from among the plurality of words (S415). The data analysis unit 135 may extract the at least one specific emotion-related word that matches a preset emotion word from among the plurality of words. The data analysis unit 135 may store the at least one specific emotion-related word in the server storage unit 120.

The counseling unit 150 included in the psychological counseling agent chatbot server 10 may calculate the emotion size for the at least one specific emotion-related word (S420) and determine whether the emotion size is greater than or equal to a set standard size (S420). S425).

The counseling unit 150 may calculate the emotion magnitude according to the emotion importance set for at least one specific emotion-related word. The emotion magnitude may be set in the server storage unit 120 as an emotion value assigned to each word representing a negative emotion.

If the emotion size is greater than or equal to a set standard size, the counseling unit 150 may determine that the user's emotion is a negative emotion and generate the counseling progress sentence asking about the motive for causing the negative emotion (S430). If the emotion size is lower than the reference size, the counseling unit 150 determines that the user's emotion is not a negative emotion, and may generate the counseling progress sentence to respond to the sentence (S435).

After the counseling progress sentence generation step (S430 or S435), the counseling unit 150 may transmit the counseling progress sentence to the server communication unit 115 and transmit it to the user terminal (S440).

At the end of the consultation, the counseling unit 150 may generate a question sentence asking whether to save the counseling sentence and the counseling progress sentence and transmit the question to the client terminal 60 (S445).

FIG. 10 is a flowchart showing an example of the operation of the psychological counseling agent chatbot device shown in FIG. 9.

Figure 10 assumes that the psychological counseling agent chatbot server 10 receives the counseling sentence “I am sad” from the user and stores the emotion-related sentence component “sad” in the mood_status value.

The psychological counseling agent chatbot server 10 may generate a question sentence about whether the user knows the cause of the emotion and provide it to the user (S455).

That is, at this time, if the user does not know the cause of the emotion, the psychological counseling agent chatbot server 10 may generate a question sentence to request a story about the negative emotion and provide it to the user. Since users may not know the cause of their emotions, they can find out the cause of their emotions by freely talking about their current negative emotions. For example, if the user is depressed for no reason, when talking about depressed feelings, it can be seen that the weather is cloudy and the user feels depressed and feels depressed.

The psychological counseling agent chatbot server 10 may provide a question sentence about why the emotion occurred (S460). In other words, if the user knows the cause of the emotion, the psychological counseling agent chatbot server 10 can generate and provide a question sentence about the specific reason why the emotion occurred.

The psychological counseling agent chatbot server 10 can provide a question sentence about the magnitude of emotion (S465). For example, users can enter the magnitude of their emotions on a 5-point scale. Here, having the user state the reason and magnitude of the emotion is emotion objectification, which is one of the counseling techniques, and has the effect of allowing the user to look at his or her own emotions.

The psychological counseling agent chatbot server 10 may determine whether the value corresponding to the size of the emotion is smaller than a preset value (S470). That is, the psychological counseling agent chatbot server 10 can determine whether the negative emotion is weak or strong depending on whether the value corresponding to the size of the emotion is smaller than, for example, 3 (a preset value). For example, if the value corresponding to the size of the emotion is 1 to 2 according to the judgment result in the emotion size determination step (S470), the psychological counseling agent chatbot server 10 may determine that the negative emotion is weak.

The psychological counseling agent chatbot server 10 may provide a question sentence to request an episode related to emotions (S475). Accordingly, the user can input episodes related to emotions.

The psychological counseling agent chatbot server 10 may provide a response sentence to an emotion-related episode (S480). In other words, the psychological counseling agent chatbot server 10 can provide a response sentence to appropriately respond to an emotion-related episode.

For example, if the user feels sad because he or she is reminded of his or her deceased dog (name: Leo), the psychological counseling agent chatbot server 10 may provide a question sentence to request a positive episode related to the dog. Thereafter, when the user inputs the input sentence “Leo liked to walk with me,” the psychological counseling agent chatbot server 10 may parse the input sentence to determine the user's episode. In addition, the psychological counseling agent chatbot server 10 sends a response sentence to empathize with the user's emotions, such as “It must have been fun,” and a question sentence to request another episode from the user, such as “Leo liked to walk.” By providing this, psychological counseling can be performed to comfort the user.

The psychological counseling agent chatbot server 10 may provide a question sentence to request additional episodes related to emotions. At this time, if the user wants to talk about additional episodes related to emotions, the user can continuously conduct a conversation with the psychological counseling agent chatbot server 10. On the other hand, if the user wants to stop the conversation, the psychological counseling agent chatbot server 10 may provide a question sentence to determine whether the negative emotions have been resolved.

On the other hand, if the value corresponding to the size of the emotion is greater than or equal to the preset value according to the judgment result in the emotion size determination step (S470), the psychological counseling agent chatbot server 10 is used to determine the potential cause of the emotion. A question sentence can be provided (S485). For example, if the value corresponding to the magnitude of the emotion is 3 to 5, the psychological counseling agent chatbot server 10 may determine that the negative emotion is strong. In this case, the psychological counseling agent chatbot server 10 may provide a question sentence to confirm that there may be other potential causes related to negative emotions.

The psychological counseling agent chatbot server 10 may provide a question sentence to request an episode related to the potential cause of the emotion (S490). For example, in response to a provided question sentence, a user may enter an episode about failing to give medicine to a dead dog when it was sick.

The psychological counseling agent chatbot server 10 may provide a response sentence to an episode related to the potential cause of the emotion (S495). For example, when a user feels sad due to guilt about not being able to give medicine to a dead dog when it was sick, the psychological counseling agent chatbot server (10) expresses the guilt by saying, "That's right. Leo will understand you too." You can provide a response sentence to alleviate the problem. As another example, the psychological counseling agent chatbot server 10 may perform functions to comfort the user, such as recommending a meditation method to relieve guilt or playing music that Leo liked. Additionally, the psychological counseling agent chatbot server 10 may provide a question sentence to request additional episodes related to the potential cause of the emotion.

The psychological counseling agent chatbot server 10 can determine whether to end the counseling (S500). The psychological counseling agent chatbot server 10 may provide a question sentence to inquire whether to end the counseling. In other words, the psychological counseling agent chatbot server 10 may ask whether the user wants to continue counseling when the user's negative emotions have been resolved to some extent. If the user wishes to continue counseling, the psychological counseling agent chatbot server 10 may repeatedly perform counseling with the user.

In contrast, the psychological counseling agent chatbot server 10 performs the 'waiting' function of counseling by waiting for the user's input for a preset time (e.g., 1 minute) when the user's negative emotions are not resolved. can do. If the preset time has passed, the psychological counseling agent chatbot server 10 may repeatedly perform counseling with the user.

Additionally, the psychological counseling agent chatbot server 10 may provide a question sentence to inquire whether or not to store the content of the counseling session with the user. For example, before ending the consultation, a question sentence can be provided to inquire about consent to saving the consultation content, such as “Mina is sad because her pet dog Leo died two weeks ago.” At this time, if the user agrees to save the counseling content, the psychological counseling agent chatbot server 10 may store the counseling content in the server storage unit 120.

Meanwhile, in the above example, the case where the mood_status value is 'sadness' was explained, but the psychological counseling agent chatbot server 10 uses the same method as described above for other negative emotions such as anger, fear, irritation, loneliness, etc. Psychological counseling can be performed.

10: Server 60: Client terminal
70: Counselor terminal 110: Server control unit
115: server communication unit 120: server storage unit
125: App management unit 130: Text conversion unit
135: Data analysis unit 140: Big data processing unit
150: Counseling unit 210: Terminal control unit
220: terminal communication unit 230: terminal input unit
240: terminal control unit 250: terminal output unit
255: display module 260: terminal storage unit

Claims (5)

Receiving voice data, which is the content of the consultation between the client and the counselor, from at least the counselor terminal (hereinafter referred to as 'user terminal') among the client terminal and the counselor terminal;
converting the voice data into text data;
transmitting the text data to the user terminal; and
Upon receiving a phrase label manipulated by the user from the user terminal, labeling the phrase label to the corresponding conversation,
Receiving a counseling sentence for artificial intelligence counseling from the client terminal, parsing the counseling sentence, and dividing the counseling sentence into a plurality of words;
extracting at least one specific word representing a user's emotion from among the plurality of words;
calculating an emotional level for the at least one specific word and generating a counseling progress sentence according to the emotional level; and
A method of providing a psychological support chatbot service, further comprising: transmitting the counseling progress sentence to the user terminal. According to claim 1,
analyzing the text data with reference to the phrase label; and
A method of providing a psychological support chatbot service, further comprising: updating a summary using the analyzed counseling content. According to claim 2,
Processing the text data as big data to generate a big data processed product; and
A method of providing a psychological support chatbot service, further comprising: deriving counseling assistance content from the big data processed product. According to claim 1,
The step of extracting the specific word includes extracting the at least one specific word that matches a preset emotional word among the plurality of words, and storing the at least one specific word in a database for consultation,
The step of generating the counseling progress sentence calculates the emotion size according to the emotional importance set for the at least one specific word,
The step of generating the counseling progress sentence includes determining that the user's emotion is a negative emotion if the emotion size is greater than a set standard size, and generating the counseling progress sentence asking about the motive for causing the negative emotion,
The counseling progress sentence asking about the motive for causing the user's emotion is a sentence asking about at least one of the cause and reason of the negative emotion and the episode about the negative emotion. Method for providing a psychological support chatbot service. A server that communicates data with at least a counselor terminal (hereinafter referred to as “user terminal”) among the client terminal and counselor terminal,
a server communication unit that receives voice data, which is the content of the consultation between the client and the counselor, from the user terminal;
a text converter that converts the voice data into text data;
A counseling app for installation is transmitted to the user terminal to be installed, the text data is displayed on the installed counseling app, and when a phrase label manipulated by the user is received from the user terminal, the phrase label is added to the conversation. App management department for labeling;
a data analysis unit that analyzes the text data with reference to the phrase label and updates a summary using the analyzed counseling content; and
It includes a big data processing unit that processes the text data into big data to generate a big data processed product, and derives counseling assistance content from the big data processed product,
The data analysis unit receives a counseling sentence transmitted from the client terminal, parses the counseling sentence, divides it into a plurality of words, and extracts at least one specific word representing the user's emotion from the plurality of words,
A psychological counseling agent chatbot server further comprising a counseling unit that calculates an emotional level for the at least one specific word and generates a counseling progress sentence according to the emotional level.

Images