Procrastination habit management chatbot system and method

Abstract

The present invention relates to a procrastination habit management chatbot system and method capable of performing AI chatbot conversations enabling natural conversations for the purpose of improving a procrastination habit, performing intention analysis for the purpose of encouraging a user to think as in actual consultation, and identifying a cause of the procrastination through a response of the user in the conversation with the chatbot. The procrastination habit management chatbot system of the present invention includes: a server including a procrastination habit improvement unit configured to improve a procrastination habit through conversation between a user and a chatbot, a response production unit of the chatbot configured to produce a response for the chatbot, and a procrastination habit classification unit configured to classify the procrastination habit; and a user terminal configured to execute a chatbot interface with the user.

Description

Procrastination Habit Management Chatbot System and Method

[0001] The present invention relates to a chatbot system and method for managing procrastination habits, and more particularly, to monitoring mental health and habits through mobile or web and providing personalized advice, and particularly, in the case of developing mental health and habit intervention through chatbot, to providing a conversation flow and logic that can monitor an individual's mental health and habits and bring about cognitive and behavioral changes.

[0002] In addition, the present invention relates to a chatbot system and method for managing procrastination habits that can provide a method for generating dialogue for chatbot interaction by simultaneously using three fields: a rule-based chatbot in which the dialogue order is determined based on a natural language processing model necessary for achieving the purpose of the chatbot or the dialogue flow, and the dialogue progresses according to the rules of each order; a retrieval-based chatbot that uses an AI classification model to output the next sentence of the chatbot from a group of chatbot answers that have been previously written, and a generative chatbot that outputs an answer using an AI generation model according to the user's response.

[0003] There is a lot of research in the field of mental health on monitoring mental health or habit management and providing personalized advice using chatbots. However, most of the monitoring is not done through natural conversation, but through questionnaire responses via chatbots.

[0004] In addition, rather than taking a psychological approach to habit management, there are many cases where only praise or habit management methods are provided. In addition, there are few chatbots that focus on procrastination, which is the most important behavior in habit management, or they are only simple rule-based chatbots, making natural conversation impossible.

[0005] Cognitive behavioral therapy is widely used as a method to modify existing procrastination behaviors. However, applying offline cognitive behavioral therapy methods to chatbots has limitations such as a long process and too many variables. To apply this, a modified version of the chatbot version of cognitive behavioral therapy sessions is required.

[0006] In order to enable natural conversation using AI models rather than existing rule-based chatbots, intent must be identified. However, the intent analysis models of existing chatbots often only identify whether the intent is a question, a command, or a request for information, and are therefore not appropriate for identifying intent within the cognitive behavioral therapy process.

[0007] Furthermore, because there was no focus on procrastination in the past, the causes were found through empirical methods such as surveys used in counseling or behavioral observations. However, it takes a lot of time to identify procrastination habits through surveys while using chatbots, and continuous measurement is impossible.

[0008] Therefore, while it is important to naturally identify the cause of procrastination in a conversation, research has continued to classify user responses, similar to the existing method of classifying the cause of procrastination in natural language.

[0009] As an example, Korean Patent Publication No. 10-2020-0164100 discloses a method for classifying user queries in a chatbot, which inputs a user query into a language model dialogue engine, determines whether a score output from the language model dialogue engine satisfies a predetermined standard, and classifies the query type into a language model type if the score satisfies the predetermined standard, and classifies the query type from the user query if the score does not satisfy the predetermined standard, stores word patterns related to searches, compares the user query with these patterns to determine whether it is a search-related type, and stores word patterns included in typical conversations in a specific field to which the chatbot is applied, and compares the user query with these patterns to determine whether it is a scenario-based model type.

[0010] However, even in this case, natural conversation is not possible for the purpose of improving the user's habits, and there is a disadvantage in that intention analysis is performed for the purpose of encouraging the user to think as in an actual consultation, and the cause cannot be identified through the user's response in the conversation with the chatbot.

[0011] The purpose of the present invention is to provide a procrastination habit management chatbot system and method that performs a natural conversational AI chatbot conversation for the purpose of improving the procrastination habit.

[0012] Another purpose of the present invention is to provide a chatbot system and method for managing procrastination habits, which performs intention analysis for the purpose of encouraging the user to think as in an actual consultation and can identify the cause of procrastination through the user's response in a conversation with the chatbot.

[0013] The procrastination habit management chatbot system according to the present invention may include a server including a procrastination habit improvement unit that performs procrastination habit improvement through conversation between a user and a chatbot, a chatbot response output unit that performs response output from the chatbot, and a procrastination habit classification unit that performs procrastination habit classification, and a user terminal that performs a chatbot interface with the user.

[0014] Here, the procrastination improvement unit can communicate with users through long and short conversations over a period of time.

[0015] Additionally, long conversations can be sequentially conducted to get to know myself, understand my strengths and weaknesses at work, reflect on my actions, deal with my fears, and see how I can change.

[0016] Here, short conversations can be sequentially followed by longer conversations about why you're getting your daily tasks done, providing personalized feedback on your plan's progress and procrastination habits, developing action strategies and practicing with a chatbot, and learning effective time management techniques.

[0017] Additionally, the chatbot's response output section can extract chatbot questions from the scenario DB within the chatbot's response output section.

[0018] Here, the chatbot's response output unit can receive a user response from the user terminal in response to a chatbot question.

[0019] Additionally, the chatbot's response output section can classify the user's response as a normal response, 'I don't know', or 'I'm skeptical' in the response classification model within the chatbot's response output section.

[0020] Here, the chatbot's response output unit can generate a sentence again if the response classification model's classification is a normal response and the sentence generation model within the chatbot's response output unit determines that the normal response ends with a question.

[0021] In addition, the chatbot's response output unit may be characterized in that, when the classification of the response classification model is 'I don't know', the chatbot performs a chatbot utterance for a supplementary explanation to a question in the supplementary explanation DB within the chatbot's response output unit and receives a user response from the user terminal.

[0022] Here, the chatbot's response output unit may be characterized by performing a chatbot utterance for an explanation of the justification for a question in the justification provision DB within the server when the classification of the answer classification model is 'skeptical' and receiving a user response from the user terminal.

[0023] Additionally, the chatbot's response output section can perform chatbot utterances for the auxiliary explanation DB and the justification provision DB, and if the user response received from the user terminal ends with a question, the sentence can be regenerated in the sentence generation model.

[0024] Here, the procrastination habit classification unit can perform preprocessing on user responses on the server.

[0025] In addition, the procrastination habit classification unit can classify the cause of procrastination by providing the procrastination probability vector for the cause of procrastination together with the result of the preprocessing in the procrastination habit analyzer in the procrastination habit classification unit.

[0026] Here, the causes of procrastination may include mental stress and physical stress, which is closely related to mental stress.

[0027] Additionally, causes of procrastination may include social relationships, which are factors that prompt or stop procrastination.

[0028] Here, causes of procrastination may include environmental factors that distract attention.

[0029] Additionally, causes of procrastination can include perfectionism, which makes it difficult to get started due to uncertainty about the future and fear of being evaluated.

[0030] Here, causes of procrastination may include lack of motivation, which reduces self-control and lowers work efficiency.

[0031] Additionally, causes of procrastination may include poor concentration or lack of self-control.

[0032] A method for managing procrastination according to another embodiment of the present invention may include a procrastination habit improvement step in which a procrastination habit improvement unit communicates with a user through a chatbot via a user terminal that performs a chatbot interface with the user to receive questions for improving the procrastination habit and responses from the user, a chatbot response output step in which a chatbot response output unit outputs a chatbot response corresponding to the user response, and a procrastination habit classification step in which a procrastination habit classification unit analyzes the user response to classify the procrastination habit.

[0033] Here, in the procrastination habit improvement stage, you can converse with users through long and short conversations over a certain period of time.

[0034] Additionally, long conversations can be sequentially conducted to get to know myself, understand my strengths and weaknesses at work, reflect on my actions, deal with my fears, and see how I can change.

[0035] Here, short conversations can be sequentially followed by longer conversations about why you're getting your daily tasks done, providing personalized feedback on your plan's progress and procrastination habits, developing action strategies and practicing with a chatbot, and learning effective time management techniques.

[0036] Additionally, the chatbot response output step may include a chatbot question extraction step that extracts a chatbot question from a scenario DB in the chatbot response output section.

[0037] Here, the response output step of the chatbot may include a user response reception step of receiving a user response from a user terminal in response to a chatbot question.

[0038] Additionally, the response output stage of the chatbot may include a user response classification stage that classifies the user response into one of a normal response, 'I don't know', and 'I am skeptical' in the response classification model within the response output section of the chatbot.

[0039] Here, in the user response classification step, if the classification is a normal response, a normal response processing step may be included to generate a sentence again in the sentence generation model in the response output section of the chatbot if the normal response ends with a question.

[0040] In addition, in the user response classification step, if the classification is 'I don't know', an auxiliary explanation processing step may be included in which the chatbot performs a chatbot utterance for an auxiliary explanation for the question in the auxiliary explanation DB in the chatbot's response output section and receives a user response from the user terminal.

[0041] Here, in the user response classification step, if the classification is 'skeptical', a justification explanation processing step may be included in which a chatbot utters a justification explanation for the question in the justification provision DB in the chatbot's response output section and receives a user response from the user terminal.

[0042] In addition, if a user response received from a user terminal for a chatbot utterance performed in the auxiliary explanation processing step and the justification explanation processing step ends with a question, a sentence regeneration step may be included in which a sentence is regenerated in the sentence generation model.

[0043] Here, the procrastination habit classification step may include a user response preprocessing step that performs preprocessing on user responses on the server.

[0044] In addition, the procrastination habit classification step may include a procrastination cause classification step that classifies the cause of procrastination by providing a procrastination habit probability vector for the cause of procrastination together with the result of the preprocessing in the procrastination habit analyzer in the procrastination habit classification unit.

[0045] Here, the causes of procrastination may include mental stress and physical stress, which is closely related to mental stress.

[0046] Additionally, causes of procrastination may include social relationships, which are factors that prompt or stop procrastination.

[0047] Here, causes of procrastination may include environmental factors that distract attention.

[0048] Additionally, causes of procrastination can include perfectionism, which makes it difficult to get started due to uncertainty about the future and fear of being evaluated.

[0049] Here, causes of procrastination may include lack of motivation, which reduces self-control and lowers work efficiency.

[0050] Additionally, causes of procrastination may include poor concentration or lack of self-control.

[0051] The chatbot system and method for managing procrastination habits according to the present invention have the advantage of performing AI chatbot conversations that enable natural conversations for the purpose of improving procrastination habits.

[0052] In addition, the chatbot system and method for managing procrastination habits according to the present invention have the advantage of performing intention analysis for the purpose of encouraging the user to think as in actual counseling, and identifying the cause of procrastination through the user's response in a conversation with the chatbot.

[0053] Figure 1 is a schematic diagram showing a procrastination habit management chatbot system according to one embodiment of the present invention.

[0054] Figure 2 is a drawing showing in detail the procrastination habit improvement part of Figure 1.

[0055] Figure 3 is a drawing showing in detail the response output section of the chatbot of Figure 1.

[0056] Figure 4 is a drawing showing in detail the procrastination habit classification section of Figure 1.

[0057] Figure 5 is a flowchart illustrating a procrastination habit management method according to one embodiment of the present invention.

[0058] Figure 6 is a flowchart showing in detail the response output step of the chatbot of Figure 5.

[0059] Figure 7 is a flowchart showing in detail the procrastination habit classification steps of Figure 5.

[0060]

[0061] Hereinafter, specific embodiments for carrying out the present invention will be described with reference to the attached drawings.

[0062] When describing the present invention, terms such as "first" and "second" may be used to describe various components. However, the components may not be limited by these terms. The terms are used solely to distinguish one component from another. For example, without departing from the scope of the present invention, the first component could be referred to as the "second component," and similarly, the second component could also be referred to as the "first component."

[0063] When it is said that a component is connected or connected to another component, it can be understood that it may be directly connected or connected to that other component, but there may also be other components in between.

[0064] The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the present invention. Singular expressions may include plural expressions unless the context clearly dictates otherwise.

[0065] In this specification, terms such as “include” or “have” are intended to indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the specification, and can be understood as not excluding in advance the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0066] Additionally, the shape and size of elements in the drawing may be exaggerated for clearer explanation.

[0067] Hereinafter, the procrastination habit management chatbot system and method according to the present invention will be described in detail with reference to the attached drawings.

[0068]

[0069] FIG. 1 is a schematic diagram showing a procrastination habit management chatbot system according to one embodiment of the present invention, and FIGS. 2 to 4 are detailed drawings for explaining FIG. 1 in detail.

[0070] Hereinafter, a procrastination habit management chatbot system according to one embodiment of the present invention will be described with reference to FIGS. 1 to 4.

[0071] First, referring to FIG. 1, a procrastination habit management chatbot system according to one embodiment of the present invention is composed of a server (100) including a procrastination habit improvement unit (110) that performs procrastination habit improvement through conversation between a user and a chatbot, a chatbot response output unit (120) that performs response output from the chatbot, and a procrastination habit classification unit (130) that performs procrastination habit classification, and a user terminal (200) that performs a chatbot interface with a user.

[0072] The procrastination habit management chatbot system according to the present invention monitors mental health and habits through conversations with a chatbot via mobile or web and provides personalized advice (mhealth / ehealth with chatbot). Various AI (Artificial Intelligence) technologies such as natural language processing technology or classification models can be incorporated as needed.

[0073] In addition, the procrastination habit management chatbot system according to the present invention can provide a method for generating a conversation for chatbot interaction by simultaneously using a rule-based chatbot in which the conversation flow and logic of the chatbot are designed using artificial intelligence and natural language processing, and the conversation order is determined based on a natural language processing model necessary for achieving the purpose of the chatbot or the conversation flow, and the conversation proceeds according to the rules of each order, a retrieval-based chatbot that uses an AI classification model to output the next sentence of the chatbot from a group of chatbot answers that have been previously written, and a generative chatbot that outputs an answer using an AI generation model according to a user's response.

[0074] Meanwhile, the procrastination habit improvement unit (110) of the present invention provides an AI chatbot conversation system capable of natural conversation for the purpose of improving the procrastination habit, thereby providing a cognitive behavioral therapy method used to improve the procrastination habit in existing offline counseling, which is described in detail in FIG. 2.

[0075] In addition, the chatbot's response output section (120) analyzes the user's response intention for the purpose of encouraging the user to help his or her thoughts as in actual counseling when conversing with the cognitive behavioral therapy chatbot for improving procrastination habits, as described in detail in Fig. 3.

[0076] Meanwhile, the procrastination habit classification unit (130) classifies the causes of procrastination into six categories and identifies and provides the user's procrastination causes through the user's response in a conversation with a chatbot, which is described in detail in Fig. 4.

[0077]

[0078] Figure 2 is a drawing showing in detail the procrastination habit improvement unit (110) of Figure 1.

[0079] As can be seen in Figure 2, the procrastination habit improvement unit (110) converses with the user through long and short conversations over a certain period of time.

[0080] Here, the long conversation proceeds sequentially through getting to know myself, knowing my strengths and weaknesses at work, reflecting on my actions, dealing with my fears, and finding out how I can change myself.

[0081] Additionally, short conversations are sequentially conducted along with longer conversations about the reason for the progress of daily tasks, the rate of plan achievement and customized feedback on procrastination habits, establishing an action strategy and practicing with the chatbot, and learning effective time management techniques.

[0082] The procrastination habit improvement unit (110) according to the present invention includes cognitive behavioral therapy sessions for procrastination treatment and some life crafting for motivation, and is conducted in 5 sessions, with one session per week, so that the chatbot can be used for a total of 5 weeks.

[0083] A session consists of one day of long conversations on a specific topic, and several days of short conversations on that topic, where you practice and receive feedback. The long conversations last from 3 to 7 minutes, and the short conversations last from 10 seconds to 3 minutes.

[0084] For example, for the first week, in the long conversation 'Getting to know myself', you can ask about hobbies, likes, career, procrastination habits, etc., and in the short conversation 'Reasons for performing daily tasks', you can have a conversation every evening about why you did / didn't do well on your tasks that day, as follows.

[0085] Chatbot (long conversation): What kind of lifestyle do you want to have?

[0086] User: I want to live and travel around.

[0087] Chatbot (short conversation): Why do you procrastinate?

[0088] User: I think I miscalculated the time.

[0089] Next, for the second week, in the long conversation, 'Know the pros and cons of working', you can ask about strengths and routines, tell them the main causes of procrastination found in the procrastination habit classification algorithm, and ask them to make short-term and long-term plans, and in the short conversation, 'Plan achievement rate and procrastination habit customized feedback', you can provide positive feedback or personalized advice to do work well according to the plan success rate.

[0090] Chatbot (long conversation): What motivates you to work?

[0091] User: When you do a good job, does life feel worthwhile and fun?

[0092] Chatbot (short conversation): You weren't just putting off work for no particular reason, were you? User: Yeah.

[0093] Chatbot: Let's plan each day a little more precisely. We'll create a plan by the hour and meticulously manage our tasks with a checklist. How about that?

[0094] Next, for the third week, in the long conversation, 'Reviewing the actions of the past week', you can think of a time when you did something on time and ask about your thoughts and feelings at that time, or think of a time when you put off something and ask about your thoughts and feelings at that time, and in the short conversation, 'Establish an action strategy and practice with the chatbot', you can share your own strategy with the chatbot and make a promise to stick to it, as shown below.

[0095] Chatbot (long conversation): Think of a time in the past week when you did a good job without procrastinating. Can you tell me more about that situation?

[0096] User: I don't think I put off work when I had important work to do the next day.

[0097] Chatbot (short conversation): We devised a strategy to achieve our goal. My strategy was to wake up at 9 a.m. What about you? If you remember, say yes. If you don't, say no.

[0098] User: Yeah

[0099] Chatbot: What was it?

[0100] User: "Same time every day!"

[0101] Next, for the fourth week, you could have a long conversation, "Dealing with Fears," where you teach people how to recognize and deal with fear, the primary psychological cause of procrastination. You could also have a short conversation, "Learning Effective Time Management Skills," where you teach users one effective way to manage their time each day, like this:

[0102] Chatbot (long conversation): Hello! Today, we're going to learn about 'worry' and 'fear', the main causes of procrastination. Additionally, I've included links to meditation and stress-relief exercises recommended by experts below.

[0103] Next, for the fifth week, during the long conversation, you can ask how much you have achieved in your long-term plan in the 'Discover Myself Change' section, or you can have a conversation like the one below to find out how much you have changed through the four sessions with the chatbot.

[0104] Chatbot (long conversation): What positive changes have you noticed before and after talking to me?

[0105] User: I feel like I've become more organized about my day, and I forget things less often.

[0106] In this way, the procrastination habit management chatbot system according to the present invention has the advantage of performing AI chatbot conversations that enable natural conversations for the purpose of improving procrastination habits.

[0107]

[0108] Figure 3 is a drawing showing in detail the response output section (120) of the chatbot of Figure 1.

[0109] As can be seen in FIG. 3, the chatbot's response output unit (120) extracts a chatbot question from a scenario DB (121) within the chatbot's response output unit (120), receives a user answer to the chatbot question from a user terminal (200), and classifies the user answer into one of a normal response, 'I don't know', and 'I'm skeptical' in the answer classification model (122) within the chatbot's response output unit (120).

[0110] Here, if the classification is a normal response, the sentence generation model (123) in the chatbot's response output unit (120) generates a sentence again if the normal response ends with a question, and if it is 'I don't know', the chatbot performs an auxiliary explanation for the question in the auxiliary explanation DB (124) in the chatbot's response output unit (120) and receives a user answer from the user terminal (200). If it is 'skeptical', the chatbot performs an explanation for the justification for the question in the justification provision DB (125) in the server (100) and receives a user answer from the user terminal (200).

[0111] At this time, if the user response received from the user terminal (200) ends with a question, the sentence can be regenerated in the sentence generation model (123).

[0112] The response generation unit (120) of the chatbot according to the present invention classifies intentions so that users can better think about answers to questions, as in actual consultations. That is, unlike existing chatbots, cases where a user answers a question strangely can be labeled as either "difficult to think about the question" (label name: "I don't know") or "unwilling to answer the question" (label name: "I'm skeptical about the question").

[0113] Accordingly, in the present invention, responses can be classified into three types, for example, if the response is correct (label name: normal response), I don't know, and I am skeptical about the question.

[0114] Here, the answer classification model (122) can be used as the intent classification model of the chatbot by fine-tuning the KoBERT pre-trained model.

[0115] Meanwhile, if the user's response is classified as a normal response in this chatbot, the chatbot's response is generated using a sentence generation model (123). The generation model used here can be a fine-tuned version of the pre-trained GPT2 model.

[0116] The data used is data from a survey of multiple users, and sentence sets categorized by keywords such as “emotions, hobbies, studies, general conversation, concentration problems, stress, and insomnia” are available from the open data set provided by AI Hub.

[0117] The data can be used for fine-tuning GPT2 by pairing the user's responses to the chatbot's questions with the expected chatbot responses.

[0118] For example, a typical response might include something related to "exercise, travel, cooking," or a clear response like, "External circumstances? I get really stressed when my plans are disrupted."

[0119] 'I don't know' can include responses such as "I don't really have a problem", "I just don't want to do anything", and "I don't know, I'm having trouble finding it these days, so I think that's what's bothering me".

[0120] 'Skeptical' can include responding in an opposing manner, such as "I don't want to answer," "I can do everything better than you," and "I'm putting off even answering."

[0121] Therefore, the procrastination habit management chatbot system according to the present invention has the advantage of being able to perform intention analysis for the purpose of encouraging the user to think as in actual counseling.

[0122]

[0123] Figure 4 is a drawing showing in detail the procrastination habit classification unit (130) of Figure 1.

[0124] As can be seen in Fig. 4, the procrastination habit classification unit (130) performs preprocessing on the user response, and provides the result of the preprocessing together with the procrastination habit probability vector for the cause of procrastination to the procrastination habit analyzer (131) in the procrastination habit classification unit (130) to classify the cause of procrastination.

[0125] Here, the causes of procrastination can be classified into six categories: mental stress and physical stress that is closely related to mental stress, social relationships that are factors that encourage or stop procrastination, environmental factors that distract attention, perfectionism that makes it difficult to start due to uncertainty about the future and fear of evaluation, lack of motivation that reduces self-control and work rate, and self-control that causes low concentration or boredom.

[0126] Meanwhile, preprocessing removes stop words from the user's responses to the chatbot's questions about the causes of procrastination in the conversation, and then performs morphological analysis using the Okt morphological analyzer provided by Konlpy. If the number of words exceeds 15, the remainder is omitted, and the number of words is unified to 15.

[0127] Meanwhile, we embed sentences into a vector of size 15*300 for each sentence using the fasttext pretrained vector dictionary, which is a Korean embedder.

[0128] The embedded sentences pass through a procrastination habit analyzer (131), which is a multi-layered CNN that inputs a vector created by a fully connected layer with filter sizes of 2*2, 3*3, 4*4, and 5*5, respectively, and a number of filters of 50, into a neural network to produce a probability vector for six causes of procrastination.

[0129] Here, the procrastination habit analyzer (131) is trained using cross entropy as a loss function and Adam as an optimizer, and after passing through the procrastination habit analyzer (131), a probability vector (1*6) corresponding to each procrastination habit factor is output.

[0130] At this point, the probability vectors of user responses to the chatbot's questions regarding multiple procrastination factors are summed, and the largest and second-largest procrastination factors are identified. Because the procrastination factors are not independent, some may overlap.

[0131] Therefore, we can first predict the procrastination factor with the largest value, and if the predicted value is different from the user's expectation, we can predict the user's procrastination factor with the second largest value.

[0132] For example, among the procrastination factors, 'mental / physical stress' can be defined as a cause of procrastination because mental stress and procrastination are closely related, and physical stress often accompanies mental stress. In the conversation, mental / physical stress can be included, such as, "When I'm not sure I can do this and feel anxious?", or "When I get anxious while worrying about the future, it interferes with my work."

[0133] Also, among the procrastination factors, 'social relationships' are an important factor that encourages or stops procrastination. Social support can be a driving force for working hard, but when problems arise in social relationships or you become entangled in relationships, it can be accompanied by mental stress and have a negative effect on procrastination. In terms of conversation content, social relationships can be included, such as, "I think it's stressful! Stress in relationships often makes me unable to do anything...ㅠ" or "I have to work, but if a friend contacts me saying they need counseling, I think I should go."

[0134] Meanwhile, among the procrastination factors, 'environmental factors' can be defined as physical and environmental factors, such as having difficulty concentrating on work or finishing work on time when distracting factors such as cell phones or social media are around. In terms of conversation, it can include distracting factors such as "I do badly when I'm in an environment where I can't concentrate," "I should watch one more YouTube video," or "Netflix..."

[0135] Also, among the procrastination factors, 'perfectionism' is highly related to conscientiousness, so it has a positive effect on procrastination to some extent, but maladaptive perfectionism can make it difficult to start work on time due to uncertainty about the future, fear of evaluation, etc. In the conversation, it can include the desire for perfection, such as "I want to do it well. I wonder if my work efficiency is low because I want the result to be good.", "Making various plans is both a good factor and a hindrance. I get a bit flustered when this situation does not come."

[0136] Meanwhile, among the factors of procrastination, 'lack of motivation' is the fundamental cause of lowering self-control and work rate, and when motivation is lacking, interest in work decreases and it becomes difficult to finish work on time. As a conversation topic, it can include non-motivating content such as "Will what I learned be helpful in this unmotivated situation?", "I feel like I'm getting more vicious by thinking the same thing", and "I don't feel like doing something when it's too difficult or I don't know why I should do it."

[0137] Also, among the factors of procrastination, 'low self-control' is a factor that is so closely related to procrastination that failure of self-control can be said to be procrastination, and includes things like lack of concentration and boredom, and is also related to environmental factors and lack of motivation. When procrastination behavior occurs due to bothersomeness and lack of concentration without a specific cause being known, it can be labeled as low self-control. In the conversation, it can include content that shows an inability to control oneself, such as "It's annoying", "Oh... I really don't want to do it and I can't concentrate. What's wrong with me?", "I think I'm getting in the way of work because I want to rest comfortably and think, 'Isn't this enough?'"

[0138] In this way, the procrastination habit management chatbot system according to the present invention has the advantage of being able to identify the cause of procrastination through the user's response in a conversation with the chatbot.

[0139]

[0140] FIG. 5 is a flowchart illustrating a procrastination habit management method according to one embodiment of the present invention, and FIGS. 6 and 7 are detailed flowcharts for explaining FIG. 5 in detail.

[0141] Hereinafter, a chatbot method for managing procrastination habits according to another embodiment of the present invention will be described with reference to FIGS. 5 to 7.

[0142] First, referring to FIG. 5, a procrastination habit management method according to another embodiment of the present invention comprises a procrastination habit improvement step (S100) in which a user and a chatbot are conversed through a user terminal (200) that performs a chatbot interface in a procrastination habit improvement unit (110) to receive questions for improving the procrastination habit and responses from the user, a chatbot response output step (S200) in which a chatbot response output unit (120) outputs a chatbot response corresponding to the user's response, and a procrastination habit classification step (S300) in which a procrastination habit classification unit (130) analyzes the user's response and classifies the procrastination habit.

[0143] The method for managing procrastination habits according to the present invention monitors mental health and habits through conversation with a chatbot via mobile or web and provides personalized advice (mhealth / ehealth with chatbot). Various AI technologies such as natural language processing technology or classification models can be incorporated as needed.

[0144] In addition, the method for managing procrastination habits according to the present invention can provide a method for generating dialogue for chatbot interaction by simultaneously using a rule-based chatbot in which the dialogue order is determined based on a natural language processing model necessary for achieving the purpose of the chatbot or the dialogue order and the dialogue progresses according to the rules of each order, a retrieval-based chatbot that uses an AI classification model to output the next sentence of the chatbot from a group of chatbot answers that have been previously written, and a generative chatbot that outputs an answer using an AI generation model according to a user's response.

[0145] Meanwhile, in the step (S100) of improving the procrastination habit according to the present invention, an AI chatbot conversation system capable of natural conversation is provided for the purpose of improving the procrastination habit, thereby providing a cognitive behavioral therapy method used to improve the procrastination habit in existing offline counseling.

[0146] Here, in the procrastination habit improvement stage (S100), the user is conversed with through long and short conversations over a certain period of time.

[0147] Additionally, long conversations sequentially involve getting to know myself, identifying my strengths and weaknesses at work, reflecting on my actions, dealing with my fears, and identifying changes in myself.

[0148] Here, short conversations are sequentially followed by longer conversations about the reason for the progress of daily tasks, customized feedback on plan achievement rate and procrastination habit, developing action strategies and practicing with chatbots, and learning effective time management techniques.

[0149] The step (S100) of improving procrastination habits according to the present invention includes cognitive behavioral therapy sessions for procrastination treatment and some life crafting for motivation, and is largely conducted in 5 sessions, with one session per week, so the chatbot can be used for a total of 5 weeks.

[0150] Here, one session consists of one day of long conversations on a specific topic, and several days of short conversations on the topic, where practice is practiced and feedback is received. The long conversations take about 3 to 7 minutes, and the short conversations take about 10 seconds to 3 minutes. Detailed examples of these are omitted as they are explained in Fig. 2.

[0151] Therefore, the procrastination habit management method according to the present invention has the advantage of performing an AI chatbot conversation that enables natural conversation for the purpose of improving the procrastination habit.

[0152] Meanwhile, the chatbot's response output stage (S200) analyzes the user's response intention for the purpose of encouraging the user to think as in actual counseling when conversing with the cognitive behavioral therapy chatbot for improving procrastination habits, and is described in detail in Fig. 6.

[0153] Meanwhile, the procrastination habit classification step (S300) names six causes of procrastination and identifies and provides the user's procrastination causes through the user's response in a conversation with a chatbot, which is explained in detail in Fig. 7.

[0154]

[0155] Figure 6 is a flowchart showing in detail the response output step (S200) of the chatbot of Figure 5.

[0156] As can be seen in FIG. 6, the chatbot response output step (S200) includes a chatbot question extraction step (S210) for extracting a chatbot question from a scenario DB (121), a user answer reception step (S220) for receiving a user answer from a user terminal (200) for the chatbot question, and a user answer classification step (S230) for classifying the user answer into one of a normal response, 'I don't know', and 'I'm skeptical' in an answer classification model (122).

[0157] The response generation step (S200) of the chatbot according to the present invention classifies the intention so that the user can better think of an answer to a question, as in a consultation.

[0158] Unlike existing chatbots, it can label cases where the user answers strangely to a question: either "I don't know" (labeled "I don't know") or "I don't want to answer" (labeled "I'm skeptical"). Therefore, responses can be categorized into three categories: "I answered correctly" (labeled "Normal Response"), "I don't know" (labeled "I'm skeptical"), and "I'm skeptical."

[0159] Here, the user response classification step (S230) is a normal response processing step (S231) in which a sentence is generated again if the normal response ends with a question in the sentence generation model (123) when the classification is a normal response, a supplementary explanation processing step (S232) in which a chatbot utterance for a supplementary explanation to a question is performed in the supplementary explanation DB (124) and a user response is received from the user terminal (200) when the classification is 'I don't know', a justification explanation processing step (S233) in which a chatbot utterance for a justification explanation to a question is performed in the justification provision DB (125) in the chatbot's response output unit (120) and a user response is received from the user terminal (200) when the classification is 'I'm skeptical', a sentence in which a sentence is generated again in the sentence generation model (123) when the user response received from the user terminal (200) for the chatbot utterance performed in the supplementary explanation processing step (S232) and the justification explanation processing step (S233) ends with a question. Includes a regeneration step (S234).

[0160] Meanwhile, detailed examples of the cases of normal response, 'skeptical', and 'I don't know' during classification are described in detail in Fig. 3, so a description thereof is omitted.

[0161] Therefore, the procrastination habit management method according to the present invention has the advantage of being able to perform intention analysis for the purpose of encouraging the user to think as in actual counseling.

[0162]

[0163] Figure 7 is a flowchart showing in detail the procrastination habit classification step (S300) of Figure 5.

[0164] As can be seen in Fig. 7, the procrastination habit classification step (S300) is composed of a user response preprocessing step (S310) in which the procrastination habit classification unit (130) performs preprocessing on the user response, and a procrastination cause classification step (S320) in which the procrastination cause is classified by providing the procrastination habit probability vector for the procrastination cause together with the result of the preprocessing to the procrastination habit analyzer (131).

[0165] Here, the causes of procrastination can be classified into six categories: mental stress and physical stress that is closely related to mental stress, social relationships that are factors that encourage or stop procrastination, environmental factors that distract attention, perfectionism that makes it difficult to start due to uncertainty about the future and fear of evaluation, lack of motivation that reduces self-control and work rate, and self-control that causes low concentration or boredom.

[0166] Meanwhile, preprocessing involves removing stop words from the user's response to the chatbot's question regarding the cause of procrastination in the conversation, and then performing morphological analysis using the Okt morphological analyzer provided by Konlpy. This is omitted here as it is described in detail in Fig. 4.

[0167] At this point, the probability vectors of user responses to the chatbot's questions regarding multiple procrastination factors are summed, and the largest and second-largest procrastination factors are identified. Because the procrastination factors are not independent, some may overlap.

[0168] Therefore, we can first predict the procrastination factor with the largest value, and if the predicted value is different from the user's expectation, we can predict the user's procrastination factor with the second largest value.

[0169] Meanwhile, examples of procrastination factors are described in detail in Fig. 4, so they are omitted here.

[0170] Therefore, the procrastination habit management method according to the present invention has the advantage of being able to identify the cause of procrastination through the user's response in a conversation with a chatbot.

[0171]

[0172] As described above, the chatbot system and method for managing procrastination habits according to the present invention have the advantage of performing AI chatbot conversations that enable natural conversations for the purpose of improving procrastination habits, and have the advantage of performing intention analysis for the purpose of encouraging the user to help with their thoughts as in actual counseling, and of identifying the cause of procrastination through the user's responses in conversations with the chatbot.

[0173]

[0174] Those skilled in the art will appreciate that the various illustrative logical blocks, modules, processors, means, circuits, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, various forms of programs or design code (referred to herein, for convenience, as software), or a combination of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

[0175] The various embodiments presented herein can be implemented as a method, apparatus, or article of manufacture using standard programming and / or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash memory devices (e.g., EEPROMs, cards, sticks, key drives, etc.). Furthermore, various storage media presented herein include one or more devices and / or other machine-readable media for storing information.

[0176] It should be understood that the specific order or hierarchy of steps in the presented processes is merely an example of exemplary approaches. It should be understood that the specific order or hierarchy of steps in the processes may be rearranged within the scope of the present invention based on design priorities. The appended method claims provide elements of various steps in a sample order, but are not intended to be limited to the specific order or hierarchy presented.

[0177] The description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments disclosed herein, but is to be construed in the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A server including a procrastination habit improvement unit that performs procrastination habit improvement by conversing with a user and a chatbot, a chatbot response output unit that performs response output from the chatbot, and a procrastination habit classification unit that performs procrastination habit classification; and A chatbot system for managing procrastination habits, comprising a user terminal that performs the above user and chatbot interface.

2. In paragraph 1, The above procrastination habit improvement unit is a procrastination habit management chatbot system characterized by conversing with the user through long and short conversations over a certain period of time.

3. In paragraph 2, The above long conversation is a chatbot system for managing procrastination habits that sequentially progresses through getting to know myself, finding out my strengths and weaknesses at work, reflecting on my actions, dealing with my fears, and finding out about my own changes.

4. In paragraph 3, The above short conversation is a chatbot system for managing procrastination habits, characterized by sequentially conducting the above long conversation along with the reasons for the degree of daily task performance, plan achievement rate and customized feedback on procrastination habits, establishing an implementation strategy and practicing with the chatbot, and learning effective time management methods.

5. In paragraph 1, A chatbot system for managing procrastination habits, characterized in that the response output section of the chatbot extracts chatbot questions from a scenario DB within the response output section of the chatbot.

6. In paragraph 5, A chatbot system for managing procrastination habits, characterized in that the response output unit of the chatbot receives a user response from the user terminal in response to the chatbot question.

7. In paragraph 6, A chatbot system for managing procrastination habits, characterized in that the response output unit of the chatbot classifies the user response into one of a normal response, 'I don't know', and 'I am skeptical' in a response classification model within the response output unit of the chatbot.

8. In paragraph 7, A chatbot system for managing procrastination habits, characterized in that the response output unit of the chatbot generates a sentence again if the normal response ends with a question in the sentence generation model within the response output unit of the chatbot when the classification of the answer classification model is a normal response.

9. In paragraph 8, A chatbot system for managing procrastination habits, characterized in that the response output unit of the chatbot performs chatbot utterance for auxiliary explanation of a question in the auxiliary explanation DB of the response output unit of the chatbot when the classification of the answer classification model is 'I don't know' and receives a user answer from the user terminal.

10. In paragraph 9, A chatbot system for managing procrastination habits, characterized in that the response output section of the chatbot performs a chatbot utterance for an explanation of the justification for a question in the justification provision DB within the server when the classification of the answer classification model is 'skeptical' and receives a user response from the user terminal.

11. In paragraph 10, A chatbot system for managing procrastination habits, characterized in that the response output section of the chatbot performs chatbot utterance for the auxiliary explanation DB and the justification provision DB, and re-generates a sentence in the sentence generation model when the user response received from the user terminal ends with a question.

12. In paragraph 11, A chatbot system for managing procrastination habits, characterized in that the procrastination habit classification unit performs preprocessing on the user response on the server.

13. In paragraph 12, The procrastination habit classification unit is a chatbot system for managing procrastination, characterized in that the procrastination habit classification unit provides a procrastination habit probability vector for the procrastination cause together with the result of the preprocessing to the procrastination habit analyzer in the procrastination habit classification unit to classify the cause of procrastination.

14. In paragraph 13, A chatbot system for managing procrastination habits, characterized in that the above causes of procrastination include mental stress and physical stress closely related to the above mental stress.

15. In paragraph 13, A chatbot system for managing procrastination, characterized in that the above procrastination causes include social relationships that are factors that encourage or stop procrastination.

16. In paragraph 13, A chatbot system for managing procrastination, characterized in that the above procrastination causes include environmental factors that distract attention.

17. In paragraph 13, A chatbot system for managing procrastination habits characterized by the above causes of procrastination including perfectionism that makes it difficult to get started due to uncertainty about the future and fear of evaluation.

18. In paragraph 13, A chatbot system for managing procrastination habits, characterized in that the above causes of procrastination include lack of motivation, which reduces self-control and lowers work rate.

19. In paragraph 13, A chatbot system for managing procrastination habits, characterized in that the above procrastination causes include low concentration or self-control that makes one feel bored.

20. A procrastination habit improvement step in which a procrastination habit improvement unit communicates with a user through a chatbot using a user terminal that performs a user-chatbot interface, asks questions for improving procrastination habits, and receives responses from the user; A chatbot response output step for outputting a chatbot response corresponding to the user's response from the chatbot's response output section; and A method for managing procrastination, comprising: a procrastination habit classification step for analyzing the user's response in a procrastination habit classification section and performing procrastination habit classification; 21. In paragraph 20, A method for managing procrastination, characterized in that in the step of improving the procrastination habit, a conversation is held with the user through long and short conversations over a certain period of time.

22. In paragraph 21, The above long conversation is a method of managing procrastination that is characterized by sequentially getting to know myself, finding out my strengths and weaknesses at work, reflecting on my actions, dealing with my fears, and finding out how I can change.

23. In paragraph 22, The above short conversation is a method for managing procrastination habits, characterized by sequentially conducting the above long conversation along with the reasons for the degree of daily task performance, plan achievement rate and customized feedback on procrastination habits, establishing an implementation strategy and practicing with a chatbot, and learning effective time management methods.

24. In paragraph 20, A method for managing procrastination habits, characterized in that the response output step of the chatbot includes a chatbot question extraction step of extracting a chatbot question from a scenario DB in the response output section of the chatbot.

25. In paragraph 24, A method for managing procrastination habits, characterized in that the response output step of the chatbot includes a user response receiving step for receiving a user response from the user terminal in response to the chatbot question.

26. In paragraph 25, A method for managing procrastination, characterized in that the response output step of the chatbot includes a user response classification step that classifies the user response into one of a normal response, 'I don't know', and 'I am skeptical' in a response classification model in the response output section of the chatbot.

27. In paragraph 26, A method for managing procrastination, characterized in that, in the above user response classification step, if the classification is a normal response, a normal response processing step is included for generating a sentence again in a sentence generation model in the response output section of the chatbot if the normal response ends with a question sentence.

28. In paragraph 27, A method for managing procrastination, characterized in that in the above user response classification step, if the classification is 'I don't know', the method comprises an auxiliary explanation processing step of performing a chatbot utterance for an auxiliary explanation for a question in the auxiliary explanation DB in the response output section of the chatbot and receiving a user response from the user terminal.

29. In paragraph 28, A method for managing procrastination, characterized in that in the above user response classification step, if the classification is 'skeptical', a justification explanation processing step is included, in which a chatbot utters a justification explanation for a question in a justification provision DB in the response output section of the chatbot and receives a user response from the user terminal.

30. In paragraph 29, A method for managing procrastination habits, characterized in that it includes a sentence regeneration step for regenerating a sentence in the sentence generation model when a user response received from the user terminal to the chatbot utterance performed in the above auxiliary explanation processing step and the above justification explanation processing step ends with a question.

31. In paragraph 30, A method for managing procrastination, characterized in that the procrastination habit classification step includes a user response preprocessing step in which the server performs preprocessing on the user response.

32. In paragraph 31, A method for managing procrastination, characterized in that in the procrastination habit classification step, the procrastination habit classification step includes a procrastination cause classification step for classifying the cause of procrastination by providing a procrastination habit probability vector for the cause of procrastination together with the result of the preprocessing in the procrastination habit analyzer in the procrastination habit classification unit.

33. In paragraph 32, A method for managing procrastination, characterized in that the above causes of procrastination include mental stress and physical stress closely related to the above mental stress.

34. In paragraph 32, A method for managing procrastination, characterized in that the above procrastination causes include social relationships that are factors that urge or stop procrastination.

35. In paragraph 32, A method for managing procrastination, characterized in that the above causes of procrastination include environmental factors that distract attention.

36. In paragraph 32, A method for managing procrastination, characterized by the above causes of procrastination including perfectionism that makes it difficult to get started due to uncertainty about the future and fear of evaluation.

37. In paragraph 32, A method for managing procrastination, characterized in that the above causes of procrastination include lack of motivation, which reduces self-control and lowers work rate.

38. In paragraph 32, A method for managing procrastination, characterized in that the above causes of procrastination include low concentration or self-control that makes one feel bored.

Images