Customized chatbot response system based on customer DISC types

Abstract

The present invention relates to a customized chatbot response system based on customer personality behavior types that estimates the customer's personality behavior (DISC) type in real time based on multidimensional input data collected from customers, and dynamically determines a response strategy based on the estimation result to provide a chatbot service.

Description

Technology Field

[0001] The present invention relates to a customized chatbot response system based on customer personality and behavioral types, and more specifically, to a customized chatbot response system based on customer personality and behavioral types that estimates the customer's personality and behavioral (DISC) type in real time based on multidimensional input data collected from the customer, and dynamically determines a response strategy based on the estimation result to provide a chatbot service. Background Technology

[0003] Recently, companies have been introducing various automated consultation systems, including chatbots, to improve the efficiency and satisfaction of customer service. These systems respond to repetitive customer inquiries through FAQ-based responses, keyword matching, and scenario-based tree structures, and some even include features that analyze conversation flow to adjust the direction of the response.

[0004] However, existing chatbot systems are structured to apply the same response scenarios uniformly without considering the psychological tendencies, communication styles, or emotional response characteristics of customers. This presents a problem in that it is difficult to respond appropriately using only standardized scenarios, even when response strategies need to be varied depending on the customer's expression style or response speed.

[0005] Furthermore, there is a problem in that there is a lack of functionality to recognize and respond early to emotional changes or signs of churn exhibited by customers during conversations, and existing customer behavior analysis primarily relies on pre-conversation surveys or CRM-based segment classification, thus failing to reflect unstructured responses occurring during real-time conversations.

[0006] Due to these limitations, the precision and satisfaction of customer service have decreased, and there have been limitations in that the effectiveness of introducing chatbots is evaluated as low, especially in the field of counseling services where empathetic communication is important.

[0007] Meanwhile, the aforementioned background technology is technical information that the inventor possessed for the derivation of the present invention or acquired during the process of deriving the present invention, and it cannot necessarily be considered publicly known technology disclosed to the general public prior to the filing of the present invention. The problem to be solved

[0009] One aspect of the present invention provides a customized chatbot response system based on customer personality behavior types that estimates the customer's personality behavior (DISC) type in real time based on multidimensional input data collected from a customer, and dynamically determines a response strategy based on the estimation result to provide a chatbot service.

[0010] The technical problems of the present invention are not limited to those mentioned above, and other unmentioned technical problems will be clearly understood by those skilled in the art from the description below. means of solving the problem

[0012] A customized chatbot response system based on customer personality and behavioral types according to one embodiment of the present invention estimates the customer's personality and behavioral (DISC) type in real time based on multidimensional input data collected from the customer, and dynamically determines a response strategy based on the estimation result to provide a chatbot service.

[0013] The customized chatbot response system based on the above customer personality and behavioral types is,

[0014] A data collection unit that collects multidimensional data including customer initial input, behavior logs, language usage, response speed, and inquiry content;

[0015] DISC type estimation unit that estimates the psychological and behavioral tendencies of customers based on the collected multidimensional data;

[0016] A customer characteristic indicator calculation unit that quantifies at least one of the customer's emotional sensitivity, decision speed, language complexity, churn potential, and information search tendency based on the estimated DISC type above;

[0017] A customer index integration unit that calculates an integrated customer index by applying propensity-based weights to the above customer characteristic indicators; and

[0018] It includes a response strategy determination unit that dynamically determines a customer response strategy based on the above-determined customer characteristic indicators or customer index and provides a customized chatbot response according to the strategy.

[0019] The above customer characteristic indicator calculation unit is,

[0020] Each characteristic indicator, such as customer emotional sensitivity, decision speed, language complexity, churn potential, and information search tendency, is calculated as a normalized numerical value according to a preset standard range or statistical standardization method, and

[0021] It is characterized by dynamically reflecting changes in customer tendencies by updating the values ​​of each characteristic indicator in real time according to the detected changes in the customer's response method, expression content, behavior logs, etc., while a conversation with the customer is taking place in real time.

[0022] The above-mentioned customer index integration department is,

[0023] For each of the multiple characteristic indicators transmitted from the customer characteristic indicator calculation unit, the D-type, I-type, S-type, and C-type propensity weights provided by the DISC type estimation unit are applied to convert the importance of each indicator into a weighted correction value adjusted to the propensity of the corresponding customer, and

[0024] It is characterized by calculating an integrated customer index that can be used as a criterion for selecting a customer response strategy by integrating multiple corrected indicator values.

[0025] The above response strategy decision department,

[0026] Whenever new input data is collected during the chatbot conversation with a customer, the customer characteristic indicator calculation unit recalculates each characteristic indicator value based on the data, and

[0027] It is characterized by the Customer Index Integration Unit updating the integrated customer index, and the Response Strategy Decision Unit re-selecting the most suitable response strategy or conversation scenario for the current situation in real time based on the updated customer index, thereby dynamically determining a customized response type according to changes in customer reactions and tendencies even during a conversation. Effects of the invention

[0029] According to one aspect of the present invention described above, by estimating the psychological behavioral tendencies of customers in real time and calculating characteristic indicators based thereon, it is possible to provide customized customer responses, unlike existing standardized chatbot scenarios.

[0030] In addition, by identifying customers' emotional sensitivity or churn potential early on and providing more intensive information or guidance to those showing signs of leaving during the conversation, the engagement retention rate can be increased and overall consultation satisfaction can be improved.

[0031] In addition, since DISC-based personality analysis and scenario design are automated without the need for separate counseling experts, advanced counseling services can be provided even in service environments centered on small and medium-sized enterprises or non-specialized personnel. Brief explanation of the drawing

[0033] FIG. 1 is a diagram illustrating the schematic configuration of a customized chatbot response system based on customer personality behavior types according to one embodiment of the present invention. Specific details for implementing the invention

[0034] The following detailed description of the invention refers to the accompanying drawings, which illustrate specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. It should be understood that various embodiments of the invention are different but need not be mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in other embodiments without departing from the spirit and scope of the invention in relation to one embodiment. It should also be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the invention. Accordingly, the following detailed description is not intended to be limiting, and the scope of the invention is limited only by the appended claims, including all equivalents to those claimed therein, provided appropriately described. Similar reference numerals in the drawings refer to the same or similar functions across various aspects.

[0035] Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the drawings.

[0036] FIG. 1 is a diagram illustrating the schematic configuration of a customized chatbot response system based on customer personality behavior types according to one embodiment of the present invention.

[0037] The customized chatbot response system based on customer personality behavior type according to the present invention aims to provide a chatbot service by estimating the customer's personality behavior (DISC) type in real time based on multidimensional input data collected from the customer and dynamically determining a response strategy based on the estimation result.

[0040] To this end, the customized chatbot response system based on customer personality and behavioral types according to the present invention includes a data collection unit, a DISC type estimation unit, a customer characteristic indicator calculation unit, a customer index integration unit, and a response strategy determination unit.

[0041] The data collection unit collects multidimensional data from customers, including information such as customer initial input, behavior logs, language usage, response speed, and inquiry content.

[0042] In other words, the data collection unit can collect and store all interaction data between the customer and the chatbot or response system in real time to multidimensionally secure input data for estimating customer propensity.

[0043] In one embodiment, the data collection unit may collect data based on initial information and a preliminary survey entered by a customer when first entering the service or starting a consultation. Additionally, the data collection unit may also collect data such as the original text of input sentences (text data), text input time intervals, multiple-choice responses, and emoticons entered by the customer into the chatbot interface. In this process, the data collection unit may also collect non-verbal behavior logs, such as the customer's mouse clicks, page navigation, scroll depth, and dwell time, so that they can be used to determine the customer's information search tendencies and attention duration.

[0044] The DISC type estimation unit analyzes the customer's multidimensional data transmitted from the data collection unit to analyze the customer's DISC type.

[0045] DISC theory classifies human behavioral types into four categories: Dominance, Influence, Steadiness, and Conscientiousness, each exhibiting distinct differences in verbal expression, decision-making speed, information preferences, and emotional response patterns. Therefore, DISC types are estimated based on customers' language usage, response patterns, and behavioral logs.

[0046] However, to date, technology for quantitatively estimating DISC types in chatbot systems and quantifying and utilizing customer characteristics based on them has not been sufficiently implemented, and in particular, a structure that dynamically calculates DISC-based indices and reflects them in real-time response strategies is difficult to find in existing technologies.

[0047] Accordingly, the present invention can provide customized response scenarios tailored to individual customers by comprehensively analyzing various information such as the customer's initial survey, input language, response speed, behavior logs, and inquiry content to estimate and correct the customer's DISC tendency and designing a response strategy based thereon.

[0048] The DISC type estimation unit calculates the proportion of each of these four types a customer exhibits in the form of continuous values ​​or probability distributions.

[0049] To this end, the DISC type estimator estimates the customer's DISC type based on data such as sentence length, sentiment ratio, proportion of positive / negative words, use of conjunctions, average response time, rate of change in response latency, click frequency, information search patterns, dropout status, scroll depth, and pre-survey response results.

[0050] The DISC type numerical estimator periodically adjusts the initial DISC type estimation results as the conversation progresses. For example, if the DISC type numerical estimator initially classified the customer as Type I based on the initial data collection results, but the customer proceeded with the conversation focusing on lengthy explanations, technical questions, and requests for information later on, it can gradually increase the Type C weight.

[0051] The customer characteristic indicator calculation unit can represent the psychological and behavioral characteristics exhibited by customers as specific quantified indicators based on propensity information transmitted from the DISC type numerical estimation unit and input data from the data collection unit.

[0052] In particular, unlike fixed customer type classification methods, the customer characteristic indicator calculation unit dynamically calculates behavior-based characteristic indicators, thereby enabling the response strategy decision unit described later to perform more precise personalized responses.

[0053] To this end, the Customer Characteristics Indicator Calculation Unit calculates a customer index that quantitatively reflects customer response characteristics, emotional sensitivity, and decision-making styles based on customers' DISC tendencies and multidimensional customer data. To achieve this, it constructs a customer characteristic indicator vector capable of quantifying and expressing individual customers' tendencies or psychological responses. This vector is composed of the following five key indicators by analyzing customers' conversational responses, language usage, and behavioral logs.

[0054] The customer characteristic indicator vector consists of an emotion sensitivity indicator (f1), a decision speed based on response speed (f2), a language complexity indicator (f3), a churn likelihood indicator (f4), and an information search tendency indicator (f5), and can be expressed in the form of such a customer characteristic indicator vector v=[f1, f2, f3, f4, f5].

[0055] The Emotional Sensitivity Score is a metric that quantifies a customer's emotional response sensitivity based on the frequency and intensity of emotional expressions (emotional words, negative words, emojis, etc.) within sentences entered by the customer. The score is calculated based on the frequency of emotional words relative to the total number of words, the ratio of negative words to positive words, and the change in emotional direction, which is the difference between the number of emotional words in the beginning and the end. Since DISC types I and S tend to react sensitively to emotional language and use an emotional tone when there are many emotional expressions, the Emotional Sensitivity Score was included in the customer characteristic indicator vector to estimate these types.

[0056] The Decision Speed ​​Score is a metric that quantifies a customer's decision-making speed based on factors such as the time taken to respond to a question and whether the question is repeated. The score is calculated based on the average of total response times, the standard deviation of response times, and the processing rate of multiple-choice responses. Since DISC types D prefer quick decisions and C prefer careful judgment, the Decision Speed ​​Score can serve as an important indicator to distinguish between Type D and Type C.

[0057] The Language Complexity Score is a metric that quantifies the complexity of a customer's linguistic expression based on the structure, length, and vocabulary diversity of the sentences they use. The score is calculated based on characteristics such as average sentence length, the ratio of unique words to total words, and the number of conjunctions. DISC Type C customers prefer analytical and logical expressions and tend to use complex sentence structures. Accordingly, the Language Complexity Score can serve as an indicator to distinguish Type C.

[0058] The Dropout Risk Score is an indicator of customer dropout risk calculated based on logs of customers abandoning a conversation or having their responses cut off. It is calculated based on characteristics such as the duration of non-response, window closing, the occurrence of session interruption events, and the number of repeated entries for the same question. Since Type D and Type I customers are sensitive to the provision of unnecessary information or delays and tend to drop out quickly, this metric can be used to adjust response strategies.

[0059] The Information-Seeking Score is a metric that quantifies the extent to which customers voluntarily click and explore FAQs, menus, and detailed information. The Information-Seeking Score can be derived based on characteristics such as the number of FAQ menu clicks, the number of search terms entered through the search bar, screen scroll depth, and dwell time. Since Type C and Type S customers make prudent decisions based on sufficient information, information-seeking behavior becomes a key factor in determining their tendencies, and responses containing detailed explanations are required.

[0060] Each of these indicators is normalized to a range of 0 to 100 and subsequently combined into an integrated Customer Index (CPI) by reflecting weights based on DISC tendencies in the Customer Index Integration Unit. Indicators can be updated in real time; for example, since emotional sensitivity may rise or fall as the conversation progresses, the rate of change over time can also be recorded.

[0061] Through these emotional sensitivity indicators (f1), decision speed based on response speed (f2), language complexity indicator (f3), churn likelihood indicator (f4), and information search tendency indicator (f5), it is possible to quantitatively reflect behavioral patterns that are difficult to identify with simple tendency (DISC) information alone, and subsequently support detailed strategic branching when determining response strategies.

[0062] The Customer Index Integration Unit calculates the final customer index by applying propensity-specific weights to each indicator and integrating them, based on multiple customer characteristic indicators quantified and transmitted from the Customer Characteristic Indicator Calculation Unit and customer propensity information provided by the DISC Type Estimation Unit. Rather than simple summation or averaging, the Customer Index Integration Unit enables more detailed customized response strategies by variably reflecting the importance of each indicator according to each customer's propensity characteristics.

[0063] The Customer Index Integration Department receives customer characteristic indicators and DISC propensity distribution information data as input.

[0064] The customer characteristic indicator is a quantified indicator value transmitted from the customer characteristic indicator calculation unit, and the DISC propensity distribution information is the propensity probability distribution of each indicator transmitted from the DISC type numerical estimation unit.

[0065] In other words, the Customer Index Integration Department comprehensively analyzes customers' preliminary surveys, conversation responses, input sentence characteristics, and behavioral logs to derive a quantitative indicator in the form of a probability distribution that shows the degree of each of the four tendencies—Type D (Dominance), Type I (Influence), Type S (Steadiness), and Type C (Conscientiousness)—to which the customer exhibits each.

[0066] In this case, rather than classifying into a single fixed type, the contribution or propensity strength of each of the four types is considered together and expressed in a distribution form such that the total sum is 1. This is a sophisticated approach that takes into account the fact that it is difficult to categorize actual human personality or reactions into a single type.

[0067] To this end, the Customer Index Integration Department first collects basic information related to DISC tendencies through simple self-assessment survey questions when customers first access the chatbot or consultation interface or sign up for the service. For example, questions such as "Do you tend to make quick, intuitive judgments when making decisions? Do you tend to make decisions after gathering sufficient information? Do you tend to express your emotions frequently?" have a high correlation with Type D, Type C, and Type I, respectively, in DISC theory.

[0068] When a conversation begins, various elements such as the tone, length, sentence structure, vocabulary used, response speed, click behavior, and information search patterns of the text entered by the customer are collected. The Customer Index Integration Unit analyzes this data in real time and adjusts the propensity score by applying weights to existing survey-based tendencies based on linguistic and behavioral patterns.

[0069] Once the correction is complete, the scores calculated for each type are organized into a single probabilistic vector. However, since these scores cannot be used directly as probabilities, their relative proportions are calculated based on the total sum of scores, and they are finally normalized into a DISC type probability distribution.

[0070] For example, if the D-type score is 15 points, the I-type score is 38 points, the S-type score is 15 points, and the C-type score is 7 points, the sum of these scores is 75 points, and each score is normalized based on 75 points. That is, the D-type score is calculated as 0.2 (15 / 75), the I-type score as 0.5 (38 / 75), the S-type score as 0.2 (15 / 75), and the C-type score as 0.1 (7 / 75).

[0071] Therefore, the final DISC propensity distribution of customers is [D:0.2, I:0.5, S:0.2, C:0.1], and this result is subsequently used as a reference value in various sub-modules, such as adjusting the weights of customer characteristic indicators, setting priorities for response strategies, and determining chatbot tone and conversation flow.

[0072] The Customer Index Integration Unit performs a final integration calculation after weighting and correcting indicator values ​​using a propensity-based weighting algorithm. That is, since the relative importance of each indicator varies by DISC type, the present invention predefines a weight matrix for each DISC type, weights and corrects indicator values ​​using this matrix, and then performs a final integration calculation.

[0073] The Customer Index Integration Department derives correction values ​​by applying a weighted average based on the DISC propensity distribution value for each indicator.

[0074] In this way, the customer index integration unit receives the changed characteristic indicator values ​​when changes in the customer's response patterns, expression methods, or behaviors occur during a conversation, and recalculates the CPI in real time by reflecting the changed DISC weights. Accordingly, the chatbot according to the present invention can actively modify its response strategy by reflecting changes in the customer's intermediate tendencies, emotional fluctuations, and accumulated fatigue.

[0075] The Response Strategy Decision Department selects and adjusts response strategies to be provided to customers in real time based on the Integrated Customer Index (CPI) received from the Customer Index Integration Department.

[0076] In particular, the response strategy decision unit according to the present invention is characterized by being able to automatically set multifaceted response elements such as sentence length, tone, response volume, conversation speed, whether to provide buttons, and level of information summary by comprehensively considering the customer's psychological tendency, response characteristics, and information search style.

[0077] To this end, the response strategy decision unit receives customer characteristic indicators, DISC type probability distributions, current dominant tendencies, and the topic or context of the customer's inquiry as input data.

[0078] The Response Strategy Decision Unit automatically selects a response strategy to provide to customers based on the Integrated Customer Index (CPI) transmitted from the Customer Index Integration Unit and propensity distribution information transmitted from the DISC Type Numerical Estimation Unit. In this context, a response strategy refers not merely to the content of the response sentence, but encompasses the entire strategic response component, which consists of various elements such as sentence length, tone, method of information presentation, and interface configuration.

[0079] The flow of strategic decision-making is carried out through the following sequential procedures.

[0080] First, the response strategy decision unit analyzes the input Customer Price Index (CPI) value and the customer's DISC propensity distribution (disc_vector). Here, the CPI is an integrated index that comprehensively reflects the customer's current response sensitivity, reaction propensity, and information preference style, while the DISC distribution is the result of quantifying tendencies based on the customer's primary psychological type. Based on these inputs, the response strategy decision unit determines which category among the predefined response strategy templates the case falls into.

[0081] Next, the response strategy decision unit identifies the pool of candidate strategies most suitable for the customer. In this process, the customer's dominant DISC type, key indicators (e.g., emotional sensitivity, decision-making speed), and the purpose of the consultation (e.g., delivery inquiry, refund request) may be considered together to provide a multi-dimensional interpretation of the customer's current psychological state and the suitability of the response.

[0082] Furthermore, to select the optimal strategy from the identified pool of strategy candidates, the response strategy decision unit sets the response components required by each individual strategy. For example, an empathetic tone and short sentences are suitable for customers with high emotional sensitivity, while providing structured information and step-by-step guidance is effective for customers with a strong Type C personality. These response strategies are organized into strategy attributes and transmitted to the chatbot response engine, influencing not only the generation of response messages but also whether buttons are provided, the method of explanation, and whether images are used.

[0083] Finally, the response strategy decision unit transmits strategy IDs or strategy configuration parameters to the chatbot scenario control unit so that the selected strategy is effectively applied, and the actual response is configured based on this. At this time, the configured strategy is synchronized to be maintained consistently in various dimensions, such as not only a single response message but also subsequent response flow, configuration of user options, and control of conversation speed.

[0084] Through this series of flows, the response strategy decision unit of the present invention can automatically select and apply a sophisticated customized response strategy based on the psychological and behavioral characteristics of customers.

[0085] For example, in the case of a customer with a Customer Index of 84.2 points and a DISC type distribution [D: 0.15, I: 0.55, S: 0.20, C: 0.10], the dominant tendency is Type I and the emotional sensitivity indicator is very high; therefore, this customer is expected to be sensitive to emotional expressions and expect an immediate response. Since a relatively definitive profile was derived with a CPI of 80 or higher, the strategic decision-making department communicates with the customer using a high-confidence, empathetic response strategy. This strategy is characterized by delivering only the essentials briefly and concisely using a pleasant and positive tone, while presenting two to three buttons for the customer to choose from (e.g., checking delivery status, changing address, etc.). Accordingly, it becomes possible to respond sensitively to the customer's emotional expectations while inducing a rapid problem-solving flow without customer churn caused by unnecessary information.

[0086] In addition, during this process, the response strategy decision unit takes into account that customers' response tendencies may change even during the conversation; therefore, instead of applying the selected response strategy uniformly as a fixed one, it updates and switches strategies in real time.

[0087] That is, in the process of responding to a customer with a CPI score of 78.4 points and a DISC type distribution of [D: 0.1, I: 0.6, S: 0.2, C: 0.1], if it is confirmed that the length of the customer's input sentences increases, the frequency of using logical conjunctions without repetitive questions increases, questions comparing the conditions of products A and B appear, and search records increase, it is determined that the customer was initially Type I but shifted to a Type C tendency during the conversation. In this case, the Customer Index Integration Department updates the new indicators and confirms that the DISC type distribution changes to [D: 0.05, I: 0.4, S: 0.25, C: 0.3].

[0088] In such cases, the strategic decision department determines that the existing strategy does not align with the customer's situation and responds to the customer using a comparison-based analysis response strategy. This response strategy features detailed product descriptions, active use of comparison tables and comparison images, and minimizes comments that lead to a decision.

[0089] To this end, the Customer Index Integration Unit repeatedly receives new data while the customer's input behavior continues. For example, if the proportion of emotional vocabulary used in sentences entered by the customer gradually decreases, the response time interval becomes longer than initially, or information search behavior increases, this indicates that the customer's tendencies or response patterns are changing.

[0090] When such changes are detected, the Customer Characteristics Indicator Calculation Unit updates each indicator in real time, and the Customer Index Integration Unit recalculates the Integrated Customer Index (CPI) by reflecting these updated indicators. The Strategy Decision Unit refers to this new Customer Index and the updated DISC propensity distribution to determine whether the previously applied strategy remains appropriate.

[0091] If it is determined that the currently applied strategy does not align with the customer's changed state, the response strategy decision unit does not maintain the existing strategy but instead searches for a new strategy template and immediately switches to a suitable strategy. For example, while a friendly tone and summary response strategy were initially applied because Type I personalities—characterized by high emotional sensitivity and quick reactions—were dominant, if the customer repeatedly entered long sentences and multi-layered questions later in the conversation, causing an increase in information-seeking tendencies and language complexity indicators, it is determined that the customer is partially exhibiting Type C personalities, leading to a switch to a logic-centered and structured information provision strategy.

[0092] These strategy updates are flexibly applied to every response without a separate restart in the chatbot response control unit, and customers can continuously receive responses tailored to their preferences naturally without interruption in the flow of conversation.

[0093] This real-time strategy update structure enables adaptive response that reflects changes in customers' psychological reactions, information acceptance, and fatigue, and allows for precise customized responses that existing fixed chatbots cannot provide.

[0094] The technology according to the present invention, as described above, may be implemented in the form of program instructions that can be executed through various computer components or implemented as an application, and may be recorded on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc., either individually or in combination.

[0095] The program instructions recorded on the above-mentioned computer-readable recording medium are those specifically designed and configured for the present invention, but may also be those known and available to those skilled in the art of computer software.

[0096] Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, and flash memory.

[0097] Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The hardware device may be configured to operate as one or more software modules to perform processing according to the present invention, and vice versa.

[0098] Although the invention has been described above with reference to embodiments, those skilled in the art will understand that various modifications and changes can be made to the invention without departing from the spirit and scope of the invention as set forth in the following claims.

Claims

Claim 1 A customized chatbot response system based on customer personality and behavioral types, which estimates a customer's personality and behavioral (DISC) type in real time based on multidimensional input data collected from a customer and dynamically determines a response strategy based on the estimation result to provide a chatbot service, wherein the customized chatbot response system based on customer personality and behavioral types comprises: a data collection unit that collects multidimensional data including a customer's initial input, behavior log, language usage, response speed, and inquiry content; a DISC type estimation unit that estimates the customer's psychological and behavioral tendencies based on the collected multidimensional data; a customer characteristic indicator calculation unit that quantifies at least one characteristic among the customer's emotional sensitivity, decision speed, language complexity, churn potential, and information search tendency based on the estimated DISC type; a customer index integration unit that calculates an integrated customer index by applying a tendency-specific weight to the customer characteristic indicator; and a response strategy determination unit that dynamically determines a customer response strategy based on the calculated customer characteristic indicator or customer index and provides a customized chatbot response according to the strategy. Claim 2 delete Claim 3 A customized chatbot response system based on a customer personality and behavior type, wherein the multidimensional data comprises at least one of: pre-survey response information that the customer responds to when first accessing consultation; text sentences and expressed emotion words that the customer inputs into a conversation interface; the customer's response speed and input time interval to a question; behavior logs such as clicks, scrolls, and FAQ searches performed by the customer within the screen; and events related to whether the customer's session is interrupted or exits. Claim 4 In paragraph 3, the customer characteristic indicator calculation unit calculates each characteristic indicator, such as the customer’s emotional sensitivity, decision speed, language complexity, churn potential, and information search tendency, into a normalized numerical value according to a preset standard range or statistical standardization method, and when a change in the customer’s response method, expression content, behavior log, etc., is detected during a conversation with the customer in real time, the numerical value of each characteristic indicator is updated in real time according to the change to dynamically reflect the change in the customer’s tendency, characterized by a customized chatbot response system based on customer personality and behavioral types. Claim 5 In claim 4, the customer index integration unit applies D-type, I-type, S-type, and C-type propensity weights provided by the DISC type estimation unit to each of the plurality of characteristic indicators transmitted from the customer characteristic indicator calculation unit, converts the importance of each indicator into a weighted correction value adjusted to match the propensity of the corresponding customer, and integrates the corrected plurality of indicator values ​​to calculate an integrated customer index that can be used as a criterion for selecting a customer response strategy, thereby providing a customized chatbot response system based on customer personality and behavior types. Claim 6 In claim 5, the response strategy determination unit is characterized by recalculating each characteristic indicator value in the customer characteristic indicator calculation unit based on the data whenever new input data is collected during the chatbot conversation process with the customer, updating the integrated customer index in the customer index integration unit, and, based on the updated customer index, re-selecting in real-time the response strategy or conversation scenario most suitable for the current situation, thereby dynamically determining a customized response type according to changes in the customer's reaction and tendencies even during the conversation, thereby forming a customized chatbot response system based on customer personality and behavior types.

Images