Method and System for Analyzing Family Emotion Networks Based on Emotional Artificial Intelligence

Abstract

An emotional artificial organ-based user emotion storage and crisis detection protection system acquires emotion data including at least one of text data, voice data, and facial expression data input by a user, processes the emotion data through an emotion analysis model to convert it into emotion state and intensity values, arranges the emotion state and intensity values ​​along a time axis to determine a time-series-based emotion rhythm, stores the emotion rhythm in a local storage embedded in an electronic device inside or outside the user's body, collects bio-signals in real time through a bio-sensor, integrates and analyzes the bio-signals with the emotion data, calculates the rate of change of the bio-signals to derive a correlation with the emotion rhythm, determines a crisis situation if the emotion rhythm falls below a specified lower limit value or the rate of change of the bio-signals exceeds a specified upper limit value, selectively executes at least one of generating a message over voice, transmitting a vibration pattern, or displaying a message through a display according to the type of the determined crisis situation, and can synchronize the bio-signals and the emotion rhythm in real time.

Description

Technology Field

[0001] The present invention relates to the fields of artificial intelligence, wearable computing, and digital healthcare. More specifically, it relates to an emotional artificial organ system and a method of operating the same, which continuously collects changes in a user's emotions and stores them in a form similar to a biological organ, fuses the stored emotional rhythms with real-time biological signals to detect a psychological crisis situation of the user, and generates and provides an automated protective response thereto.

[0002] In addition, the present invention can be applied to the field of technology for constructing emotional networks that convert collected emotional data into medical information and support emotional connections with designated targets, such as family members. Background Technology

[0004] With the advancement of artificial intelligence technology and the widespread adoption of wearable devices, research aimed at recognizing and analyzing user emotions in real time is actively underway. Existing technologies have primarily focused on identifying momentary emotional states through text, voice, or biometric signals, and using this information to improve user experience or provide feedback on specific services.

[0005] However, this approach has tended to treat emotions as one-off or short-term data. Human emotions possess rhythmic characteristics that change and accumulate over time, profoundly impacting an individual's psychological stability and health status. Existing technologies have shown limitations in systematically storing and managing this long-term accumulation of emotions and rhythmic changes, and in predicting or preemptively responding to psychological crisis situations based on this data.

[0006] In addition, while biosignals such as heart rate (BPM) and gross skin response (GSR) collected from wearable devices are closely related to emotional states, there have been insufficient attempts to integrate them with long-term emotional rhythm data to build in-depth crisis detection and protection systems. There is a growing need for technology that goes beyond simply recognizing emotions when a user faces psychological difficulties to immediately provide practical emotional support or protective measures.

[0007] Consequently, there is a growing demand for an integrated emotional artificial organ system and method that continuously stores and manages emotions like biological organs, combines them with biosignals to precisely detect crisis situations, and further provides automated protective responses to promote the psychological stability and safety of users.

[0008] delete Prior art literature

[65535] Korean Registered Patent No. 10-2662412 Korean Registered Patent No. 10-2757044 The problem to be solved

[0009] The present invention aims to provide a system and method for continuously collecting user emotional input, converting it into time-series-based emotional waveforms and tension scores to generate an emotional rhythm, and storing this rhythm long-term like a biological organ in a local storage within the body or a wearable device.

[0010] In addition, the present invention aims to provide a precise crisis detection algorithm and system that detects a user's emotional crisis situation (e.g., sudden emotional decline, fear, anxiety, etc.) by comprehensively analyzing the rate of change of stored emotional rhythms, specific keywords, and linked biosensor (BPM, GSR, etc.) data.

[0011] The present invention aims to provide a system and method for automatically generating and performing an appropriate protective response tailored to the user's state and situation in response to a detected crisis situation. This may include silent text messages, voice-based comfort (TTS), haptic vibration notifications, or a function to notify a designated guardian of the crisis situation.

[0012] The present invention aims to provide a system and method that systematically records a user's emotional rhythm, crisis detection records, protective response logs, etc., and converts them into a standardized medical data format as needed to support their use in individual mental health management or medical counseling. means of solving the problem

[0014] The user emotion storage and crisis detection protection system based on an emotion artificial organ includes a memory and a processor for storing instructions, and when the instructions are executed by the processor, the system continuously collects the user's emotion input, converts it into a time-series-based emotion waveform and tension score, generates an emotion rhythm, stores the generated emotion rhythm long-term like a biological organ in a local storage embedded in the body or a wearable device, detects the user's emotion tension and danger signals in real time by combining the rate of change of the stored emotion rhythm, keyword analysis, and physiological signal data, automatically generates and executes a protection response including a silent comfort message, screen guidance, and vibration notification in response to the detected crisis situation, and controls the system to convert the emotion state record into a medical data format so that it can be linked with a healthcare system. Effects of the invention

[0016] The method and system for storing user emotions and detecting crises based on an emotional artificial organ according to the present invention has the effect of enabling in-depth understanding and tracking of an individual's psychological state by storing and managing the user's emotions as long-term rhythms that accumulate and change over time, rather than as short-term information.

[0017] In addition, the present invention has the effect of contributing to the protection of the user's mental health and the enhancement of safety by enabling an immediate response through a crisis detection algorithm that fuses stored emotional rhythms and real-time bio-signals, by preemptively or in real-time detecting psychological crisis situations that the user may not recognize or find difficult to express.

[0018] The present invention provides various forms of automated protective responses, such as voice, text, and vibration, to detected crisis situations, thereby providing immediate emotional support and a sense of security to the user, and, if necessary, inducing intervention by a guardian or expert to prevent or mitigate dangerous situations.

[0019] The present invention provides emotion monitoring and protection functions in a continuous and non-invasive manner within the user's daily life by implementing an emotion artificial organ system in a wearable device or an implantable device, and has the effect of converting collected data into assets that can be utilized in the medical and healthcare fields.

[0020] This invention has the effect of presenting a new paradigm for future emotion-based healthcare and protection devices by linking emotional data with family networks to strengthen emotional bonds and manage well-being indicators, thereby supporting human relationships through technology and building a social emotional safety net. Brief explanation of the drawing

[0022] FIG. 1 is a diagram illustrating the overall structure of an artificial intelligence-based system according to one embodiment. FIG. 2 is a diagram illustrating the learning of a neural network according to one embodiment. FIG. 3 is a diagram illustrating the configuration of an artificial intelligence model according to one embodiment. FIG. 4 is a block diagram showing the configuration of a user emotion storage and crisis detection system based on an emotion artificial organ according to one embodiment. FIG. 5 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment. FIG. 6 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment. FIG. 7 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment. Specific details for implementing the invention

[0023] Hereinafter, embodiments are described in detail with reference to the attached drawings. However, various modifications may be made to the embodiments, and thus the scope of the patent application is not limited or restricted by these embodiments. It should be understood that all modifications, equivalents, and substitutions to the embodiments are included within the scope of the rights.

[0024] Specific structural or functional descriptions of the embodiments are disclosed for illustrative purposes only and may be modified and implemented in various forms. Accordingly, the embodiments are not limited to the specific disclosed forms, and the scope of this specification includes modifications, equivalents, or substitutions that fall within the technical concept.

[0025] Terms such as "first" or "second" may be used to describe various components, but these terms should be interpreted solely for the purpose of distinguishing one component from another. For example, the first component may be named the second component, and similarly, the second component may be named the first component.

[0026] When it is stated that a component is "connected" to another component, it should be understood that it may be directly connected to or joined to that other component, or that there may be other components in between.

[0027] The terms used in the embodiments are for illustrative purposes only and should not be interpreted as intended to be limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0028] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the embodiments pertain. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0029] In addition, when describing with reference to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the embodiments, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the embodiments, such detailed description is omitted.

[0030] The embodiments can be implemented in various forms of products such as personal computers, laptop computers, tablet computers, smartphones, televisions, smart home appliances, intelligent automobiles, kiosks, and wearable devices.

[0031] Artificial Intelligence (AI) systems are computer systems that implement human-level intelligence; unlike existing rule-based smart systems, they are systems in which machines learn and make decisions autonomously. As AI systems improve in recognition accuracy and gain a more accurate understanding of user preferences with continued use, existing rule-based smart systems are gradually being replaced by deep learning-based AI systems.

[0032] Artificial intelligence technology consists of machine learning and component technologies utilizing machine learning. Machine learning is an algorithmic technology that autonomously classifies and learns the characteristics of input data, while component technologies are technologies that mimic the cognitive and judgmental functions of the human brain by utilizing machine learning algorithms such as deep learning, and are comprised of technological fields such as linguistic understanding, visual understanding, reasoning / prediction, knowledge representation, and motion control.

[0033] The various fields where artificial intelligence technology is applied are as follows. Linguistic understanding refers to technologies that recognize, apply, and process human language and text, including natural language processing, machine translation, dialogue systems, question answering, and speech recognition / synthesis. Visual understanding refers to technologies that perceive and process objects like human vision, including object recognition, object tracking, image search, people recognition, scene understanding, spatial understanding, and image enhancement. Inference and prediction refers to technologies that logically reason and predict by judging information, including knowledge / probability-based inference, optimization prediction, preference-based planning, and recommendation. Knowledge representation refers to technologies that automatically process human experiential information into knowledge data, including knowledge construction (data generation / classification) and knowledge management (data utilization). Motion control refers to technologies that control the autonomous driving of vehicles and the movement of robots, including motion control (navigation, collision, driving) and manipulation control (behavior control).

[0034] Generally, to apply machine learning algorithms to real-world situations, training is performed using a trial-and-error method due to the inherent characteristics of the fundamental methodologies. In particular, deep learning requires hundreds of thousands of iterations. Since it is impossible to execute this in a real physical external environment, training is instead performed through simulations that virtually recreate the actual physical environment on a computer.

[0035] In the present invention, Artificial Intelligence (AI) refers to a technology that imitates human learning ability, reasoning ability, and perceptual ability, and implements them on a computer, and may include concepts such as machine learning and symbolic logic. Machine Learning (ML) is an algorithmic technology that classifies or learns the characteristics of input data on its own. AI technology can analyze input data as a machine learning algorithm, learn from the results of the analysis, and make judgments or predictions based on the results of the learning. Furthermore, technologies that mimic the functions of the human brain, such as cognition and judgment, by utilizing machine learning algorithms can also be understood as falling within the category of AI. For example, technological fields such as linguistic understanding, visual understanding, reasoning / prediction, knowledge representation, and motion control may be included.

[0036] Machine learning can refer to the process of training neural network models using experience in processing data. Through machine learning, computer software can improve its own data processing capabilities. A neural network model is constructed by modeling the correlations between data, and these correlations can be expressed by multiple parameters. A neural network model extracts and analyzes features from given data to derive correlations between them; machine learning can be defined as the process of optimizing the model's parameters by repeating this process. For example, a neural network model can learn the mapping (correlation) between inputs and outputs for data given as input-output pairs. Alternatively, even when only input data is provided, a neural network model can derive regularities between the given data and learn those relationships.

[0037] An artificial intelligence learning model or neural network model can be designed to implement the structure of the human brain on a computer and may include multiple network nodes that have weights and simulate neurons of a human neural network. The multiple network nodes may have interconnected relationships by simulating the synaptic activity of neurons, where neurons exchange signals through synapses. In an artificial intelligence learning model, multiple network nodes may be located in layers of different depths and exchange data according to convolutional connections. The artificial intelligence learning model may be, for example, an Artificial Neural Network (ANN) or a Convolutional Neural Network (CNN). As an embodiment, the artificial intelligence learning model may be machine learned according to methods such as supervised learning, unsupervised learning, and reinforcement learning. Machine learning algorithms for performing machine learning may include Decision Tree, Bayesian Network, Support Vector Machine, Artificial Neural Network, Ada-boost, Perceptron, Genetic Programming, and Clustering.

[0038] Among these, CNNs are a type of multilayer perceptron designed to use minimal preprocessing. CNNs consist of one or more convolutional layers and standard artificial neural network layers stacked on top, additionally utilizing weights and pooling layers. Thanks to this structure, CNNs can fully utilize two-dimensional input data. Compared to other deep learning architectures, CNNs demonstrate good performance in both image and audio fields. CNNs can also be trained using standard backpropagation. CNNs have the advantage of being easier to train than other feedforward artificial neural network techniques and using a small number of parameters.

[0039] Convolutional networks are neural networks comprising sets of nodes with bounded parameters. Many computer vision tasks have been significantly improved, driven by the increased size of available training data and the availability of computational power, combined with algorithmic advancements such as discriminative linear units and dropout training. In the case of massive datasets, such as those available for many tasks today, outfitting is not critical, and increasing the network size improves test accuracy. Optimal utilization of computing resources becomes a limiting factor. To address this, distributed, scalable implementations of deep neural networks can be employed.

[0041] FIG. 1 is a diagram illustrating the overall structure of an artificial intelligence-based system according to one embodiment.

[0042] As illustrated in FIG. 1, an artificial intelligence-based system (100) may include a plurality of user terminals (110-1, 110-n), a server (120), and a database (130). This system adopts a distributed architecture and operates based on a client-server model, with each component playing a unique role and contributing to increasing the efficiency and scalability of the entire system. According to one embodiment, the database (130) is depicted as being configured separately from the server (120), but is not limited thereto; depending on system design and operational efficiency, the database (130) may be integrated within the server (120). This integrated configuration has the advantage of improving data access speed and reducing system complexity. For example, the server (120) may include a plurality of artificial intelligence models and processing units for performing machine learning algorithms, and these can implement various types of deep learning and machine learning technologies (e.g., CNN, RNN, Transformer, reinforcement learning, etc.) to respond to user requests and provide intelligent services. According to one embodiment, a plurality of user terminals (110-1, 110-n), a server (120), and a database (130) can be connected to communicate with each other through a network (N), which enables real-time data exchange and smooth interaction between system components.

[0043] According to one embodiment, the network (N) may perform wireless or wired communication between a plurality of user terminals (110-1, 110-n), a server (120), a database (130), etc., and is located in the center of FIG. 1 and serves as a hub connecting all components. For example, the network may perform wireless communication according to methods such as 5G, LTE (Long-Term Evolution), LTE-A (LTE Advanced), CDMA (Code Division Multiple Access), WCDMA (Wideband CDMA), WiBro (Wireless BroadBand), WiFi (Wireless Fidelity), Bluetooth, NFC (Near Field Communication), GPS (Global Positioning System), or GNSS (Global Navigation Satellite System). The 5G network supports high-speed data transmission (up to 20Gbps), ultra-low latency (1ms or less), and large-scale connectivity, making it suitable for real-time processing of large-capacity AI models. For example, the network (N) may be configured to perform wired communication using methods such as USB (Universal Serial Bus), HDMI (High Definition Multimedia Interface), RS-232 (Recommended Standard 232), Ethernet, fiber optic cable, or POTS (Plain Old Telephone Service). In particular, in data center environments, high-performance network technologies such as 400Gbps Ethernet or InfiniBand can be utilized for communication between large-scale AI computing clusters.

[0044] According to one embodiment, user terminals (110-1, 110-n) are various client devices that access the system and can be implemented in various forms such as smartphones, tablets, desktop computers, wearable devices, and IoT devices. These terminals are responsible for transmitting service requests through a user interface and receiving and displaying results processed by a server (120). Each terminal may also run a lightweight AI model locally, which can reduce network latency and enhance privacy protection. Direct connection or peer-to-peer (P2P) communication between terminals may be possible, as indicated by the dotted line, which is particularly useful in distributed learning or edge computing scenarios.

[0045] According to one embodiment, the server (120) is the central processing unit of the system, responsible for processing user requests, executing artificial intelligence models, and managing data. The server is equipped with a high-performance processor (CPU, GPU, TPU, NPU, etc.) for large-scale computation, large-capacity memory, and a stable operating system, and may be operated in a virtualized manner on a cloud infrastructure. The server performs training and inference functions for complex deep learning models and transmits the results to user terminals or stores them in a database. In addition, the stability and reliability of the system can be ensured through functions such as load balancing, error recovery, and security management.

[0046] According to one embodiment, a database (130) is a storage facility for storing and managing various data, and is depicted in a cylindrical shape in FIG. 1. Data stored in the database (130) is data acquired, processed, or used by at least one component of a plurality of user terminals (110-1, 110-n) or a server (120), and may include software (e.g., programs), user profiles, training datasets, model weights, log records, etc. Structured relational data may be stored in a SQL-based database (MySQL, PostgreSQL, etc.), and unstructured large-scale data may be stored in a NoSQL database (MongoDB, Cassandra, etc.) or a distributed file system (Hadoop HDFS, etc.). The database (130) may include volatile and / or non-volatile memory, where volatile memory (RAM) is used for caching or temporary data processing requiring fast data access, and non-volatile memory (SSD, HDD, tape drive, etc.) is used for permanent data storage. Databases also ensure the security and integrity of data through data encryption, access control, and backup and recovery mechanisms.

[0047] This artificial intelligence-based system (100) can be utilized in various application fields and can provide services such as user behavior analysis, recommendation systems, natural language processing, image and video recognition, and predictive analysis. In particular, the distributed architecture and scalable design of the system enable stable performance to be maintained even as the number of users and data volume increase.

[0049] FIG. 2 is a diagram illustrating the learning of a neural network according to one embodiment.

[0050] As illustrated in FIG. 2, the learning device can train a neural network (123) to classify review responses received from multiple user terminals (110-1, ���) by item. Additionally, the learning device can train a neural network (123) to extract user stay history from user movement path information. The neural network (123) is also called an artificial neural network and is a computational model designed inspired by the neural structure of the human brain, specialized in recognizing and learning complex patterns within data. According to one embodiment, the learning device may be a separate entity from the server (120), but it may also be implemented integrated into the same system, so it is not limited thereto.

[0051] According to one embodiment, the neural network (123) includes an input layer (121) into which training samples are input and an output layer (125) that outputs training outputs, and can be learned based on the difference between the training outputs and labels (i.e., actual correct data). Here, the labels are defined based on items corresponding to review responses (e.g., service quality, price satisfaction, cleanliness, etc.) and can be defined based on user dwell history (place visited, time spent, frequency of visit, etc.) corresponding to movement path information. The neural network (123) is connected as a group of multiple nodes and is defined by weights between the connected nodes and an activation function that activates the nodes. Various forms of the activation function may be used, such as sigmoid, hyperbolic tangent (tanh), and ReLU, which enables the network to learn non-linear patterns.

[0052] According to one embodiment, the learning device can train a neural network (123) using a Gradient Descent (GD) technique or a Stochastic Gradient Descent (SGD) technique. The GD technique is a method of updating weights all at once by calculating gradients based on the entire dataset, while the SGD technique is a method of increasing computational efficiency and the possibility of escaping a local optimum by updating weights more frequently using only a randomly selected portion of data (mini-batch). The learning device can use a loss function designed by the outputs and labels of the neural network. The loss function is an important element that quantifies the difference between the model's predicted value and the actual correct answer to suggest the direction of learning.

[0053] The learning device can calculate the training error using a predefined loss function. The loss function can be predefined with labels, outputs, and parameters as input variables, where the parameters can be set by weights within the neural network (123). For example, the loss function can be designed in the form of Mean Square Error (MSE), entropy, etc. MSE is calculated as the squared mean of the difference between the predicted value and the actual value and is mainly used for regression problems, while cross-entropy is a function that measures the difference between the predicted probability distribution and the actual distribution and is suitable for classification problems. In addition, loss functions that are less sensitive to outliers, such as Huber Loss, can be utilized, and various techniques or methods can be employed in the embodiments where the loss function is designed.

[0054] According to one embodiment, the learning device can identify weights that influence the training error using a backpropagation technique. Backpropagation is a process of calculating the degree to which each weight contributes to the final error while propagating the error calculated in the output layer toward the input layer, and is performed through differential calculations using the chain rule. Here, the weights are relationships between nodes within the neural network (123). The learning device may use an SGD technique using labels and outputs to optimize the weights found through the backpropagation technique. For example, the learning device may update the weights of a loss function defined based on labels, outputs, and weights using an SGD technique. This process operates by calculating the gradient (the derivative of the loss function with respect to the weights) and adjusting the weights in the direction of this gradient to gradually decrease the value of the loss function.

[0055] According to one embodiment, the learning device extracts first objects from a review response, obtains first labels which are items corresponding to the first objects, applies the first objects to a first neural network to generate first training outputs corresponding to the first objects, and can train the first neural network based on the first training outputs and the first labels. At this time, the first objects may include key keywords, sentence structures, sentiment expressions, etc. extracted from the review text, and thereby learn the ability to identify the core content of the review and classify it into the corresponding items.

[0056] According to one embodiment, the learning device extracts second objects from movement path information, obtains second labels which are user dwell history corresponding to the second objects, applies the second objects to a second neural network to generate second training outputs corresponding to the second objects, and can train the second neural network based on the second training outputs and the second labels. The second objects may consist of data such as the user's GPS coordinate sequence, movement speed, stopping point, and movement pattern, thereby learning the ability to determine where and for how long the user stayed.

[0057] According to one embodiment, a learning device can generate first training feature vectors based on constituent features (e.g., sentence structure, word frequency, part-of-speech distribution), positional features (e.g., location of important words, location of key sentences), and pattern features (e.g., repetitive expressions, usage patterns of specific phrases) of a review response. These feature vectors are generated through a process of converting raw text data into a numerical form that can be processed by a neural network, and various methods such as TF-IDF, Word2Vec, and BERT may be employed to extract features.

[0058] According to one embodiment, the learning device can generate second training feature vectors based on constituent features of movement path information (e.g., shape of the path, estimation of the means of movement), length features (e.g., total travel distance, travel time, distance between each point), and pattern features (e.g., repeated visit patterns, movement patterns by day of the week / time of the day). Since location data has continuous values ​​over time, time-series data processing techniques or spatial data analysis methodologies may be applied, and various methods may be employed to extract features.

[0059] According to one embodiment, the learning device can obtain training outputs by applying first training feature vectors to the neural network (123). This process is also called feedforward and is a process in which input data passes through each layer of the network to generate a final prediction value. The learning device can train the review item extraction algorithm of the neural network (123) based on the training outputs and first labels. The learning device can train the review item extraction algorithm of the neural network (123) by calculating training errors corresponding to the training outputs and optimizing the connection relationships of nodes within the neural network (123) to minimize the training errors. This process is repeated over several epochs, and in each epoch, the entire training dataset is processed and weights are updated. The server (120) can automatically extract and classify items from new review responses using the first neural network that has been trained.

[0060] According to one embodiment, the learning device can obtain training outputs by applying second training feature vectors to the neural network (123). The learning device can train the user stay history acquisition algorithm of the neural network (123) based on the training outputs and second labels. The learning device can train the user stay history acquisition algorithm of the neural network (123) by calculating training errors corresponding to the training outputs and optimizing the connection relationships of nodes within the neural network (123) to minimize the training errors. Since movement path data has temporal continuity, a recurrent neural network (RNN) structure such as LSTM (Long Short-Term Memory) or GRU (Gated Recurrent Unit) can be applied, which enables effective learning of the patterns of the sequence data. The server (120) can obtain user stay history from movement path information using the second neural network after training is complete, and the information thus obtained can be utilized in various application fields such as improving location-based services, building customized recommendation systems, and analyzing traffic patterns.

[0062] FIG. 3 is a diagram illustrating the configuration of an artificial intelligence model according to one embodiment.

[0063] An artificial intelligence model according to one embodiment may include an input layer, a hidden layer, and an output layer. This multi-layer structure is effective for learning complex patterns and solving non-linear problems, and is utilized as a core structure, particularly in deep learning models. The input layer is a layer related to input values ​​input into the artificial intelligence model and serves to receive data coming from the outside. As illustrated in FIG. 3, the input layer consists of three nodes, and each node may represent a specific feature or attribute of the data. The number of nodes in the input layer may vary depending on the dimension or characteristics of the data to be processed, and may be determined by the number of pixels in the case of image data, the dimension of the word vector in the case of text data, etc.

[0064] According to one embodiment, a feature map can be output by performing a MAC (multiply-accumulate) operation and an activation operation on the input values ​​in the hidden layer. In FIG. 3, the hidden layer consists of four nodes, and these nodes are connected to all nodes of the input layer to form a fully connected structure. The MAC operation may be an operation that multiplies the input values ​​by their corresponding weights and sums the multiplied values, and this acts as the basic unit of information processing in a neural network.

[0065] According to one embodiment, the activation operation may be an operation that inputs the result of a MAC operation into an activation function and outputs a result value. The activation function is a key element that enables a model to learn complex patterns by introducing non-linearity into the result of a linear operation. The activation function may be of various types. For example, the activation function may include a sigmoid function (which outputs a value between 0 and 1 and is useful for expressing probability), a tangent function (which outputs a value between -1 and 1 and is suitable for balanced data), a ReLU function (which outputs 0 for negative inputs and outputs it as is for positive inputs, mitigating the vanishing gradient problem), a Leaky ReLU function (which applies a linear function with a small gradient to negative inputs), a MaxOut function (which selects the maximum value among several linear functions), and / or an ELU function (which is an exponential linear unit function that provides a smooth saturation curve for negative inputs), but there are no limitations on the types thereof. Recently, new activation functions such as Swish and Mish are also being studied and are contributing to the improvement of model performance.

[0066] According to one embodiment, the hidden layer may be composed of at least one layer, and the more layers there are, the more complex patterns can be learned, resulting in a deep learning structure. For example, when the hidden layer is composed of a first hidden layer and a second hidden layer, the first hidden layer performs MAC operations and activation operations based on the input value of the input layer to output a feature map, and the feature map, which is the result value from the first hidden layer, can become the input value for the second hidden layer. The feature map is an intermediate representation representing features extracted from input data; in image processing, it can represent visual features such as edges, textures, and shapes, and in text processing, it can represent semantic patterns or syntactic structures. The second hidden layer can perform MAC operations and activation operations based on the feature map, which is the result value of the first hidden layer, and through this hierarchical processing, increasingly abstract and high-dimensional features can be extracted.

[0067] According to one embodiment, the output layer may be a layer associated with the result of an operation performed in the hidden layer, and in FIG. 3, it consists of three nodes. The number of nodes in the output layer varies depending on the type of problem to be solved; for binary classification problems, it may be one node, for multi-class classification problems, it may be equal to the number of classes, for regression problems, it may be determined by the dimension of the value to be predicted, and for sequence generation problems, it may be determined by the length of the output sequence. In the output layer, a special activation function such as a softmax function or a sigmoid function is generally used to generate a final predicted value or a probability distribution.

[0068] In one embodiment, the learning model learns syllable (character) patterns that are frequently combined and used in a given corpus to automatically learn the boundaries of compound words and named entities, integrates object information from a first UI source with object information rendered in a browser to create a learning object information file, and uses the learning object information file to generate training data for training a deep learning network. In this process, natural language processing (NLP) technology is utilized, and architectures such as word embedding, recurrent neural networks (RNN), and Transformers may be applied. Additionally, the learning model receives data from various domains of a support system, standardizes the data from the various domains into an integrated format based on at least one standardization method corresponding to each of the various domains, learns and infers data from a specific domain, determines information to be transmitted for standardization from the data of the specific domain, and performs post-processing on the data from the various domains.

[0069] According to one embodiment, the first UI source includes an XML file, and the training object information file includes an input JSON file for feature learning and an output JSON file that serves as label data during training. This structured data format allows the model to process it easily and is particularly useful for analyzing UI elements of web-based applications. The output JSON file includes a file containing HTML DOM Tree information implemented in compliance with web standards, and the various domains include at least one of a RAN (radio access network), a transport, or a core. A RAN refers to a wireless access network where a mobile device connects to a cellular network; a transport refers to a network layer responsible for data transmission; and a core refers to the central part of a network responsible for central processing and routing. Post-processing may include a correlation function, which plays an important role in analyzing relationships between data collected from various domains and discovering meaningful patterns. Through correlation analysis, the model can identify the causes of anomalies or performance degradation occurring in complex systems and improve prediction accuracy.

[0072] *71 FIG. 4 is a block diagram showing the configuration of a user emotion storage and crisis detection system based on an emotion artificial organ according to one embodiment.

[0073] A system (400) according to one embodiment may include a processor (420) and memory (430), and some of the illustrated components may be omitted or substituted. Although the system diagram briefly depicts only the core components, the actual implementation may include various additional components such as an input / output controller, a system bus, a graphics processing unit (GPU), a network interface card (NIC), and a storage device controller. A system (400) according to one embodiment may be a server or a terminal; if implemented as a server, it may be operated in a virtualized environment as part of a cloud infrastructure or implemented as a physical hardware server, and if implemented as a terminal, it may be implemented in various forms such as a desktop computer, laptop, tablet, smartphone, embedded system, or IoT device. According to one embodiment, the processor (420) is a component capable of performing operations or data processing regarding the control and / or communication of each component of the system (400), and may be composed of one or more processors. In modern processor architectures, multi-core designs are common, which improve parallel processing performance by integrating multiple independent processor cores within a single chip. For example, configurations such as dual-core, quad-core, and octa-core are possible, and each core has independent cache memory (L1, L2), and some cache (L3) can be shared among cores.

[0074] The memory (430) can store information related to the method described above or a program in which the method described above is implemented. The memory subsystem is designed with a hierarchical structure and is composed of several levels according to access speed and capacity. The memory (430) may be volatile memory or non-volatile memory. Volatile memory is memory in which stored data is lost when the power is turned off, and is mainly implemented in the form of DRAM (Dynamic RAM) or SRAM (Static RAM). DRAM stores each bit in the form of charge in a capacitor and requires periodic refreshing, whereas SRAM uses flip-flop circuits to maintain data, so it does not require refreshing but has higher cost and power consumption due to the use of more transistors. The memory (430) can store various file data, and the stored file data can be updated according to the operation of the processor (420). The memory management unit (MMU) is responsible for converting virtual memory addresses to physical memory addresses, managing page tables, and maintaining memory protection and cache consistency.

[0075] According to one embodiment, the processor (420) can execute a program and control the device (400). The processor can operate through basic pipeline stages of instruction fetch, decode, execute, and write-back. The code of the program executed by the processor (420) can be stored in memory (430). In addition to the application, the operating system, device driver, middleware, system services, etc. are loaded into memory and executed, and these provide basic system functions such as hardware resource management, process scheduling, interrupt handling, and file system management. The operations of the processor (420) can be performed by loading instructions stored in memory (430). In this process, the program counter (PC) points to the address of the next instruction to be executed, the instruction register (IR) stores the instruction currently being executed, and various general-purpose registers store operation data and intermediate results. The system (400) can be connected to an external device (e.g., a personal computer or a network) through an input / output device (not shown in the drawing) and exchange data. The input / output system can operate on an interrupt-based or polling-based basis and can transfer data directly between memory and the input / output device without the intervention of a processor through a Direct Memory Access (DMA) controller.

[0076] According to one embodiment, data exchange between memory (430) and processor (420) is performed via a system bus, which consists of an address bus, a data bus, and a control bus. The address bus specifies the memory location that the processor intends to access, the data bus transmits actual data, and the control bus transmits control information such as read / write signals. The system (400) may be operated under an operating system that supports advanced features such as multitasking, virtual memory, memory protection, and separation of privileges, and if a real-time operating system (RTOS) is used, it may satisfy deterministic response times and strict time constraints. For artificial intelligence applications, the memory (430) stores a neural network model, weight parameters, a training dataset, intermediate calculation results, etc., and the processor (420) may perform inference and learning by executing forward propagation and backpropagation algorithms. The system (400) can ensure high availability and fault tolerance by applying technologies such as clustering, load balancing, and failover mechanisms for scalability, and can implement technologies such as data encryption, access control, secure booting, and memory protection for security.

[0078] FIG. 5 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment.

[0079] The operations described through FIG. 5 can be implemented based on instructions that can be stored in a computer recording medium or memory (e.g., memory (430) of FIG. 4). The order of each operation of FIG. 5 may be changed, some operations may be omitted, and some operations may be performed simultaneously.

[0080] In operation 510, the system (400) can continuously collect the user's emotional input, convert it into a time-series-based emotional waveform and tension score, and generate an emotional rhythm. In this process, the system (400) can collect emotional data through various channels, such as the user's text input, voice patterns, and facial expression changes. The collected emotional data can be processed through an emotional analysis model, such as HuggingFace Transformers, and converted into emotional labels (e.g., joy, sadness, anger, fear, etc.) and intensity values. The system (400) can arrange this emotional data along a time axis to form a time-series-based emotional waveform and quantify the tension score for each emotional state into a value within a normalized numerical range (e.g., 0.0 to 1.0). The emotional rhythm thus generated can represent the user's emotional state changes as a continuous flow.

[0081] In operation 520, the system (400) can store the generated emotional rhythms long-term, like biological organs, in local storage embedded in the body or a wearable device. In this storage process, the system (400) can utilize SQLite and flash memory-linked local storage to store the emotional data as structured data including timestamps, emotional labels, scores, and alert statuses. Unlike conventional ephemeral emotion recognition systems, the emotional rhythms stored in this way can continuously record and manage the user's emotional state, similar to how biological organs function. The emotional data stored in the local storage provides a basis for analyzing patterns of emotional change over time over the long term, which can be used as important information for a comprehensive understanding of the user's emotional health status.

[0082] In operation 530, the system (400) can detect the user's emotional tension and danger signals in real time by combining the rate of change of stored emotional rhythms, keyword analysis, and physiological signal data, and can automatically generate and execute a protective response in response to the detected crisis situation. The system (400) can determine a crisis situation if the emotional tension score drops sharply, if danger keywords (e.g., I'm scared, I'm frustrated) are detected, or if abnormalities are detected in biosignals such as BPM (heart rate) and GSR (gastroskin response). Upon detection of such a crisis, the system (400) can automatically generate and execute a protective response appropriate to the situation. The protective response can be provided in various forms, such as voice over messages via Coqui TTS, vibration patterns via Haptic signals, or silent text messages via a display, and each protective response can be determined according to a predefined mapping table for each emotion type. In addition, the system (400) can control the conversion of the emotional state record into a medical data format so that it can be linked with a healthcare system.

[0083] According to one embodiment, the system (400) continuously collects the user's emotional input, converts it into a time-series-based emotional waveform and tension score, generates an emotional rhythm, stores the generated emotional rhythm long-term like a biological organ in a local storage built into the body or a wearable device, detects the user's emotional tension and danger signals in real time by combining the rate of change of the stored emotional rhythm, keyword analysis, and physiological signal data, automatically generates and executes a protective response including a silent comfort message, screen guidance, and vibration notification in response to the detected crisis situation, and controls the conversion of the emotional state record into a medical data format so that it can be linked with a healthcare system.

[0084] According to one embodiment, the system (400) may continuously collect emotion-related data from various sources, such as the user's text input, voice utterance, or connected application activity. This continuous collection may be intended to identify the user's emotional changes over time and understand them as a continuous flow rather than fragmentary information. The system (400) may convert the collected raw data into emotion labels, such as joy, sadness, anger, and fear, and tension scores representing the intensity of the corresponding emotion, by utilizing a natural language processing (NLP) model or a voice analysis engine. This converted data is arranged along a time axis to form a time-series-based emotion waveform, and the fluctuation pattern of this waveform itself may be defined as the user's unique emotion rhythm.

[0085] According to one embodiment, the system (400) may store the generated emotional rhythm data in a local storage in the form of a lightweight database such as SQLite or flash memory of a wearable device or an implanted device owned by the user, rather than a cloud server, in order to protect the user's privacy and increase data accessibility. Storing it like a biological organ means a permanent and organic method of data storage and utilization in which data is continuously accumulated like a part of the body and the user's condition is determined based on it, rather than being consumed once.

[0086] According to one embodiment, the system (400) may use a multifaceted approach to detect a crisis situation. First, by analyzing the rate of change of emotional scores in stored emotional rhythm data, it may detect patterns in which positive emotions rapidly decrease or negative emotions rapidly increase within a short period of time. At the same time, it may detect specific keywords that suggest a crisis situation, such as "difficult," "scared," or "want to die," in text or voice input in real time. Additionally, it may detect rapid changes in physiological signal data, such as heart rate (BPM) or gross skin response (GSR), collected from wearable sensors, and combine this with the results of emotional rhythm changes and keyword analysis to improve the accuracy of judgment regarding the crisis situation. This combination may be implemented through a weighting model or a machine learning-based classifier, thereby enabling a comprehensive judgment of the user's emotional tension and danger signals.

[0087] According to one embodiment, the system (400) can automatically execute a response to immediately support and protect the user when a crisis situation is determined. For example, considering the possibility that the user is in a public place, it may display a comforting message on the device screen silently, such as "It's okay, take a breath," or provide a vibration notification of a specific pattern that provides a sense of stability through a haptic motor. Additionally, it may provide guidance on simple meditation or breathing exercises through screen guidance. These protective responses may be customized according to the severity or type of the detected crisis, the user's presets, etc.

[0088] According to one embodiment, the system (400) can systematically record and manage all collected and analyzed emotional data (emotional rhythms, crisis detection records, protective response logs, related physiological signals, etc.). With the user's consent, these records can be converted into a standard medical data format such as HL7 FHIR (Fast Healthcare Interoperability Resources) or other formats that are interoperable with healthcare systems.

[0090] FIG. 6 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment.

[0091] The operations described through FIG. 6 can be implemented based on instructions that can be stored in a computer recording medium or memory (e.g., memory (430) of FIG. 4). The order of each operation of FIG. 6 may be changed, some operations may be omitted, and some operations may be performed simultaneously.

[0092] In operation 610, the system (400) can generate time-series-based emotional waveform data for emotional rhythm analysis and quantify tension scores for each emotion into values ​​within a normalized numerical range. In this process, the system (400) can utilize various emotion analysis models (e.g., HuggingFace Transformers) to multidimensionally analyze the user's emotional state and convert it into waveform data arranged along a time axis. Emotional tension scores can be expressed as normalized values ​​between 0.0 and 1.0, which can numerically represent the intensity and duration of the emotion. Through this quantification process, the system (400) can convert subjective emotional states into objectively measurable data, thereby laying the foundation for analyzing patterns of emotional change and detecting crisis situations.

[0093] In operation 620, the system (400) utilizes a local storage linked to SQLite and flash memory to store emotional data in a data structure including timestamps, emotional labels, scores, and warning states, and can collect real-time physiological signals from biosensors including BPM (heart rate) and GSR (gastroskin response) to integrate and analyze with the emotional data. In this process, the system (400) can calculate the rate of change of each sensor data to derive a correlation with the emotional score. The local storage can be implemented in a wearable device or an embedded system, thereby allowing emotional and biometric data to be collected and stored in a continuous and non-invasive manner during the user's daily life. Through integrated analysis, the system (400) learns patterns between emotional states and physiological responses, and based on this, can develop a more accurate ability to detect crisis situations.

[0094] In operation 630, the system (400) determines a crisis situation when the emotion score falls below a preset lower threshold or the rate of change of the biosignal exceeds a preset upper threshold, and can selectively execute an appropriate protective response according to the type of crisis situation determined. The threshold for determining a crisis situation can be customized for each user, and the system (400) can build a personalized crisis detection model by learning the user's normal emotion patterns and biosignal baseline. When a crisis situation is detected, the system (400) can selectively execute at least one of generating a voice message over Coqui TTS, transmitting a vibration pattern via Haptic signals, or displaying a silent text message via a display, depending on the severity and type of the situation. Each protective response can be determined according to a predefined mapping table for each emotion type, which can provide optimized support by taking into account the user's current situation and preferences.

[0095] In operation 640, the system (400) can synchronize collected biometric data with emotional rhythm data in real time by linking with wearable APIs including Fitbit and Apple HealthKit, and can activate an API that automatically sends a real-time report to a pre-designated guardian upon detection of a crisis. Additionally, the system (400) can build a database that can be used for long-term emotional health tracking and medical consultation by logging all changes in emotional state and the execution history of protective responses. Through linkage with wearable APIs, the system (400) can perform more accurate crisis detection by integrating biometric data collected from various wearable devices. Through the guardian notification function, in the event of a serious crisis, an immediate notification can be sent to a designated guardian or medical professional to request additional support. These functions can play an important role in ensuring the safety of users, particularly those at risk of self-harm or those suffering from mental health issues.

[0096] According to one embodiment, the system (400) generates time-series-based emotional waveform data for emotional rhythm analysis and quantifies tension scores for each emotion into values ​​within a normalized numerical range; stores emotional data in a data structure including timestamps, emotional labels, scores, and warning states using SQLite and flash memory-linked local storage; collects real-time physiological signals from biosensors including BPM (heart rate) and GSR (gastroskin response) and integrates and analyzes them with emotional data, calculates the rate of change of each sensor data to derive a correlation with the emotional score; determines a crisis situation when the emotional score falls below a preset lower threshold or the rate of change of the biosignal exceeds a preset upper threshold; selectively executes at least one of generating a voice overlay message via Coqui TTS, transmitting a vibration pattern via a Haptic signal, or displaying a silent text message via a display according to the type of crisis situation determined, wherein each protective response is determined according to a mapping table predefined for each emotion type; synchronizes collected biosignal data with emotional rhythm data in real time by linking with wearable APIs including Fitbit and Apple HealthKit; and, upon detection of a crisis, pre-designated It is possible to control the operation of an API that automatically sends real-time reports to guardians, and to build a database that can be utilized for long-term emotional health tracking and medical consultations by logging all changes in emotional state and the execution history of protective responses.

[0097] According to one embodiment, the system (400) may generate time-series waveform data representing changes in emotion over time based on input data to analyze emotional rhythms more precisely. At this time, a tension score representing the intensity of each emotion (e.g., joy, sadness) is assigned, and this score may be expressed as a value within a normalized numerical range that allows for consistent comparison, such as between 0 and 1 or between -1 and +1. This may be intended to objectively compare and track various emotional states.

[0098] According to one embodiment, the system (400) may utilize a lightweight database system, such as SQLite, which is widely used in smartphones or wearable devices, or flash memory to efficiently manage and store such quantified emotion data. The data may be stored in a structured form that includes not only emotion labels and scores, but also a timestamp indicating the exact time when the emotion was recorded, an emotion label indicating the type of emotion, a score indicating the intensity, and a warning status field indicating whether a crisis situation has occurred. This can facilitate the retrieval, analysis, and utilization of the data.

[0099] According to one embodiment, the system (400) may utilize physiological indicators such as heart rate (BPM) or skin electrical response (GSR) to more accurately determine a psychological state. To this end, it may collect real-time biosignals by linking with APIs provided by various wearable devices and platforms, such as Fitbit and Apple HealthKit, and synchronize this data with emotional rhythm data. Rather than simply looking at current values, it may perform a more in-depth analysis by calculating the rate of change per hour of each sensor data and comparing it with changes in emotional scores to derive a correlation between a specific emotional state and a physiological response.

[0100] According to one embodiment, the system (400) may establish clear criteria for determining a crisis situation. For example, a rule-based or machine learning-based algorithm may be used to determine a crisis situation when conditions are met, such as when a user's tension score drops sharply below a specific threshold (e.g., 0.2) or when the rate of change in BPM exceeds a specific threshold (e.g., an increase of 30 or more times per minute) relative to a stable state. This threshold may be adjusted to suit the characteristics of individual users.

[0101] According to one embodiment, when a crisis situation is identified, the system (400) can select and execute the most appropriate protective response according to the type of situation (e.g., acute stress, deepening of depression) and severity. For example, it may deliver a comforting message in a warm voice using text-to-speech technology such as Coqui TTS, transmit a vibration pattern that provides a sense of stability using the haptic function of a wearable device, or display quiet words of comfort on the screen. Which response to execute may be determined according to a predefined mapping table or a learned model, such as vibration + a comforting message when fear is detected, or voice comfort when sadness is detected.

[0102] According to one embodiment, the system (400) may activate an API that can automatically send a notification to a designated guardian, such as a family member or friend, when a serious crisis situation is detected, provided that the user has previously agreed to and designated it. This notification may include the fact that a crisis situation has occurred and, if necessary, the user's location information, and this can serve as an important means to ensure the user's safety.

[0103] According to one embodiment, the system (400) may record all activities, such as changes in emotion, changes in vital signs, detection of crisis situations, and execution of protective responses, as detailed logs and store them in a database. The accumulated data may aim to build a valuable resource that not only helps the user track and understand their emotional health status over the long term but can also be used as objective data in consultations with medical professionals when necessary.

[0105] FIG. 7 is a flowchart illustrating a method for storing user emotions and detecting crises based on an emotional artificial organ according to one embodiment.

[0106] The operations described through FIG. 7 can be implemented based on instructions that can be stored in a computer recording medium or memory (e.g., memory (430) of FIG. 4). The order of each operation of FIG. 7 may be changed, some operations may be omitted, and some operations may be performed simultaneously.

[0107] In operation 710, the system (400) can classify and store the user's emotional data by family member and accumulate emotional associations for each member. In this process, the emotional association can be calculated by multiplying the frequency and intensity of emotional expressions related to the member. The system (400) can build an emotional profile for each relationship by continuously monitoring the emotional state when the user mentions family members or interacts with them, and classifying and storing this by member. The emotional association accumulates over time and can show emotional patterns occurring in a relationship with a specific family member. This data can be used as a basis for understanding the dynamics of family relationships, identifying potential relationship problems early, and strengthening emotional bonds within the family.

[0108] In operation 720, the system (400) can analyze the degree of emotional contagion among family members. In this process, the probability that a similar emotion will appear in another member within 24 hours of a specific member's negative emotion occurring can be calculated and quantified as an emotional contagion coefficient to construct an emotional influence network within the family. Through the analysis of emotional contagion, the system (400) can identify patterns of how emotions are propagated and shared within the family. This analysis can help understand the impact of one member's negative emotional state on other members and provide insights to support the emotional well-being of the entire family. The emotional influence network can help to more clearly understand the emotional dynamics within the family by visualizing and quantifying the influence each member has on the emotional atmosphere within the family.

[0109] In operation 730, the system (400) can generate emotional legacy based on collected emotional data. In this process, text messages are encrypted using AES-256 encryption and voice messages are converted into Mel spectrograms, which are then stored in a Redis-based emotional message storage system along with the sender, recipient, emotional label, intensity value, and time of creation. Integrity can be ensured by assigning a unique hash value to each emotional legacy. Emotional legacy can function as a digital legacy that preserves and transmits meaningful emotional moments or messages shared among family members over a long period. The system (400) can strengthen emotional bonds between family members by securely storing messages generated during special anniversaries, important family events, or moments of deep emotional connection, and delivering them to recipients at appropriate times. The security and reliability of such emotional legacy can be ensured through advanced encryption technology and integrity verification mechanisms.

[0110] In operation 740, the system (400) can calculate well-being indicators in real time, including the emotional resonance rate and the dissonance rate among family members. In this process, the resonance rate can be defined as the rate at which the same emotional label appears consecutively, and the dissonance rate as the rate at which opposing emotional labels appear alternately. Additionally, the system (400) can derive a relationship intimacy score by analyzing the correlation between the emotional response delay time and the intensity of emotional expression. Furthermore, the system (400) can apply a generation-specific emotional interpretation algorithm that reflects the characteristics of the parent generation having a high rate of indirect expression and the child generation having a high rate of direct expression by analyzing the generational emotional expression patterns of family members. These well-being indicators and emotional pattern analysis can provide useful information for measuring the level of emotional harmony among family members, understanding the differences in emotional expression methods between generations, and evaluating the health of family relationships.

[0111] In operation 750, the system (400) can identify a member who acts as an emotional hub within the family. In this process, each member is represented as a node and emotional interactions as edges, and a centrality index is calculated to identify the central figure of the emotional network and predict the impact on the entire family when the member's emotional state changes. Additionally, the system (400) can automatically generate a message for relationship restoration if the emotional dissonance rate exceeds a threshold value set for each user. In this process, the topic and time that showed the highest resonance rate in the past positive emotional history with the family member can be extracted, and a customized recall-inducing message can be constructed and suggested through a natural language generation model. Through these functions, the system (400) can detect conflict or emotional discord within the family and support the family's emotional well-being and the formation of healthy relationships by providing customized interventions for relationship restoration.

[0112] According to one embodiment, the system (400) classifies and stores user emotional data by family member and records the emotional correlation for each member cumulatively, wherein the emotional correlation is calculated as the value obtained by multiplying the frequency and intensity of emotional expressions related to the member; analyzes the emotional contagion among family members, calculates the probability that a similar emotion appears in another member within 24 hours after a specific member's negative emotion occurs, and quantifies this as an emotional contagion coefficient to construct an emotional influence network within the family; generates emotional legacy based on the collected emotional data, applies AES-256 encryption to text messages and converts voice messages into Mel spectrograms, stores them in a Redis-based emotional message storage system along with the sender, receiver, emotional label, intensity value, and time of creation, and assigns a unique hash value to each emotional legacy to ensure integrity; calculates well-being indicators including the emotional resonance rate and dissonance rate among family members in real time, where the resonance rate is defined as the rate at which the same emotional label appears consecutively and the dissonance rate is defined as the rate at which opposing emotional labels appear alternately; additionally analyzes the correlation between emotional response latency and emotional expression intensity to derive a relationship intimacy score, and [describes] the generation of family members By analyzing emotional expression patterns, a generation-specific emotion interpretation algorithm is applied that reflects the characteristics of the parent generation, which has a high proportion of indirect expressions, and the child generation, which has a high proportion of direct expressions. This algorithm identifies the family member acting as the emotional hub, represents each member as a node and emotional interactions as edges, calculates centrality indicators to identify the central figure in the emotional network, predicts the impact on the entire family when that member's emotional state changes, and automatically generates a message for relationship restoration when the emotional dissonance rate exceeds a threshold value set for each user. Furthermore, it can be controlled to extract the topic and time period that showed the highest resonance rate from the past positive emotional history with that family member, and construct and propose a customized recall-inducing message through a natural language generation model.

[0113] According to one embodiment, the system (400) can analyze and utilize the user's emotional data within a family relationship network, going beyond simply treating it as personal data. To this end, the system (400) can classify and store emotions by linking them to a specific family member (e.g., mother, son) when the user mentions that specific family member (e.g., mother, son) during emotional expression (text, voice, etc.) or during a conversation with that specific member. Furthermore, the system can calculate an emotional correlation by synthesizing how often and how strongly emotions are expressed toward each member and record this cumulatively. This correlation can be used as an indicator to identify the emotional importance or intensity of a specific family relationship.

[0114] According to one embodiment, the system (400) can calculate emotional contagion to analyze how emotions influence each other among family members. For example, if a pattern of one member expressing a negative emotion (e.g., stress, sadness) and another member expressing a similar negative emotion within a certain period of time (e.g., 24 hours) appears repeatedly, the probability can be calculated and quantified as an emotional contagion coefficient. By combining these coefficients, an emotional influence network can be constructed to visualize or analyze who has a significant emotional influence on whom within the family, which can help in understanding the emotional health of the entire family.

[0115] According to one embodiment, the system (400) can provide a function to safely preserve precious emotional exchanges between family members as an emotional legacy. Text messages that the user deems important or that the system deems positive and intense can be protected by a strong encryption algorithm such as AES-256, and voice messages can be converted into a Mel spectrogram form that can be stored efficiently while preserving their characteristics. The data processed in this way can be stored in an in-memory database system that allows for fast access, such as Redis, along with the sender, recipient, emotional information, and time. Additionally, by assigning a unique hash value to each legacy, data tampering can be prevented and integrity can be guaranteed.

[0116] According to one embodiment, the system (400) can calculate well-being indicators in real time to evaluate the quality of family relationships. The emotional resonance rate represents the rate at which family members exchange positive emotions with one another and are emotionally in agreement, and the emotional dissonance rate represents the rate at which conflicting or negative emotions clash. In addition, the depth of the relationship can be assessed from various angles by deriving a relationship intimacy score through the analysis of how quickly a response to a message is received (response delay time) and how strongly one responds with emotion.

[0117] According to one embodiment, the system (400) can improve the accuracy of emotion interpretation by taking into account differences in communication styles between generations. For example, if data analysis reveals that the parent generation tends to express emotions indirectly while the child generation tends to express them directly, the system can learn these generational emotion expression patterns and apply a customized emotion interpretation algorithm tailored to each generation. This can contribute to reducing misunderstandings and increasing mutual understanding.

[0118] According to one embodiment, the system (400) can identify a member who acts as an emotional hub within the family by utilizing social network analysis techniques. After constructing a network in which family members are nodes and emotional interactions (message exchange, emotional contagion, etc.) are edges, a centrality score is calculated to determine who is at the center of emotional exchange. Through this, the influence that the emotional state of this central figure may have on the entire family can be predicted, and preemptive support can be considered if necessary.

[0119] According to one embodiment, the system (400) may suggest intervention for relationship recovery when the emotional dissonance rate among family members rises above a certain threshold and difficulties in the relationship are detected. The system may analyze past interaction records with the relevant member to identify specific conversation topics or shared experiences (times) that showed the most positive and strong emotional resonance. Then, by utilizing a natural language generation model such as GPT, it may generate and suggest to the user a customized recall-inducing message that prompts conversation by recalling these positive memories.

[0121] According to one embodiment, the system (400) generates a 768-dimensional vector of user utterance using BERT Embedding, extracts stylistic features including sentence length, ending pattern, interjection frequency, question ratio, and average word length, classifies speech types through K-means clustering, converts them into speech vectors, performs cumulative learning by associating emotional rhythms and speech patterns by specific topics or people, stores them in a 3-dimensional tensor structure of topic-emotion-speech, and quantifies the association strength by calculating the weights between each dimension using cosine similarity, searches for a past topic most similar to the current conversation topic when conversing with a third party, extracts emotional rhythms and speech vectors connected to the topic, and generates a response that reproduces the user's usual speech style by injecting the extracted speech features as style parameters into the prompt of a GPT-based language model, collects user feedback on the generated response, scores positive feedback as +1 and negative feedback as -1, adjusts the growth level of the AI ​​object upward whenever the cumulative score passes a specific interval, and gradually expands the vocabulary complexity and diversity of emotional expressions available at each level. Ethical filters are applied to responses generated by AI objects, and text containing profanity, discriminatory expressions, or personal information is detected using regular expressions and blacklist dictionaries. When detected, the relevant parts are automatically replaced with neutral expressions or the entire response is regenerated. Furthermore, if different tone patterns appear for the same emotion at different times during the tone learning process, this is modeled as a circadian rhythm to apply differentiated tone patterns for morning, day, evening, and night times, and the length and complexity of the response can be dynamically controlled to reflect user fatigue.

[0122] According to one embodiment, the system (400) can generate a 768-dimensional vector of a user's utterance by utilizing BERT Embedding. BERT Embedding is a pre-trained language model widely used in the field of natural language processing that can generate context-aware word representations. This 768-dimensional vector can capture even fine nuances by representing the semantic characteristics of the user's utterance in a high-dimensional space. Along with this vector, the system (400) can extract stylistic features including sentence length, ending patterns, interjection frequency, question ratio, and average word length. Stylistic features are important factors in identifying a user's unique language usage patterns, thereby enabling the construction of a personalized speech profile.

[0123] According to one embodiment, the system (400) can classify speech types through K-means clustering and convert them into speech vectors. K-means clustering is an unsupervised learning algorithm that can be effective for grouping data points with similar characteristics. The system (400) can use this algorithm to classify various utterances of a user into several representative speech types, and each type can be represented by a unique vector. This speech vectorization may be a process of converting the user's linguistic characteristics into a form that is easy to process in a machine learning model by numerically representing them.

[0124] According to one embodiment, the system (400) can learn cumulatively by associating emotional rhythms and speech patterns for specific topics or people. In this process, it can be stored as a 3D tensor structure of topic-emotion-speech, and the strength of association can be quantified by calculating the weights between each dimension as cosine similarity. The 3D tensor structure is a mathematical model capable of efficiently representing and processing multidimensional data, and can organize data while maintaining complex relationships. Cosine similarity is a method of measuring the angle cosine value between two vectors, and can quantify the directional similarity of vectors as a value between -1 and 1. In this way, the system (400) can learn and store the correlation between the user's emotional state and speech for a specific topic or person.

[0125] According to one embodiment, the system (400) can search for past topics most similar to the current topic of conversation when conversing with a third party. Subsequently, it can extract the emotional rhythm and speech vectors associated with the topic and inject the extracted speech features as style parameters into the prompt of a GPT-based language model to generate a response that reproduces the user's usual speech style. GPT (Generative Pre-trained Transformer) is a natural language generation model pre-trained with large-scale text data, capable of generating text of various styles and contents. By providing the user's unique speech features as parameters to this model, the system (400) can generate a natural response as if the user were speaking directly. This may be a key function that enables a digital persona or AI agent to interact with a third party while maintaining the user's linguistic characteristics.

[0126] According to one embodiment, the system (400) can collect user feedback on the generated response. In this process, positive feedback is scored as +1 and negative feedback as -1, and the growth level of the AI ​​object can be adjusted upward whenever the accumulated score passes a specific interval. The vocabulary complexity and diversity of emotional expressions available at each level can be expanded step by step. This feedback-based growth system may be a mechanism that enables the AI ​​object to gradually develop and adapt according to the user's preferences and reactions. The expansion of vocabulary complexity and emotional expressions by level may enable the AI ​​object to communicate in more sophisticated and diverse ways as it grows.

[0127] According to one embodiment, the system (400) may apply an ethical filter to a response generated by an AI object. In this process, text containing profanity, discriminatory expressions, or personal information may be detected through regular expressions and a blacklist dictionary. If detected, the corresponding part may be automatically replaced with a neutral expression or the entire response may be regenerated. Regular expressions are formal languages ​​used to search for and manipulate strings of specific patterns, and can efficiently identify specific patterns within text. The application of an ethical filter may serve as an important safeguard to ensure that the AI ​​object generates only safe and appropriate content while complying with social norms and ethical guidelines.

[0128] According to one embodiment, the system (400) can model different speech patterns for the same emotion at different times of the day as a circadian rhythm during the speech learning process. Through this, differentiated speech patterns can be applied according to morning, day, evening, and night times, and the length and complexity of the response can be dynamically adjusted by reflecting the user's fatigue. A circadian rhythm is a biological process that repeats with a cycle of approximately 24 hours and can also affect human emotions and language expressions. By reflecting these biological rhythms in the language generation model of an AI object, the system (400) can generate more natural and situationally appropriate responses that take into account the user's language usage patterns and emotional state at different times of the day. Additionally, by detecting the user's fatigue and adjusting the complexity of the response accordingly, the user experience can be optimized and effective communication can be promoted.

[0130] According to one embodiment, the system (400) predicts the trend of a user's emotional change through Prophet-based time series analysis and creates an emotional prediction model for periods including 7, 14, and 30 days, applying different weights for each period to improve prediction accuracy, automatically generates an emotional report via FastAPI and sends it to a designated recipient, the report including an emotional change graph, major emotional keywords, and predicted emotional trends, performs an emotional analysis of the recipient's response to the sent report to calculate the positive / negative response ratio and updates the relationship change indicator, while also considering the response time and response length to calculate the relationship quality score, and based on the updated relationship indicator, the response generation module of the system dynamically adjusts the response tone and content, and controls the use of more intimate expressions as the relationship quality score increases.

[0131] According to one embodiment, the system (400) can predict the trend of a user's emotional change through Prophet-based time series analysis. Prophet is a time series prediction tool developed by Facebook that can provide a powerful prediction model that takes into account seasonality and holiday effects. The system (400) can utilize this tool to generate an emotional prediction model for periods including 7, 14, and 30 days, and can improve prediction accuracy by applying different weights to each period. By applying different weights to short-term, medium-term, and long-term predictions, the system (400) can capture various patterns of emotional change over time and generate more accurate prediction results.

[0132] According to one embodiment, the system (400) can automatically generate an emotion report via FastAPI and send it to a designated recipient. FastAPI is a modern, high-performance web framework that can provide asynchronous processing and automatic documentation capabilities. The report generated by the system (400) may include an emotion change graph, key emotion keywords, and predicted emotion trends. This report provides a comprehensive overview of the user's emotional state and can help a designated recipient (family, medical professionals, caregivers, etc.) monitor the user's emotional health and provide necessary support.

[0133] According to one embodiment, the system (400) can analyze the sentiment of the recipient's response to a transmitted report to calculate the positive / negative response ratio and update the relationship change indicator. In this process, the response time and response length can also be considered to calculate the relationship quality score. Sentiment analysis is a natural language processing technique that identifies the polarity (positive, negative, neutral) of emotions or opinions expressed in text, and the system (400) can quantitatively evaluate the recipient's response through this. The response time and length can be used as indirect indicators representing the level of interest and engagement in the relationship, and the relationship quality score calculated by synthesizing these elements can serve as an important indicator for measuring the level of emotional connection between the user and the recipient.

[0134] According to one embodiment, the system (400) allows the response generation module of the system to dynamically adjust the response tone and content based on updated relationship indicators. In this process, the system can control the use of more intimate expressions as the relationship quality score increases. The dynamic adjustment of the response tone can promote natural and effective interaction between the user and the recipient by applying an appropriate communication style that matches the current state of the relationship. Adjusting the level of intimacy based on the relationship quality score enables customized communication that takes into account the developmental stage and characteristics of the relationship, which can contribute to maintaining and strengthening the long-term relationship.

[0136] According to one embodiment, the system (400) converts a user’s emotional rhythm pattern into a musical waveform, maps emotional intensity to volume, emotional change rate to frequency, and emotional type to timbre to generate a unique emotional signature for each individual, converts the generated emotional signature into the frequency domain using a Fast Fourier Transform (FFT) algorithm, calculates cosine similarity to automatically match users with similar emotional signatures to form an emotional resonance network, and when a crisis is detected, collects anonymized comfort messages from other users within the emotional resonance network, encrypts the sender’s identification information with a hash function when collecting messages, extracts only the message content, and transmits it, measures the effect of the received comfort message by the emotional change rate before and after receiving the message, learns the most effective comfort pattern using a machine learning model, and controls the system to prioritize the use of the learned pattern in future crisis response.

[0137] According to one embodiment, the system (400) can convert a user's emotional rhythm pattern into a musical waveform. In this process, emotional intensity can be mapped to volume, the rate of change in emotion to frequency, and the type of emotion to timbre to generate a unique emotional signature for each individual. The conversion into a musical waveform is a method of expressing abstract emotional data in a physically recognizable form, thereby enabling the visualization and auditory representation of emotional patterns. The emotional signature is a digital representation that encapsulates each user's unique emotional characteristics and can be utilized for personal identification and emotional state analysis.

[0138] According to one embodiment, the system (400) can convert the generated emotion signature into the frequency domain using a Fast Fourier Transform (FFT) algorithm. The FFT is an efficient algorithm for converting signals in the time domain into the frequency domain and can be used to analyze the frequency components of complex waveforms. The system (400) can form an emotion resonance network by calculating cosine similarity based on the converted data and automatically matching users with similar emotion signatures. The emotion resonance network is a connection structure of users with similar emotion patterns, through which users can have the opportunity to share and empathize with each other's emotional experiences.

[0139] According to one embodiment, the system (400) can collect anonymized comfort messages from other users within the emotional resonance network when a crisis is detected. In this process, when collecting messages, the sender's identification information can be encrypted using a hash function, and only the message content can be extracted and transmitted. A hash function is an encryption technique that maps data of arbitrary size to a value of fixed size, and can provide a one-way transformation that prevents the restoration of the original data. The anonymization process can create a safe and reliable emotional sharing environment by enabling emotional support while protecting the sender's privacy.

[0140] According to one embodiment, the system (400) can measure the effectiveness of a received comforting message by the rate of change in emotion before and after receiving the message. Through this, the most effective comforting pattern can be learned using a machine learning model, and the learned pattern can be prioritized for use in future crisis response. The rate of change in emotion is an indicator that quantifies the difference in emotional state before and after a specific intervention (such as a comforting message), allowing for an objective evaluation of the effectiveness of the intervention. Learning effective comforting patterns through a machine learning model enables the system (400) to develop more effective crisis response strategies over time, which can contribute to providing customized support for the user's emotional recovery and stability.

[0142] According to one embodiment, the system (400) combines emotional rhythm data and biosignal patterns to construct an emotion-physiological response model for each user, derives a correlation coefficient between a specific emotional state and a biosignal through regression analysis, detects intrinsic emotional changes occurring even in the absence of external stimuli based on the constructed model, detects minute fluctuations in biosignals relative to a baseline to identify an unconscious stress state, and in response to the identified unconscious stress, transmits low-frequency vibrations in the range of 1 Hz to 30 Hz through a wearable device or plays ultra-low frequency sounds in the range of 40 Hz to 80 Hz through a bone conduction speaker to activate the parasympathetic nervous system to induce emotional stabilization, applies a Fourier transform to emotional rhythm data accumulated over a long period to extract periodic patterns, and controls the system to predict seasonal affective disorder or periodic emotional patterns through a seasonal decomposition algorithm.

[0143] According to one embodiment, the system (400) can construct a user-specific emotion-physiological response model by combining emotion rhythm data and biosignal patterns. In this process, a correlation coefficient between a specific emotional state and a biosignal can be derived through regression analysis. Regression analysis is a statistical method that models and analyzes the relationship between variables, and can quantify the correlation between a dependent variable (emotional state) and an independent variable (biosignal). By capturing the association between an individual's unique physiological response pattern and emotional state, the user-specific emotion-physiological response model can provide a basis for more accurately inferring and predicting emotional states based on objective biosignal data.

[0144] According to one embodiment, the system (400) can detect intrinsic emotional changes that occur even in the absence of external stimuli based on a constructed model. In this process, it can identify an unconscious stress state by detecting minute fluctuations in biosignals relative to a baseline. Intrinsic emotional changes are changes in emotional state that occur internally without distinct external factors, and often the user themselves may not be aware of them. Identifying an unconscious stress state can play an important role in preventing serious emotional crises through early intervention, which can contribute to the user's overall mental health management.

[0145] According to one embodiment, the system (400) can transmit a specific range of vibrations or sounds through a wearable device or a bone conduction speaker in response to identified unconscious stress. Specifically, low-frequency vibrations in the range of 1 Hz to 30 Hz can be transmitted through a wearable device, or ultra-low frequency sounds in the range of 40 Hz to 80 Hz can be played through a bone conduction speaker to activate the parasympathetic nervous system and induce emotional stabilization. Low-frequency vibrations and ultra-low frequency sounds can resonate with the body's natural vibration frequencies to induce a relaxation response. The activation of the parasympathetic nervous system promotes a "rest-digest" response, which can bring about stress-relieving effects such as a decrease in heart rate, a drop in blood pressure, and the relaxation of muscle tension, which can help restore the user's physiological and emotional stability.

[0146] According to one embodiment, the system (400) can extract periodic patterns by applying a Fourier transform to emotional rhythm data accumulated over a long period. Additionally, it can predict seasonal affective disorder or periodic emotional patterns through a seasonal decomposition algorithm. The Fourier transform is a mathematical technique that converts a signal in the time domain into the frequency domain and can be effective in identifying periodic components in complex waveforms. The seasonal decomposition algorithm is a method for separating trend, seasonal, periodic, and irregular components from time-series data and can be useful for analyzing pattern changes over time. Through this analysis, the system (400) can identify and predict periodic emotional patterns such as seasonal affective disorder (SAD), thereby providing proactive intervention and personalized support, which can contribute to promoting the user's long-term emotional well-being.

[0148] According to one embodiment, the system (400) may operate to build and maintain a model of the user's emotional homeostasis, going beyond simply storing the user's emotional rhythm. This may mean establishing a baseline or optimal range of the most stable and healthy emotional rhythm for each user through long-term data learning. When the system (400) detects that the current emotional rhythm deviates from this optimal range, it may not merely notify the user of a crisis situation, but may provide subtle feedback (e.g., changes in micro-vibration patterns, adjustment of background color tone) to help the user gently return to a state of emotional balance, or suggest activities that can induce a positive rhythm (e.g., a short walk, listening to specific music). This may implement a preventive approach aimed at actively maintaining a healthy state, rather than a response after the onset of illness.

[0149] According to one embodiment, the system (400) can analyze the interaction between the user's internal emotional rhythm as well as the external environment rhythm surrounding the user. For example, the system (400) can compare and analyze external environment data collected through sensors (e.g., light intensity, noise) and APIs (e.g., weather information, news headlines, schedules) of a smartphone or wearable device with emotional rhythm data in real time. Through this, it is possible to identify the impact of specific weather (e.g., rainy days) or social issues (e.g., a surge in negative news) on the user's emotional rhythm and learn the user's emotional sensitivity to specific environmental factors. Furthermore, based on these analysis results, the system (400) can support the user in better managing their emotions by warning the user in advance of environmental factors that may have a negative emotional impact, or conversely recommending environments that may promote a positive emotional rhythm (e.g., quiet parks, specific times of day).

[0150] According to one embodiment, the system (400) may include a function for detecting emotional foreshadowing that occurs before an obvious crisis signal occurs. This may mean identifying, through a machine learning model, subtle changes in the emotional rhythm waveform, such as the peak of positive emotion gradually decreasing, the volatility of the emotional waveform increasing abnormally, or patterns in which specific negative emotions appear frequently at a low intensity. By detecting these foreshadowing symptoms, the system (400) may perform a preemptive protection function to prevent the crisis in advance or mitigate its intensity by alerting the user before a full-blown crisis situation arrives or by performing mild emotional interventions (e.g., suggesting mood-altering activities, providing positive feedback).

[0151] According to one embodiment, the system (400) can build a protective response optimization loop by tracking and learning how the response actually affected the user's emotional rhythm after executing a protective response. That is, the system (400) can monitor how quickly and how stably the user's emotional tension score recovers after executing a specific protective response (e.g., a voice up message). By accumulating this feedback data, the system (400) can learn what kind of protective response (e.g., voice, text, vibration intensity and pattern, etc.) is most effective for each user and for different types of crisis situations.

[0153] According to one embodiment, the system (400) can analyze the phase interference pattern of the emotional rhythm to detect the synchronization phenomenon with the human biological rhythm and integrate it into a crisis prediction model. This phase interference pattern includes fine signals that are considered noise and removed in general emotion analysis systems, and the system (400) can amplify and pattern these fine signals through a Multi-scale Wavelet Transform to mathematically model the resonance phenomenon between the user's biological rhythm (e.g., circadian, lunar, seasonal cycle) and the emotional rhythm. Through this, it becomes possible to operate an emotional artificial organ in harmony with the user's entire biological system, going beyond simple emotion recognition.

[0154] According to one embodiment, the system (400) can efficiently store large-capacity emotional data in a micro-store using a pulse demodulation-based emotional compression algorithm. This algorithm selectively extracts only the major inflection points of the emotional rhythm and compresses the remaining data into a form that can be restored through mathematical interpolation, thereby enabling long-term emotional data ranging from months to years to be preserved within the limited storage space of a wearable device. In particular, the system (400) can maintain an optimal balance between storage efficiency and data accuracy by applying an adaptive sampling technique that automatically increases the data sampling density during emotionally significant moments and decreases the sampling density during emotionally stable periods.

[0155] According to one embodiment, the system (400) can implement an emotional artificial organ self-evolution mechanism that applies the principles of neuroplasticity. This mechanism can gradually adjust the structural parameters of the system according to the user's emotional patterns and reactions. Specifically, the system (400) can apply a Hebbian Learning Rule that strengthens neural network connections for pattern recognition for repeatedly occurring emotional patterns and weakens connection strengths for rarely occurring patterns. Through this, the system can gradually adapt to the user's unique emotional expression methods and patterns, thereby developing personalized emotional recognition and response capabilities.

[0156] According to one embodiment, the system (400) can detect micro-emotional oscillation patterns to detect precursors of an emotional crisis in advance. These micro-emotional oscillations are subtle instability in emotional rhythms that appear hours to days before major emotional changes occur, and are phenomena that are difficult to capture by general emotion analysis systems. By applying the Hilbert-Huang Transform, an advanced signal processing technique, to amplify and pattern these micro-oscillations, the system (400) can predict and take preventive measures before an emotional crisis fully manifests. This can play an important role in preventing acute mental health crises, such as suicidal impulses or severe depressive episodes.

[0157] According to one embodiment, the system (400) may apply dynamic protective response synthesis technology that considers the user's biological characteristics and emotional context when a crisis situation occurs. This technology does not simply execute predefined response messages or actions, but can dynamically synthesize an optimal protective response by analyzing various factors in real time, such as the user's current physiological state (autonomic nervous system activity, estimated hormone levels), the type and severity of the crisis, effective response patterns in similar past situations, and the user's personal preferences. Through this, the system (400) can provide customized emotional support optimized for each situation and user, rather than a uniform crisis response.

[0158] According to one embodiment, the system (400) can implement a vibration therapy module utilizing a bioresonance phenomenon. This module can analyze the user's current emotional state and physiological indicators to generate micro-vibrations of a specific frequency and pattern corresponding to them and transmit them through a wearable device. In particular, the system (400) can learn each individual's unique bio-vibrational signature and adjust the vibration pattern accordingly to induce an optimal physiological response. This bioresonance-based vibration therapy can promote physical and emotional stability in emotional crisis situations through effects such as restoring balance to the autonomic nervous system, reducing cortisol levels, and promoting endorphin secretion.

[0160] The embodiments described above may be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible.

[0161] The embodiments described above may be implemented as hardware components, software components, and / or combinations of hardware and software components. For example, the devices, methods, and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are also possible.

[0162] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or command the processing unit independently or collectively. Software and / or data may be permanently or temporarily embodied in any type of machine, component, physical device, virtual equipment, computer storage medium or device, or transmitted signal wave in order to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0163] Although the embodiments have been described above with reference to the limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, suitable results may be achieved even if the described techniques are performed in a different order than described, and / or if the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

Claims

Claim 1 A system for storing user emotions and detecting crises based on an emotional artificial organ, comprising: a memory for storing instructions; and a processor, wherein, when the instructions are executed by the processor, the system acquires emotional data including at least one of text data, voice data, and facial expression data input by a user, processes the emotional data through an emotional analysis model to convert it into emotional state and intensity values, arranges the emotional state and intensity values ​​along a time axis to determine a time-series-based emotional rhythm, stores the emotional rhythm in a local storage embedded in an electronic device inside or outside the user's body, collects bio-signals in real time through a bio-sensor, integrates and analyzes the bio-signals with the emotional data, calculates the rate of change of the bio-signals to derive a correlation with the emotional rhythm, determines a crisis situation when the emotional rhythm falls below a specified lower limit value or the rate of change of the bio-signals exceeds a specified upper limit value, selectively executes at least one of generating a message over voice, transmitting a vibration pattern, or displaying a message through a display according to the type of the determined crisis situation, and synchronizes the bio-signals and the emotional rhythm in real time. Claim 2 A system according to claim 1, wherein, when the instructions are executed by the processor, the system classifies and stores emotional data by family member and accumulates and records the emotional correlation for each member, wherein the emotional correlation is calculated by multiplying the frequency and intensity of emotional expressions related to each member, and analyzes the emotional contagion among family members, calculates the probability that a similar emotion appears to another member within a specified time after a specific member's negative emotion occurs, and quantifies this as an emotional contagion coefficient to construct an emotional influence network within the family. Claim 3 In claim 2, the above instructions, when executed by the processor, allow the system to calculate a well-being index including an emotional resonance rate and a dissonance rate among family members in real time, wherein the emotional resonance rate is defined as the rate at which the same emotional label appears consecutively and the dissonance rate is defined as the rate at which opposing emotional labels appear alternately, and if the dissonance rate exceeds a threshold value set for each user, the system automatically generates a message for restoring the relationship between family members, extracts the topic and time that showed the highest resonance rate in the past positive emotional history with the family member, and constructs and proposes a customized recall-inducing message through a natural language generation model.

Images