Intelligent health state analysis system and method based on artificial intelligence

Abstract

The invention discloses an intelligent health state analysis system and method based on artificial intelligence, and relates to the technical field of health data analysis, and the method comprises the steps: obtaining multi-dimensional data of a user and meaning anchor point data defined by the user; acquiring mark data of the user for associating behaviors or feelings with the significance anchor points; on the basis of all data, a personalized significance network graph model is constructed and dynamically maintained, and the weight of edges is combined with data-driven statistical association analysis and user mark-driven semantic connection learning; executing network path analysis, and identifying an effective association path between the current state and behavior of the user and the significance anchor point; personalized and meaning-oriented feedback information is generated based on the identified path, and connection between experience and personal meaning is revealed; and presenting feedback and performing closed-loop optimization through user evaluation. According to the method, the personalized significance network is constructed, and the feedback directly pointing to the personal significance of the user is provided, so that the internal motivation of the user, the long-term compliance of behavior change and the overall happiness are expected to be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of health data analysis, and in particular to an artificial intelligence-based health status intelligent analysis system and method. Background Art

[0002] Health self-management is crucial for preventing chronic diseases and improving quality of life. Existing technologies include a large number of health applications that use mobile devices and wearable sensors to collect users' physiological, behavioral, and environmental data. These applications typically provide functions such as data visualization, goal setting, general health knowledge push, and behavioral reminders based on simple rules or basic data patterns. For example, pedometer applications encourage users to reach their daily step goals, sleep trackers display sleep stage distribution, and some applications even use basic machine learning models for status prediction or risk assessment.

[0003] However, existing technologies generally face the problem of not being able to effectively stimulate users' intrinsic motivation and maintain long-term behavioral compliance. Users often lose interest after the initial novelty and find it difficult to integrate healthy behaviors into their daily lives and persist in them over the long term. This is largely because: 1) The collected data is often disconnected from the user's personal life meaning and deep goals, making it difficult for users to perceive the deeper meaning behind their daily behavioral choices; 2) The feedback information provided is usually universal, data-based, or directive, lacking connection with individual uniqueness and intrinsic value pursuits, making it difficult to trigger users' intrinsic driving force; 3) Although behavioral change theory emphasizes the importance of autonomy, competence, and consistency with values ​​to intrinsic motivation. Summary of the Invention

[0004] The present invention provides an artificial intelligence-based health status intelligent analysis system and method to solve the problem in the prior art that users often lose interest after the initial novelty and find it difficult to integrate healthy behaviors into daily life and persist in them for a long time.

[0005] In a first aspect, the present invention provides an artificial intelligence-based health status intelligent analysis method, comprising: A: Acquire the target user's time-stamped physiological data, behavioral data, and environmental data, and receive the user's input meaning anchor data representing their health-related goals through user interaction; B: obtaining, through the user interaction, semantic association tag data input by the user, wherein the tag data explicitly records the semantic association between a specific behavioral event or subjective feeling state determined by the user and the semantic anchor data; C: Based on all acquired data, construct and periodically or event-driven dynamically maintain a personalized meaning network model for the target user; D: When a preset trigger logic is met, performing network path analysis on the personalized meaning network model to identify a valid association path that meets preset conditions, starting from a source node corresponding to the current user state or behavior and reaching a target node corresponding to the meaning anchor data; E: Based on the attributes of the identified valid association paths, a feedback strategy library is invoked and feedback generation rules are matched to generate personalized and meaning-oriented feedback information. The feedback information is intended to reveal to the user the specific connection between their current experience and their personal meaning anchors, as learned by the model. F: Through the user interaction, presenting the generated feedback information to the target user in a preset format.

[0006] Preferably, while or after presenting the feedback information, interactive controls for evaluating the effectiveness or relevance of the feedback are provided to the user; the collected evaluation data is processed and analyzed through the configured feedback learning module; the analysis results are mapped into weight adjustment signals for relevant edges in the personalized meaning network model, or used to update the priority or trigger conditions of the strategies in the feedback strategy library, so as to achieve adaptive optimization of the system.

[0007] In a second aspect, the present invention further provides an artificial intelligence-based health status intelligent analysis system, comprising: Multi-dimensional data acquisition interface module for interacting with external sensors and user input devices; A personalized meaning network modeling engine, including a data-driven association learner and a semantic connection learner, and equipped with a dynamic maintenance mechanism; Meaning-oriented feedback engine, including trigger logic judgment unit, network path analyzer and feedback generator; User interaction module, used to present information, obtain user tags and feedback; Feedback learning and model optimization module is used to achieve closed-loop adaptive adjustment.

[0008] Preferably, a model interpretability submodule is constructed. When requested by the user or determined by the system as necessary, the model interpretability submodule is configured to perform at least one of the following operations: generating an evidence summary about a specific association edge based on the meaning network model and related original data fragments, the summary including the number of data points supporting the association, an association strength index, and applicable user tag records; for a certain generated meaning-oriented feedback, visualizing the key network path on which it is based or explaining its generation logic.

[0009] The technical solution provided by this application has at least the following technical effects or advantages: The personalized meaning network model constructed by the present invention can directly reveal the personalized connection learned by the model between the user's current experience and the unique meaning anchor points set by the user. This feedback goes beyond simple data presentation or behavioral suggestions and gives daily experience a deeper personal meaning.

[0010] Enhance users' awareness and understanding of the significance of healthy behaviors: By continuously receiving this meaning-oriented feedback, users can more clearly understand how their behavioral choices, even small daily behaviors, are related to their long-term goals, thereby deepening their understanding and recognition of the intrinsic value of a healthy lifestyle.

[0011] Improving intrinsic motivation and long-term behavioral compliance: This invention builds and utilizes a personalized meaning network to provide meaning-oriented feedback, directly impacting the underlying drivers of behavior change—the connection between behavior and personal meaning—rather than simply focusing on behavioral appearances or external incentives. This is fundamentally different from existing technologies, such as simple data association, goal check-ins, general suggestions, and even basic personalized recommendations.

[0012] Since the present invention helps users integrate their daily behaviors with their personal meaning framework, thus promoting a sense of life purpose and value consistency, in addition to behavioral changes, it may also indirectly but significantly improve users' deeper mental health indicators such as psychological resilience, subjective well-being, and life satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a flow chart of the health status intelligent analysis method based on artificial intelligence of the present invention; Figure 2 This is the architecture diagram of the health status intelligent analysis system based on artificial intelligence. DETAILED DESCRIPTION

[0014] The present invention relates to an AI-based intelligent health status analysis system and method, which aims to overcome the limitations of traditional health advice, stimulate the user's intrinsic motivation, and improve long-term self-management effects by building a personalized model that reflects the user's unique associations and providing feedback connected to the user's deep goals and values.

[0015] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention, rather than all, are shown in the drawings.

[0016] like Figure 1 The flowchart of the health status intelligent analysis method based on artificial intelligence is shown in the figure. The method includes the following steps: A: Acquire the target user's time-stamped physiological data, behavioral data, and environmental data, and receive the user's input meaning anchor data representing their health-related goals through user interaction; B: obtaining, through the user interaction, semantic association tag data input by the user, wherein the tag data explicitly records the semantic association between a specific behavioral event or subjective feeling state determined by the user and the semantic anchor data; C: Based on all acquired data, construct and periodically or event-driven dynamically maintain a personalized meaning network model for the target user; All data are obtained through step A and step B; D: When a preset trigger logic is met, performing network path analysis on the personalized meaning network model to identify a valid association path that meets preset conditions, starting from a source node corresponding to the current user state or behavior and reaching a target node corresponding to the meaning anchor data; E: Based on the attributes of the identified valid association paths, a feedback strategy library is invoked and feedback generation rules are matched to generate personalized and meaning-oriented feedback information. The feedback information is intended to reveal to the user the specific connection between their current experience and their personal meaning anchors, as learned by the model. F: Through the user interaction, presenting the generated feedback information to the target user in a preset format.

[0017] like Figure 2 As shown in the architecture diagram of the health status intelligent analysis system based on artificial intelligence, the system provided by the embodiment of the present invention includes the following modules: The multi-dimensional data acquisition interface module, personalized meaning network modeling engine, meaning-oriented feedback engine; user interaction module, and optionally, feedback learning and model optimization module and model interpretability subsystem module. These modules are implemented by one or more processors executing instructions stored in non-transitory memory. Data flows into the acquisition module from external sources (such as sensors and user input). The modeling engine processes data to construct a personalized meaning network model. The feedback engine generates meaning-oriented feedback based on this model, which is presented to users through the user interaction module and collected for evaluation. The feedback learning module uses the evaluation to optimize the model or strategy, forming a closed loop.

[0018] 1. The multi-dimensional data acquisition interface module is responsible for comprehensively and accurately obtaining data related to the user's health status and personal significance.

[0019] Objective data collection: Configure standardized interfaces for accessing multiple data sources, including smart wearable devices such as smart watches and wristbands; built-in sensors in mobile terminals such as accelerometers and global positioning system modules; and external application interfaces such as calendar application programming interfaces and weather application programming interfaces.

[0020] Implement hierarchical sampling strategies based on data type characteristics: Acquire high-temporal-resolution physiological data, such as beat-by-beat or second-level heart rate data for heart rate variability (HRV) analysis and second-level electrodermal activity (EDA).

[0021] Collect minute- or hour-level medium-frequency data, such as step count, activity intensity classification statistics, duration of each stage in the sleep architecture report, skin temperature or blood oxygen saturation trends.

[0022] Collect environmental and behavioral data that is aggregated daily, hourly, or triggered by specific events, such as screen usage time by application category, scene tags inferred based on location, such as home, office, commuting, ambient light and noise levels, and specific keywords in the calendar.

[0023] All collected data is accurately timestamped and aggregated to the processing center through reliable synchronization mechanisms, such as device-side software development kits and cloud-based application programming interfaces, with local data buffering capabilities to cope with network fluctuations.

[0024] Acquisition of meaning anchor data: A guided interface is provided through the user interaction module for users to set their core health-related goals such as "increasing energy" and "improving emotional stability".

[0025] The system assigns a unique internal target identifier to each meaning anchor defined or selected by the user, and allows the user to subsequently add, delete, and modify it.

[0026] Acquisition of meaningfully associated tag data: Provides a convenient user interaction method for receiving meaningfully associated tags input by the user; the tagging process can be initiated by the user after experiencing specific behaviors or feelings, or triggered by the system in a non-mandatory prompt manner when it recognizes a specific situation, such as when completing an exercise record.

[0027] The labeling process guides users to select or confirm the current behavior / feeling event and select one or more from the list of meaning anchors they have set to associate.

[0028] The system records semantic connection data including timestamp, behavior node identifier or feeling node identifier, and associated target identifier list.

[0029] As a preference, natural language input processing: a natural language processing unit can be configured with the user's explicit authorization and full notification of privacy risks.

[0030] This unit applies keyword extraction, sentiment analysis, relationship extraction and other technologies to process user-input text such as health diaries, identify potential behaviors, feelings, goals, and their associations, and submit the identification results to the user for confirmation as auxiliary input for meaning network modeling.

[0031] 2. The personalized meaning network modeling engine is responsible for building and dynamically maintaining a personalized meaning network model for each user. The model is implemented as a graph data structure containing a set of nodes and a set of weighted edges.

[0032] Node management: Maintain the user's node collection, including the following types of nodes: Objective data feature nodes: represent quantified physiological, behavioral, and environmental characteristics and their status values; Subjective feeling node: represents the subjective state marked by the user and its rating or category; Behavior node: represents the specific health-related behavior performed by the user and its occurrence attributes; Goal node: represents the health goal set by the user; Dynamically update a node set and its properties based on new user input or data analysis results.

[0033] Data-driven correlation edge learner: Input the user's long-term, pre-processed multi-dimensional feature time series data.

[0034] Core processing flow (periodic execution): Step 1: Candidate association pair screening: Calculate the long-term preliminary correlation measurement values ​​between all predefined objective data feature nodes, subjective feeling nodes, and behavior nodes.

[0035] Set a preliminary relevance judgment threshold.

[0036] Node pairs whose correlation metric exceeds the threshold are screened out as candidate association pairs for further analysis.

[0037] Step 2: Temporal causality or correlation depth verification (for each candidate correlation pair, denoted as node A and node B): Extract the time-aligned historical series data corresponding to node A and node B, and record them as time series A and time series B.

[0038] Performing time-series causal inference analysis: Initializing causal assumptions.

[0039] Apply the selected causal test method based on predictive power: evaluate whether using the past values ​​of time series A to predict the future values ​​of time series B significantly reduces the forecast error compared to using only the past values ​​of time series B, or vice versa.

[0040] Decision logic: Based on the significance of the reduction in prediction error, determine whether there is a unidirectional predictive association from A to B or from B to A, a bidirectional association / common driver, or no direct predictive association. Record the inferred directionality.

[0041] Perform statistical correlation / dependence strength calculation: calculate the correlation strength value between time series A and time series B at synchronization or different time lags. The calculation can adopt methods such as cross-correlation function peak calculation or mutual information calculation.

[0042] Step 3: Comprehensive determination of edge attributes: Fusion logic: Design a set of decision rules to determine whether to add or update an edge connecting node A and node B in the graph based on the directional evidence and strength values ​​obtained in step 2.

[0043] Rule example: Strong association rule: If the causal inference shows a clear direction and the association strength value exceeds the "high confidence threshold", a strong data-driven edge is created or updated, and the weight is positively correlated with the strength value.

[0044] Moderate association rule: If there is only evidence of association strength and the strength exceeds the "medium confidence threshold", an undirected data-driven edge is created or updated with a weight related to the strength.

[0045] Weak or no association rule: If the evidence is insufficient or the strength is below a threshold, do not create or reduce the weight of an existing edge.

[0046] Weight sign determination: The sign of the edge weight (positive / negative) is determined based on the sign of the original sequence correlation coefficient or preset domain knowledge.

[0047] User tag-driven semantic link learner: Input: A stream of user-generated meaning-association annotation records, including timestamps, source node identifiers (behavior / feeling nodes), a list of target node identifiers (target nodes), and optional contextual information.

[0048] Core processing flow (event-driven): Step 1: Identify associated pairs: For each labeled event, determine the source node and target node pair.

[0049] Step 2: Initialize or find edges: Find the semantic edge connecting this pair of nodes in the graph. If none exists, create one and assign a basic weight value.

[0050] Step 3: Calculate weight increment: Set a default weight increase value.

[0051] Adjust the weight increment according to the contextual adjustment factor calculation logic: The logic considers the contextual information of the marked event, such as the sentiment intensity of the user's annotation, the importance level of the marked situation, or the time interval since the last marking of the same connection, to calculate a contextual adjustment factor.

[0052] The final weight increment is equal to the default increase value multiplied by the context adjustment factor.

[0053] Step 4: Update weights: Add the calculated weight increment to the current weight of the edge, and optionally apply a maximum weight limit.

[0054] Step 5: Processing user negative feedback: If the system receives a clear negative evaluation from the user on the feedback generated based on this edge, a punitive weight decay factor (less than 1) is applied to reduce the weight of the edge, and a minimum weight limit can be set.

[0055] Network integration and dynamic maintenance unit: Integration: Store data-driven association edges and semantic connection edges in a unified graph data structure, and distinguish their sources through edge attributes.

[0056] Dynamic maintenance logic (including at least one of the following): Weight decay: A time-dependent decay function is periodically applied to data-driven association edges, reducing the weight of edges that have not been supported by new data for a long time. Semantic connection edges are mainly updated based on user labeling and feedback.

[0057] Node attribute update: Update the attribute value of the corresponding node in the graph with the latest data in real time or quasi-real time.

[0058] Topology adjustment: respond to user editing operations on meaningful anchors, and synchronously add, delete, and modify nodes and related edges in the graph.

[0059] 3. The meaning-oriented feedback and reflection engine is responsible for querying the meaning network model at the right time, generating personalized feedback, and stimulating users' intrinsic motivation.

[0060] Scenario and trigger monitoring: Manage trigger logic using state machines or rule engines.

[0061] Continuously monitor predefined trigger conditions, which include at least: the user completing a key health behavior event, the user submitting a subjective feeling mark, a significant change in the state related to the user's goal, reaching a preset periodic reflection time point, or receiving an explicit query request from the user.

[0062] Personalized meaning network query and analysis: Input the context information of the triggering event and the user's personalized meaning network model.

[0063] Core processing flow: Step 1: Determine query parameters: Based on the trigger type, determine the starting node set and target node set for graph traversal, which are usually the user's target nodes.

[0064] Step 2: Perform constrained graph traversal: Initialize the queue to be visited, maintain the set of visited nodes and the current path information.

[0065] Iteratively expand the path, performing pruning and constraint checking during the expansion process: checking whether the current path length exceeds the maximum depth limit; checking whether the current path cumulative weight meets the minimum path weight threshold; performing loop detection to avoid infinite loops.

[0066] If the end node of the path belongs to the target node set, it is recorded as a valid path.

[0067] Continue iterating until the termination condition is met.

[0068] Step 3: Path Evaluation and Screening: For all valid paths found, a pre-defined path evaluation function is applied to calculate a comprehensive score for each path. The path evaluation function aims to quantify the overall strength, reliability, and relevance of the association represented by a path to the current context. Its calculation principle integrates at least the following information: a) Cumulative edge weight of a path: This is calculated based on the cumulative weights of all edges along the path. For weights representing association strength (e.g., ranging from 0 to 1), this can be calculated using weight multiplication, weighted summation, or the minimum edge weight along the path. The specific calculation method is pre-defined.

[0069] Weight processing: When performing cumulative calculations, consider the edge type, whether it is data-driven or semantic, and the direction, positive or negative. For example, you can calculate positive and negative path strengths separately, or convert negative weights into penalty terms.

[0070] b) Path length factor: Shorter, more direct paths between nodes represent stronger connections or influences.

[0071] Calculation method: Introduce a factor that is inversely proportional to the path length.

[0072] c) Node recent activity or importance factor: If the nodes involved in the path are recently active, for example, the relevant objective data has changed significantly recently, or the user has recently marked related behaviors / feelings, or the node itself is given a high importance, then the relevance of the path may be higher.

[0073] Calculation method: Each node on the path is assigned an activity / importance score, which can be calculated based on the node's latest update time, change magnitude, user settings, etc., and these scores are integrated into the total score of the path by summing, averaging, or weighted average.

[0074] Preferably, d) edge confidence or evidence strength factor: If the meaning network model can provide a confidence score for the learning result of each edge, for example, based on the amount of data supporting the association, the statistical significance p-value, or user feedback evaluation, then the path with higher confidence is more reliable.

[0075] Calculation method: The confidence scores of each edge on the path are integrated into the total score by multiplication, minimum value, or weighted average.

[0076] Comprehensive Score Calculation: The final comprehensive path score is the quantified result of the above factors through a preset weighted sum or product method: Path Score = f(Cumulative Edge Weight, Path Length Factor, Node Activity Factor, [Edge Confidence Factor]). The specific form of function f is system-preset or configurable.

[0077] Feedback content generation and personalization: The key path list and its attributes filtered out in the previous step, as well as a structured feedback strategy library.

[0078] Core processing flow: Step 1: Match feedback strategy / template: Analyze the characteristics of the critical path, such as the starting point, end point, main intermediate node types, overall correlation trends, etc.

[0079] These path features are matched with the rule conditions in the feedback policy library to find the most applicable feedback template identifier.

[0080] Step 2: Content filling and structuring: The feedback strategy library stores a natural language expression framework for realizing a variety of preset meaning-oriented functions, including confirmation and reinforcement of behavioral meaning, attribution and understanding of the source of feelings, feedback related to goal progress, and switching of meaning perspectives in response to challenges.

[0081] The specific names or labels of the nodes involved in the path are read from the graph and accurately filled into the reserved positions of the selected template to form structured feedback content.

[0082] Preferably, step three: apply template-based filling or more complex natural language generation technology to convert structured content into fluent and natural text.

[0083] Based on learned user communication preferences such as historical feedback or user settings, the tone, style, and level of detail of the text are personalized.

[0084] User feedback processing and learning: Input the effectiveness or relevance evaluation data of feedback information provided by users through interactive controls.

[0085] Core processing flow: Step 1: Feedback attribution: Link user comments back to the specific network paths and edges that generated the feedback.

[0086] Step 2: Calculate the adjustment signal: Weight adjustment logic: If a user repeatedly rates an edge-based feedback as "irrelevant," a negative weight adjustment is calculated and applied to the edge's weight. If the user rates it as "very useful," a positive adjustment is calculated. The size of the adjustment is related to the strength or consistency of the ratings.

[0087] Strategy optimization logic: Calculate the average evaluation scores of different feedback templates or strategies, periodically update the selection priority of each strategy in the strategy library, and give priority to strategies with better historical evaluations.

[0088] Step 3: Apply adjustments: Update edge weights in the meaning network model, or adjust parameters in the feedback strategy library.

[0089] 4. The user interaction and visualization module is responsible for all interactions between users and the system: it is used for users to authorize data collection, set meaning anchors, mark meaning associations, and receive feedback information. It presents meaning-oriented feedback in a non-intrusive manner, provides interactive controls for user feedback evaluation, and provides user-configurable privacy setting options.

[0090] As a preference, the fifth model interpretability subsystem module is designed to enhance system transparency and user trust. It is configured to respond to user queries or specific system triggers.

[0091] The functions include at least one of the following: Generate a summary of the evidence for a specific edge in the meaning network model, displaying the quantitative metrics supporting the association and the associated user tagging history. For a specific feedback generation process, visualize the key network paths it relies on or provide a simplified explanation of its generation logic.

[0092] In one embodiment, coping with work stress and maintaining an exercise habit: User A: A software engineer who has been under a lot of work pressure recently and often needs to work overtime.

[0093] Health goals set (meaning anchors): Goal_ReduceStress (reduce work stress); Goal_MaintainExercise (maintain the habit of running at least 3 times a week); Goal_BetterSleep (improve sleep quality); Authorized data sources: smartwatches (heart rate, HRV, sleep, activity), mobile phones (screen time, location, calendar), in-app subjective feeling tagging interface, and meaning association tagging interface.

[0094] 2. Initial meaningful network construction (assuming it has been running for a while): Data driven associations, some examples: Long hours of work (calendar / screen time) --- [Negative, Strong] --> Sleep duration; Sleep duration --- [positive, moderate] --> next day HRV (RMSSD); Next day HRV (RMSSD) --- [positive, moderate] --> subjective energy score; Running behavior --- [positive, strong] --> short-term HRV increase --- [follow-up] --> long-term resting heart rate decrease trend; Running behavior --- [positive, moderate] --> subjective stress relief (occasional marker); High-intensity workdays (calendar / screen time) --- [Negative, Moderate] --> Probability of running behavior; Semantic connections (partial examples, from user tags): Running behavior --- [positive, strong] --> Goal_MaintainExercise; Running behavior --- [Positive, weak] --> Goal_ReduceStress (users occasionally mark feeling less stressed after running); Getting enough sleep --- [Positive, Strong] --> Goal_BetterSleep; 3. Specific scenario triggers and system responses (Wednesday night): Scenario: User A had a stressful day and worked overtime until 9 o'clock in the evening before returning home. He felt tired and missed his planned run.

[0095] Data collection: The calendar shows that the day is packed with meetings.

[0096] Screen time shows that work apps are used for a longer period of time.

[0097] The smartwatch recorded low daytime HRV and the activity level did not meet the standard.

[0098] Users marked their subjective feelings in the app: "Low energy (2 / 5)" and "High stress (4 / 5)".

[0099] Triggering event: User submits subjective feeling mark.

[0100] Meaningful network query and analysis (performed by the feedback engine): Source nodes: subjective feeling (low energy), subjective feeling (high stress), behavior node (missed exercise), objective data (low HRV), objective data (high workload).

[0101] Target Node: The target node of all users.

[0102] Path analysis found that: Path 1: High Workload -> Low HRV -> Low Energy (strong data-driven association) Path 2: Behavior conflicts with goal: High workload -> Missed exercise -> Negative connection to Goal_MaintainExercise Strong semantic connection Path 3 Potential Positive Connections: Searching for behavioral nodes that are positively connected to Goal_ReduceStress, we found that running behavior and adequate sleep are historically strongly associated.

[0103] Feedback generation, strategy selection and content filling: Matching strategy: Identify users experiencing "goal frustration" and "negative feelings," with clear objective data correlations (workload -> physiological indicators -> feelings), and a conflict between behavior and goals. Select a feedback strategy focused on "understanding attribution + meaning reconstruction / perspective switching."

[0104] Generate feedback text: Attribution and empathy: "I understand you're feeling low on energy and stressed today. Data suggests your high workload today may have impacted your physiological state, with lower HRV, which is often associated with feelings of decreased energy." Connecting a behavior to a goal: "Noticed that you missed your planned run today, which deviated from your goal of maintaining your exercise habit (Goal_MaintainExercise)." Reframing and Positive Perspective: Perhaps prioritizing rest tonight would be a good way to restore balance now and support your long-term goals? If the model learns that a user has improved their sleep in the past by using a specific relaxation technique after a similar overtime job, "Data shows that your sleep quality improved when you tried [specific relaxation technique, such as reading before bed] in similar situations in the past. Consider trying it tonight?" User interaction: The user receives this push notification or card in the app.

[0105] Feedback rating: Users can rate this feedback as "helpful", "average", or "irrelevant".

[0106] Subsequent tagging: If the user adopts the suggestions, such as ensuring sleep, the positive feelings will be marked the next day and may be associated with "rest" or "happiness", further strengthening the relevant network connection.

[0107] 4. System learning and iteration: User-rated feedback: If the user rates the feedback as "helpful," the system will slightly increase the confidence of the network path based on which the feedback was generated, or prioritize it for future selection. If the user rates it as "irrelevant," the confidence of the relevant path will be reduced.

[0108] Model update: The system periodically uses new objective data and labeled data to retrain or fine-tune the meaning network model so that the association relationship more accurately reflects the user's latest status and cognition.

[0109] The methods and systems described herein can be implemented using a computing device equipped with a processor and memory. The memory stores computer-executable instructions, and the processor executes these instructions to implement the functions of the aforementioned modules. The computing device can be a user's mobile terminal, an edge computing device, a cloud server, or a combination thereof. The modules of the system can communicate and exchange data via a network.

[0110] This specific implementation method elaborates on the internal operating logic of each main functional module of the system, and is intended to enable those skilled in the art to understand and implement the detailed basis and instructions. It should be emphasized that the above description constitutes a specific and preferred embodiment, but the concept of the present invention is not limited to this. Any equivalent transformation, modification or improvement based on the core spirit of the present invention, which does not deviate from the technical principles and scope disclosed in this specification, should be considered to fall within the scope of protection required by the present invention as long as it can achieve the same or similar technical effects.

Claims

1. An intelligent health status analysis method based on artificial intelligence, characterized in that: The method comprises: A: Acquire the target user's time-stamped physiological data, behavioral data, and environmental data, and receive the user's input meaning anchor data representing their health-related goals through user interaction; B: obtaining, through the user interaction, semantic association tag data input by the user, wherein the tag data explicitly records the semantic association between a specific behavioral event or subjective feeling state determined by the user and the semantic anchor data; C: Based on all acquired data, construct and periodically or event-driven dynamically maintain a personalized meaning network model for the target user; D: When a preset trigger logic is met, performing network path analysis on the personalized meaning network model to identify a valid association path that meets preset conditions, starting from a source node corresponding to the current user state or behavior and reaching a target node corresponding to the meaning anchor data; E: Based on the attributes of the identified valid association paths, a feedback strategy library is invoked and feedback generation rules are matched to generate personalized and meaning-oriented feedback information. The feedback information is intended to reveal to the user the specific connection between their current experience and their personal meaning anchors, as learned by the model. F: Through the user interaction, presenting the generated feedback information to the target user in a preset format.

2. The method for intelligent health status analysis based on artificial intelligence according to claim 1, characterized in that: The personalized meaning network model is stored and represented using a graph data structure, wherein the nodes of the graph include objective data feature nodes, subjective feeling nodes, behavior nodes, and target nodes generated according to the data classification, and the edges of the graph represent the weighted associations learned between the nodes; the weights and types of the edges are determined by combining the following steps: C1: Perform data-driven correlation analysis on the time series corresponding to the objective data, subjective feelings, and behavioral data to determine the statistical correlation strength and direction between nodes; C2: Based on the meaning association tag data, establish or strengthen the semantic connection weight between the behavior node or feeling node and the target node.

3. The method for intelligent health status analysis based on artificial intelligence according to claim 1 or 2, characterized in that: The execution of data-driven association analysis includes: preprocessing the multi-dimensional user status data to extract multiple quantified feature time series; selecting at least one algorithm from a preset time series causal inference algorithm set to process the feature time series to identify node pairs with predictive leading relationships; selecting at least one algorithm from a preset statistical correlation or dependency measurement algorithm set to calculate the correlation strength between the feature time series; and combining the output results of the time series causal inference algorithm and the correlation or dependency measurement algorithm to determine the existence, applicable direction, and quantified weight value of data-driven association edges according to preset fusion rules.

4. The method for intelligent health status analysis based on artificial intelligence according to any one of claims 1 to 3, characterized in that: The establishment or strengthening of the semantic connection weight between the behavior node or the feeling node and the target node includes: parsing the meaning association tag data, extracting the behavior node or the feeling node associated in each tagging event as the source node, and the target node as the target node; for each identified association pair, searching or creating the corresponding semantic connection edge in the meaning network model; applying at least one of the event frequency-based cumulative update rule and the context-based weight adjustment rule to calculate the updated weight of the semantic connection edge.

5. The method for intelligent health status analysis based on artificial intelligence according to any one of claims 1 to 4, characterized in that: The dynamic maintenance of the personalized meaning network model includes performing at least one of the following processes: periodic edge weight decay processing, reducing the weight of the associated edges that have not obtained new data-driven evidence or user tag support within a preset time window according to a preset decay function; Node status update processing: updating the attribute values ​​of objective data nodes, subjective feeling nodes, or behavior nodes based on the latest acquired data; The network topology structure is dynamically adjusted in response to the user's editing operation on the target node, and the adjustment includes adding new nodes and deleting invalid nodes and their associated edges.

6. The method for intelligent health status analysis based on artificial intelligence according to any one of claims 1 to 5, characterized in that: The execution of network path analysis includes: determining one or more source nodes associated with the current triggering event or user status; setting the target of the path search to the target node corresponding to the meaning anchor point data; using a graph traversal algorithm configured with a maximum search depth limit or a minimum path weight threshold, starting from the source node, searching for all valid paths to the target node in the meaning network model; for each valid path searched, calculating its comprehensive weight or significance score according to a preset path evaluation function, wherein the evaluation function at least integrates the weight of each edge on the path, the path length factor, and the recent activity information of the nodes involved in the path.

7. The method for intelligent health status analysis based on artificial intelligence according to any one of claims 1 to 6, characterized in that: The generation of personalized and meaning-oriented feedback information includes: matching a corresponding feedback template identifier from a structured feedback strategy library based on the attributes of the highest-weighted effective path identified by the network path analysis; the feedback strategy library pre-stores a natural language expression framework for implementing various meaning-oriented functions, including behavioral significance confirmation and reinforcement, attribution and understanding of the source of feelings, feedback related to goal progress, and switching of meaning perspectives for challenge response; and filling specific node information from the attributes of the identified highest-weighted effective path into the matched feedback template to generate preliminary feedback text.

8. The method for intelligent health status analysis based on artificial intelligence according to any one of claims 1 to 7, characterized in that: The method further includes: providing the user with interactive controls for evaluating the effectiveness or relevance of the feedback while or after presenting the feedback information; processing and analyzing the collected evaluation data through a configured feedback learning module; and mapping the analysis results into weight adjustment signals for relevant edges in the personalized meaning network model, or for updating the priorities or triggering conditions of the strategies in the feedback strategy library to achieve adaptive optimization of the system.

9. The health status intelligent analysis method system based on artificial intelligence is characterized by: The system is configured to perform the method according to any one of claims 1 to 8, and the system comprises: Multi-dimensional data acquisition interface module for interacting with external sensors and user input devices; A personalized meaning network modeling engine, including a data-driven association learner and a semantic connection learner, and equipped with a dynamic maintenance mechanism; Meaning-oriented feedback engine, including trigger logic judgment unit, network path analyzer and feedback generator; User interaction module, used to present information, obtain user tags and feedback; Feedback learning and model optimization module is used to achieve closed-loop adaptive adjustment.

10. The artificial intelligence-based health status intelligent analysis system according to claim 9, characterized in that: The system also includes: constructing a model interpretability submodule, and when the user requests or the system determines that it is necessary, the model interpretability submodule is configured to perform at least one of the following operations: generating an evidence summary about a specific association edge based on the meaning network model and related original data fragments, the summary including the number of data points supporting the association, the association strength index, and the applicable user tag record; for a certain generated meaning-oriented feedback, visualizing the key network path on which it is based or explaining its generation logic.

Images