A language interactive health guidance method and system

Abstract

The application is suitable for the technical field of health guidance, and particularly relates to a language interactive health guidance method and system, the method comprising: constructing a health dictionary of food materials, wherein the health dictionary is composed of a food material item, an intake amount item and a score item; receiving disease course data uploaded by a user, and adjusting the score item; editing test data of ingested food materials, and scoring by using the health dictionary to obtain a health score, which is written into the test data; generating a training set, and training a pre-created diet behavior scoring model; and collecting real-time video data by using a camera pre-integrated in a dining area of the user. The application can provide reliable behavior reference for health management of the user by selecting a forerunner, improve the success rate of health guidance, enhance the interestingness of participation, realize the change from passive reminding to active management, and effectively guarantee the long-term stable execution of health guidance.

Description

Technical Field

[0001] This invention relates to the field of health guidance technology, and in particular to a language-interactive health guidance method and system. Background Technology

[0002] Health guidance through language interaction refers to a human-computer interaction method that uses voice or natural language dialogue to conduct two-way communication with users, thereby achieving health monitoring, behavior reminders, risk intervention, and habit management.

[0003] In real life, interactive guidance is mostly implemented using preset templates or question-and-answer structures. For example, it may be based on fixed sentence patterns for daily broadcasts or use simple questions and answers to prompt "Have you completed your exercise?" or "Have you controlled your diet?" Its core logic is usually based on fixed rules and lacks comprehensive analysis of user behavior patterns, current situational changes, and individual health differences. Therefore, in the long-term use, the voice content that users receive every day is often highly repetitive, and its semantic structure, expression, and tone of voice are relatively monotonous. Users are prone to "cognitive desensitization" of the prompts, that is, although they can hear the broadcast content, the brain's assessment of its importance gradually decreases, resulting in interactive fatigue.

[0004] Therefore, "how to combine user diet for interactive guidance" is the technical problem that this invention needs to solve. Summary of the Invention

[0005] The purpose of this invention is to provide a language-interactive health guidance method and system to solve the problem of "how to combine user diet for interactive guidance" mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A language-interactive health guidance method, the method comprising:

[0008] A health dictionary for food ingredients is constructed, which consists of food ingredient items, intake items, and scoring items. The system receives disease course data uploaded by users, adjusts the scoring items, edits test data of food ingredient intake, and uses the health dictionary to score the food ingredients to obtain a health score. This score is then written into the test data to generate a training set and trains a pre-created dietary behavior scoring model.

[0009] Using cameras pre-integrated in the user's dining area, real-time video data is collected, input into the dietary behavior scoring model, and a corresponding health score is output and dynamically updated. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted.

[0010] Using pre-created risk assessment rules, determine whether there are risk characteristics in the preference features. If so, generate voice information and send it to the speaker devices in the dining area for playback, activate the user's corresponding health monitoring device, generate a usage reminder, and send it to the preset terminal.

[0011] Create an interval period that corresponds one-to-one with the health monitoring device. After the interval period countdown ends, determine whether the health monitoring device has generated vital sign parameters. If not, issue the usage reminder again. If so, cluster all users into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. In the same category, select the users' corresponding pioneers, activate two-way invitations, establish friend relationships, and send health reminders in parallel.

[0012] Furthermore, the step of constructing a health dictionary for food ingredients, wherein the health dictionary consists of food ingredient items, intake items, and rating items, and receiving disease history data uploaded by users includes:

[0013] Tags generated from disease course data are inserted into the health dictionary to create a personalized dictionary, where each user has a corresponding personalized dictionary;

[0014] The scoring items are fine-tuned using the vital signs parameters.

[0015] Furthermore, the step of using cameras pre-integrated in the user's dining area to collect real-time video data and inputting it into the dietary behavior scoring model includes:

[0016] Configure edge devices in the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices;

[0017] The personalized dictionary is uploaded to an edge device for storage.

[0018] Furthermore, the step of obtaining the corresponding health score from the output and dynamically updating it includes:

[0019] Establish the correspondence between time and health score, and plot the trend graph with time as the horizontal axis and the corresponding health score as the vertical axis;

[0020] The trend chart is divided into several segments, wherein the segments include at least an upward segment and a downward segment, a monitoring report is generated, and the report is sent to a preset terminal.

[0021] Furthermore, the step of generating voice information and sending it to the speaker devices in the dining area for playback includes:

[0022] Audio data of users is collected using audio pickup devices pre-installed in the dining area;

[0023] An audio interaction mechanism is embedded into the audio pickup device, and activation conditions for the audio pickup device are set.

[0024] Furthermore, the step of clustering all users into several categories includes:

[0025] From the categories mentioned above, a standard category is selected, a health sharing platform is built, and the preference characteristics of users in the standard category are published on the health sharing platform.

[0026] Obtain sharing authorization from standard users, capture snapshots of the dining area from real-time video data, and grant viewing access.

[0027] Furthermore, the system includes:

[0028] The module is used to build a health dictionary of food ingredients, which consists of food ingredient items, intake items, and scoring items. It receives disease course data uploaded by users, adjusts the scoring items, edits the test data of food ingredient intake, and uses the health dictionary to score the food ingredients to obtain a health score, which is written into the test data to generate a training set and train a pre-created dietary behavior scoring model.

[0029] The extraction module is used to collect real-time video data using cameras pre-integrated in the user's dining area, input the data into the dietary behavior scoring model, output the corresponding health score, and dynamically update it. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted.

[0030] The judgment module is used to determine whether there are risk characteristics in the preference characteristics using pre-created risk assessment rules. If so, it generates voice information and sends it to the speaker devices in the dining area for playback, activates the user's corresponding health monitoring device, generates a usage reminder, and sends it to the preset terminal.

[0031] The module is used to create an interval duration that corresponds one-to-one with the health monitoring device. After the interval duration countdown ends, it is determined whether the health monitoring device has generated vital sign parameters. If not, the usage reminder is issued again. If so, all users are clustered into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. Within the same category, the users' corresponding pioneers are selected, a two-way invitation is activated, a friend relationship is established, and health reminders are sent in parallel.

[0032] Furthermore, the building module includes:

[0033] An insertion unit is used to insert tags generated from disease course data into the health dictionary to generate a personalized dictionary, where each user corresponds to a personalized dictionary.

[0034] The fine-tuning unit is used to fine-tune the scoring items using the vital sign parameters.

[0035] Furthermore, the extraction module includes:

[0036] A configuration unit is used to configure edge devices within the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices;

[0037] An upload unit is used to upload the personalized dictionary to an edge device for storage;

[0038] Establish a unit to establish the correspondence between time and health score, and plot the trend graph with time as the horizontal axis and the corresponding health score as the vertical axis;

[0039] The segmentation unit is used to segment the trend graph into several segments, wherein the segments include at least an upward segment and a downward segment, generate a monitoring report, and send it to a preset terminal.

[0040] Furthermore, the determination module includes:

[0041] The acquisition unit is used to collect users' audio data using sound pickup devices pre-installed in the dining area;

[0042] The setting unit is used to embed a language interaction mechanism into the sound pickup device and set the activation conditions of the sound pickup device.

[0043] Compared with the prior art, the beneficial effects of the present invention are:

[0044] By creating a health dictionary, users' nutritional intake can be quantified, providing a data foundation for health guidance. Simultaneously, adjusting scoring items enables personalized health management, further improving the accuracy of health guidance. Real-time video data from the dining area, combined with a dietary behavior scoring model, allows for real-time scoring of user intake without user intervention, significantly reducing interaction fatigue and greatly improving long-term satisfaction with interactive guidance. Selecting pioneers provides reliable behavioral references for users' health management, increasing the success rate of health guidance, enhancing engagement, and shifting from passive reminders to proactive management, effectively ensuring the long-term stable implementation of health guidance. Attached Figure Description

[0045] Figure 1A flowchart of the language-interactive health guidance method provided in this embodiment of the invention;

[0046] Figure 2 This is a first sub-flowchart of the language interactive health guidance method provided in an embodiment of the present invention;

[0047] Figure 3 This is a second sub-flowchart of the language-interactive health guidance method provided in an embodiment of the present invention;

[0048] Figure 4 This is a third sub-flow diagram of the language interactive health guidance method provided in an embodiment of the present invention;

[0049] Figure 5 This is a fourth sub-flowchart of the language-interactive health guidance method provided in an embodiment of the present invention;

[0050] Figure 6 This is a block diagram of the language-interactive health guidance system provided in an embodiment of the present invention;

[0051] Figure 7 A block diagram illustrating the components of the building modules in the language-interactive health guidance system provided in this embodiment of the invention;

[0052] Figure 8 A block diagram illustrating the components of the extraction module in the language interactive health guidance system provided in this embodiment of the invention;

[0053] Figure 9 This is a block diagram of the judgment module in the language interactive health guidance system provided in an embodiment of the present invention;

[0054] Figure 10 This is a block diagram of the creation module in the language interactive health guidance system provided in an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0056] In Example 1, Figure 1 The implementation flow of the language interactive health guidance method provided in this embodiment of the invention is illustrated below, and will be described in detail below:

[0057] S100: Construct a health dictionary for food ingredients, which consists of food ingredient items, intake items, and rating items. Receive disease course data uploaded by users, adjust the rating items, edit test data of ingested food ingredients, and use the health dictionary to score the food ingredients to obtain a health score, which is written into the test data to generate a training set and train a pre-created dietary behavior scoring model.

[0058] A health dictionary is constructed, consisting of food items, intake items, and rating items. This health dictionary is essentially a set of rating maps that scores a user's food intake during meals. The system receives medical history data uploaded by users to a pre-set platform (a health management platform). This data can be simply understood as past medical records. The system analyzes this data to extract disease type, stage, complications, and trends in key indicators, and adjusts the rating items in the health dictionary accordingly. For example, if a user has gallstones, the ratings for foods such as fried chicken, French fries, fried pastries, fatty meat, and pork belly should be reduced. Lower food ratings indicate a more adverse health impact on the user.

[0059] Test data is created to simulate the actual food intake of users. Essentially, it's manually or program-generated sample data primarily used for model training, and not collected from real users. Humans score the test data according to a health dictionary, obtaining a health score. A higher health score indicates that the combination and quantity of food consumed by the user are more in line with nutritional balance standards. The health scores are written into the corresponding fields of the test data for data annotation. The test data and corresponding health scores are integrated to generate a training set. This training set is used to train the dietary behavior scoring model, which is built based on a deep learning algorithm. This model can learn the non-linear relationship between food types, intake amounts, and health scores, enabling rapid assessment of the user's health score during meals.

[0060] S200: Using cameras pre-integrated in the user's dining area, real-time video data is collected, input into the dietary behavior scoring model, and a corresponding health score is output and dynamically updated. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted.

[0061] Cameras are deployed in the user's dining area to collect real-time video data during the meal. These cameras are installed by the user and can be turned on only during mealtimes or continuously. The collected real-time video data is input into a dietary behavior scoring model, outputting a health score. It's important to note that this health score reflects the user's current dietary status and changes as the meal progresses. During the meal, the health score is dynamically updated. When the health score falls below a threshold, image recognition and semantic analysis algorithms are used to convert the camera-captured dining scene into a text description. This text description includes the types and quantities of food on the plate, their combination, eating speed, and cooking methods. User dietary preferences are then feature-extracted, revealing preference characteristics. These characteristics include a preference for high-oil, high-salt, and high-sugar foods, low intake of vegetables and fruits, and repeated selection of certain food types. For example, a user's preference characteristics might be: a liking for fried foods, desserts, and sugary drinks, and a low intake of vegetables or fruits.

[0062] In this embodiment, the camera can not only monitor the food on the plates in the dining area, but also, with the user's authorization, collect and analyze the user's behavior during the dining process. Behavioral characteristics include: chewing speed, whether a mobile phone is used, and dining time.

[0063] S300: Using pre-created risk assessment rules, determine whether there are risk characteristics in the preference features. If so, generate voice information and send it to the speaker device in the dining area for playback, activate the user's corresponding health monitoring device, generate a usage reminder, and send it to the preset terminal.

[0064] A risk assessment rule is created to determine whether certain behaviors pose a health risk. For example, a risk assessment rule might state: consuming cured meat products for three consecutive days is considered a high-salt, high-fat risk behavior, indicating a potential health risk. If a user's preference characteristics match the risk characteristics, it suggests a potential health problem. A voice message is then generated and played through speakers in the dining area, such as televisions, radios, or the user's mobile device. For example, a voice message might read: "You have consumed cured meat products for three consecutive days, posing a high-salt, high-fat risk. It is recommended that you avoid cured meats and choose other foods instead."

[0065] Upon detecting risk characteristics, the system activates the user's corresponding health monitoring devices. These devices are various equipment installed in the user's home for daily health monitoring, such as blood glucose meters, pulse oximeters, blood pressure monitors, and electrocardiogram (ECG) monitors. A usage reminder is sent to the user's pre-set terminal, reminding the user to take measurements on time. It's important to note that not all of the user's health monitoring devices are activated; only those corresponding to the risk characteristics are activated. Continuing with the example above, if the risk characteristics are high salt and high fat, the corresponding health monitoring devices would be a blood pressure monitor and a blood glucose meter.

[0066] S400: Create an interval duration that corresponds one-to-one with the health monitoring device. After the interval duration countdown ends, determine whether the health monitoring device has generated vital sign parameters. If not, issue the usage reminder again. If so, cluster all users into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. In the same category, select the user's corresponding pioneer, activate the two-way invitation, establish a friend relationship, and send health reminders in parallel.

[0067] Each health monitoring device is assigned a corresponding interval, which refers to the time interval required for measurement after a meal. After the user finishes eating, a countdown begins for each interval. If the health monitoring device has not generated the latest vital signs parameters when the countdown ends, it means the user has not yet taken measurements. A reminder is then sent to the preset terminal to remind the user to take measurements promptly and guide them to complete the necessary monitoring. If the health monitoring device generates vital signs parameters before the countdown ends, all users are clustered into several categories based on these parameters. Each category corresponds to a group of users with similar dietary preferences, including categories such as: light diet, oily diet, and long-term heavy oily diet. Within each category, a "pioneer" is identified for each user. Pioneers are individuals who have accumulated more experience with a particular health behavior or lifestyle habit over a longer period; in other words, pioneers have a longer history of unhealthy eating habits. For example, if users A and B both have a long-term diet high in oil and meat, and A's triglyceride level is 7.9 mmol / L, far exceeding the normal range, while B's triglyceride level is also high at 1.8 mmol / L, then A is B's pioneer. A two-way invitation is sent to both users, establishing a friend relationship. Simultaneously, targeted health tips are sent to both, including dietary adjustments, exercise suggestions, indicator monitoring, and mutual support reminders. The advantage of this approach is that it facilitates better dialogue between the two users, fostering a health management atmosphere of mutual supervision and encouragement. By leveraging the pioneer's life experience, it helps users gradually improve their unhealthy eating habits, enhancing the effectiveness of health guidance.

[0068] In Example 2, Figure 2The diagram shows the first sub-flowchart of the language-interactive health guidance method provided in this embodiment of the invention. The following details the steps of constructing a health dictionary of food ingredients, which consists of food ingredient items, intake items, and rating items, and receiving disease data uploaded by users:

[0069] S101: Insert tags generated from disease course data into the health dictionary to generate a personalized dictionary, where each user corresponds to a personalized dictionary.

[0070] Tags are generated using disease progression data and inserted into a health dictionary. These tags are displayed as notes or links. The health dictionary containing these tags is defined as a personalized dictionary, with each user having their own personalized dictionary. In this embodiment, by constructing a personalized dictionary, a personalized rating standard can be established for each user; in other words, different users may rate the same food ingredient differently.

[0071] S102: Fine-tune the scoring items using the vital signs parameters.

[0072] In addition to being based on disease course data, scoring criteria should also take into account physical signs and parameters.

[0073] In Example 3, Figure 3 The second sub-flowchart of the language-interactive health guidance method provided in this embodiment of the invention is shown. The following details the steps of using a camera pre-integrated in the user's dining area to collect real-time video data and inputting it into the dietary behavior scoring model:

[0074] S201: Configure edge devices in the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices.

[0075] Identify edge devices within the dining area, such as smart gateways or AI camera terminals. Deploy dietary behavior scoring models onto these edge devices and connect cameras and health monitoring devices to them for data interaction.

[0076] S202: Upload the personalized dictionary to the edge device for storage.

[0077] The user's personalized dictionary is uploaded to the edge device for storage, so that it can be directly accessed when performing dining video analysis, voice interaction or risk identification, reducing dependence on the cloud, reducing response latency and improving privacy protection capabilities.

[0078] In Example 4, Figure 3The second sub-flowchart of the language interactive health guidance method provided in this embodiment of the invention is shown. The steps of obtaining the corresponding health score from the output and dynamically updating it are described in detail below:

[0079] S203: Establish the correspondence between time and health score, and plot a trend graph with time as the horizontal axis and the corresponding health score as the vertical axis.

[0080] The health score is updated periodically, and the update time is recorded. Using this time point as the horizontal axis and the corresponding health score as the vertical axis, a trend graph is plotted. The trend graph is a time series curve of the health score change.

[0081] S204: Divide the trend graph into several segments, wherein the segments include at least an upward segment and a downward segment, generate a monitoring report, and send it to a preset terminal.

[0082] The system iterates through the inflection points of the curves in the trend chart, defines the intervals where the trend is consistently positive and the slope is greater than a set value as the rising segment, and defines the intervals where the trend is consistently negative and the slope is less than another set value as the falling segment. The changes in the health score in the rising or falling segment are written into a preset template to obtain a monitoring report, which is then sent to a preset terminal. The monitoring report is also a health trend analysis report, which can show the changes in the user's dietary habits over a certain period.

[0083] In Example 5, Figure 4 The diagram shows a third sub-flowchart of the language-interactive health guidance method provided in this embodiment of the invention. The steps of generating voice information and sending it to the speaker devices in the dining area for playback are described in detail below:

[0084] S301: Collects users' audio data using a pre-installed sound pickup device in the dining area.

[0085] S302: Embed a language interaction mechanism into the sound pickup device and set the activation conditions for the sound pickup device.

[0086] Environmental audio data is collected using audio pickup devices installed in the dining area. A local language interaction mechanism is embedded in these devices, enabling voice wake-up, semantic recognition, and command parsing capabilities. The audio pickup devices can be integrated into cameras or edge devices. The language interaction mechanism refers to the method of interacting with users through language using cameras, edge devices, and audio pickup devices. In this embodiment, the camera is mainly used to collect video data during the user's dining process. Therefore, the camera and audio pickup devices are not always powered on, but only during breakfast, lunch, and dinner periods. The activation condition is automatic activation at the predetermined time. However, the camera and audio pickup devices can also be manually set or activated via voice keywords.

[0087] In Example 6, Figure 5 The fourth sub-flowchart of the language interactive health guidance method provided in this embodiment of the invention is shown. The step of clustering all users into several categories is described in detail below:

[0088] S401: Select a standard category from the categories, build a health sharing platform, and publish the preference characteristics of standard category users to the health sharing platform.

[0089] Create a "standard class," which refers to users with superior dietary habits, consistently high health scores with minimal fluctuations, and low frequency of risk triggers. Simply put, standard class users have relatively reasonable dietary structures and excellent nutrient intake ratios. Build a health sharing platform, primarily to aggregate dietary behavior data from standard class users and facilitate health discussions among all users. With the authorization of standard class users, their corresponding preference characteristics will be published on the health sharing platform to provide reference for the dietary behaviors of other users.

[0090] S402: Obtain sharing authorization from standard users, capture snapshots of the dining area from real-time video data, and grant viewing permissions.

[0091] Initiate a sharing authorization request to standard users, capture a snapshot of the standard users' dining area, and grant access to view it, so as to realize the sharing, dissemination, and positive guidance of healthy demonstration samples.

[0092] Figure 6 This diagram illustrates the structural block diagram of a language-interactive health guidance system provided in an embodiment of the present invention. The language-interactive health guidance system 1 includes:

[0093] The construction module 11 is used to construct a health dictionary of food ingredients, which consists of food ingredient items, intake items and scoring items. It receives disease course data uploaded by users, adjusts the scoring items, edits the test data of food ingredients, and uses the health dictionary to score the food ingredients to obtain a health score, writes it into the test data, generates a training set, and trains a pre-created dietary behavior scoring model.

[0094] Extraction module 12 is used to collect real-time video data using cameras pre-integrated in the user's dining area, input the data into the dietary behavior scoring model, output the corresponding health score, and dynamically update it. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted.

[0095] The judgment module 13 is used to determine whether there are risk characteristics in the preference characteristics using the pre-created risk assessment rules. If so, it generates voice information and sends it to the speaker device in the dining area for playback, activates the health monitoring device corresponding to the user, generates a usage reminder, and sends it to the preset terminal.

[0096] Module 14 is used to create an interval duration that corresponds one-to-one with the health monitoring device. After the interval duration countdown ends, it is determined whether the health monitoring device has generated vital sign parameters. If not, the usage reminder is issued again. If so, all users are clustered into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. Within the same category, the user's corresponding pioneer is selected, a two-way invitation is activated, a friend relationship is established, and health reminders are sent in parallel.

[0097] Figure 7 This diagram illustrates the composition of a building module 11 in a language-interactive health guidance system provided in an embodiment of the present invention. The building module 11 includes:

[0098] The insertion unit 111 is used to insert tags generated from disease course data into the health dictionary to generate a personalized dictionary, wherein each user corresponds to a personalized dictionary.

[0099] The fine-tuning unit 112 is used to fine-tune the scoring item using the vital sign parameters.

[0100] Figure 8 This diagram illustrates the structural composition of the extraction module 12 in the language interactive health guidance system provided in an embodiment of the present invention. The extraction module 12 includes:

[0101] Configuration unit 121 is used to configure edge devices in the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices;

[0102] Upload unit 122 is used to upload the personalized dictionary to an edge device for storage;

[0103] Establish unit 123 to establish the correspondence between time and health score, and draw a trend graph with time as the horizontal axis and the corresponding health score as the vertical axis;

[0104] The segmentation unit 124 is used to segment the trend graph into several segments, wherein the segments include at least an upward segment and a downward segment, generate a monitoring report, and send it to a preset terminal.

[0105] Figure 9 This diagram illustrates the structural composition of the judgment module 13 in the language interactive health guidance system provided in an embodiment of the present invention. The judgment module 13 includes:

[0106] The acquisition unit 131 is used to acquire the user's audio data using a pickup device pre-installed in the dining area;

[0107] Setting unit 132 is used to embed a language interaction mechanism into the sound pickup device and set the activation conditions of the sound pickup device.

[0108] Figure 10 This diagram illustrates the structural composition of the creation module 14 in the language interactive health guidance system provided in an embodiment of the present invention. The creation module 14 includes:

[0109] The publishing unit 141 is used to select a standard category from the categories, build a health sharing platform, and publish the preference characteristics of standard category users to the health sharing platform.

[0110] Open unit 142 is used to obtain sharing authorization from standard users, capture snapshots of the dining area from real-time video data, and grant viewing permissions.

[0111] The construction module 11 is mainly used to complete step S100, the extraction module 12 is mainly used to complete step S200, the judgment module 13 is mainly used to complete step S300, and the creation module 14 is mainly used to complete step S400.

[0112] The insertion unit 111 is mainly used to complete step S101, and the fine-tuning unit 112 is mainly used to complete step S102.

[0113] The configuration unit 121 is mainly used to complete step S201, the upload unit 122 is mainly used to complete step S202, the establishment unit 123 is mainly used to complete step S203, and the splitting unit 124 is mainly used to complete step S204.

[0114] The acquisition unit 131 is mainly used to complete step S301, and the setting unit 132 is mainly used to complete step S302.

[0115] The publishing unit 141 is mainly used to complete step S401, and the opening unit 142 is mainly used to complete step S402.

[0116] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0117] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0118] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A language-interactive health guidance method, characterized in that, The method includes: A health dictionary for food ingredients is constructed, which consists of food ingredient items, intake items, and scoring items. The system receives disease course data uploaded by users, adjusts the scoring items, edits test data of food ingredient intake, and uses the health dictionary to score the food ingredients to obtain a health score. This score is then written into the test data to generate a training set and trains a pre-created dietary behavior scoring model. Using cameras pre-integrated in the user's dining area, real-time video data is collected, input into the dietary behavior scoring model, and a corresponding health score is output and dynamically updated. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted. Using pre-created risk assessment rules, determine whether there are risk characteristics in the preference features. If so, generate voice information and send it to the speaker devices in the dining area for playback, activate the user's corresponding health monitoring device, generate a usage reminder, and send it to the preset terminal. Create an interval period that corresponds one-to-one with the health monitoring device. After the interval period countdown ends, determine whether the health monitoring device has generated vital sign parameters. If not, issue the usage reminder again. If so, cluster all users into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. In the same category, select the users' corresponding pioneers, activate two-way invitations, establish friend relationships, and send health reminders in parallel.

2. The language-interactive health guidance method according to claim 1, characterized in that, The step of constructing a health dictionary for food ingredients, wherein the health dictionary consists of food ingredient items, intake items, and rating items, and receiving disease data uploaded by users includes: Tags generated from disease course data are inserted into the health dictionary to create a personalized dictionary, where each user has a corresponding personalized dictionary; The scoring items are fine-tuned using the vital signs parameters.

3. The language-interactive health guidance method according to claim 2, characterized in that, The step of using cameras pre-integrated in the user's dining area to collect real-time video data and inputting it into the dietary behavior scoring model includes: Configure edge devices in the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices; The personalized dictionary is uploaded to an edge device for storage.

4. The language-interactive health guidance method according to claim 1, characterized in that, The steps of obtaining the corresponding health score from the output and dynamically updating it include: Establish the correspondence between time and health score, and plot the trend graph with time as the horizontal axis and the corresponding health score as the vertical axis; The trend chart is divided into several segments, wherein the segments include at least an upward segment and a downward segment, a monitoring report is generated, and the report is sent to a preset terminal.

5. The language-interactive health guidance method according to claim 3, characterized in that, The step of generating voice information and sending it to the speaker equipment in the dining area for playback includes: Audio data of users is collected using audio pickup devices pre-installed in the dining area; An audio interaction mechanism is embedded into the audio pickup device, and activation conditions for the audio pickup device are set.

6. The language-interactive health guidance method according to claim 5, characterized in that, The step of clustering all users into several categories includes: From the categories mentioned above, a standard category is selected, a health sharing platform is built, and the preference characteristics of users in the standard category are published on the health sharing platform. Obtain sharing authorization from standard users, capture snapshots of the dining area from real-time video data, and grant viewing access.

7. A language-interactive health guidance system, characterized in that, The system includes: The module is used to build a health dictionary of food ingredients, which consists of food ingredient items, intake items, and scoring items. It receives disease course data uploaded by users, adjusts the scoring items, edits the test data of food ingredient intake, and uses the health dictionary to score the food ingredients to obtain a health score, which is written into the test data to generate a training set and train a pre-created dietary behavior scoring model. The extraction module is used to collect real-time video data using cameras pre-integrated in the user's dining area, input the data into the dietary behavior scoring model, output the corresponding health score, and dynamically update it. When the health score is less than a threshold, the real-time video data is described in text and preference features are extracted. The judgment module is used to determine whether there are risk characteristics in the preference characteristics using pre-created risk assessment rules. If so, it generates voice information and sends it to the speaker devices in the dining area for playback, activates the user's corresponding health monitoring device, generates a usage reminder, and sends it to the preset terminal. The module is used to create an interval duration that corresponds one-to-one with the health monitoring device. After the interval duration countdown ends, it is determined whether the health monitoring device has generated vital sign parameters. If not, the usage reminder is issued again. If so, all users are clustered into several categories, which include at least: light diet, oily diet, and long-term heavy oily diet. Within the same category, the users' corresponding pioneers are selected, a two-way invitation is activated, a friend relationship is established, and health reminders are sent in parallel.

8. The language-interactive health guidance system according to claim 7, characterized in that, The building module includes: An insertion unit is used to insert tags generated from disease course data into the health dictionary to generate a personalized dictionary, where each user corresponds to a personalized dictionary. The fine-tuning unit is used to fine-tune the scoring items using the vital sign parameters.

9. The language-interactive health guidance system according to claim 8, characterized in that, The extraction module includes: A configuration unit is used to configure edge devices within the dining area, deploy the dietary behavior scoring model to the edge devices, and connect the camera and health monitoring device to the edge devices; An upload unit is used to upload the personalized dictionary to an edge device for storage; Establish a unit to establish the correspondence between time and health score, and plot the trend graph with time as the horizontal axis and the corresponding health score as the vertical axis; The segmentation unit is used to segment the trend graph into several segments, wherein the segments include at least an upward segment and a downward segment, generate a monitoring report, and send it to a preset terminal.

10. The language-interactive health guidance system according to claim 9, characterized in that, The judgment module includes: The acquisition unit is used to collect users' audio data using sound pickup devices pre-installed in the dining area; The setting unit is used to embed a language interaction mechanism into the sound pickup device and set the activation conditions of the sound pickup device.

Images