Intelligent healthy diet recommendation system for severe obese patients

Abstract

The invention relates to the technical field of data association, and provides a healthy diet intelligent recommendation system for a severe obese patient, and the system comprises the steps: obtaining a healthy diet record and a plurality of pre-recommended foods of a target patient; obtaining a distinguishing contribution index according to the content and score of the nutrient substances; obtaining a reference weight according to the distinguishing contribution index; according to the distinguishing contribution index and the reference weight, obtaining a distinguishing contribution degree; obtaining the distinguishing degree according to the distinguishing contribution degree and the distinguishing contribution index; obtaining taste similarity according to the content of the nutrient substances and the distinguishing degree; obtaining representative food and a score of the representative food according to the score and the taste similarity; and according to the taste similarity between the representative food and the pre-recommended food and the score, performing food recommendation on the target patient. According to the invention, food recommendation is carried out through the score of the patient and the taste similarity between food, and the treatment experience of the patient is improved.

Description

Technical Field

[0001] This invention relates to the field of data association technology, specifically to a smart recommendation system for healthy eating for patients with severe obesity. Background Technology

[0002] Severe obesity refers to individuals whose weight significantly exceeds the normal range and whose BMI is 40 or higher. These patients often suffer from multiple health problems, such as diabetes, hypertension, sleep apnea, heart disease, and joint disorders. Severe obesity not only increases the risk of chronic diseases but also severely impacts patients' quality of life and survival expectations. For severely obese patients, healthy dietary management is crucial. A proper diet can help control weight, improve insulin sensitivity, and lower key health indicators such as blood lipids and blood sugar, thereby alleviating symptoms and preventing complications. Furthermore, scientific dietary control can improve patients' nutritional status, enhance immunity, and improve their quality of life. Because severely obese patients often have abnormal metabolism and energy expenditure, personalized dietary plans not only aid in weight loss but also ensure they receive adequate nutritional support during the weight loss process, avoiding the risk of malnutrition.

[0003] In the process of recommending healthy eating, the dietary preferences of severely obese patients may change over time, especially during treatment. As the patient's health condition improves or changes, they may gradually adapt to new eating habits, which will lead to changes in dietary preferences. A single, fixed healthy eating plan may not be able to meet the patient's dietary needs, and a more targeted eating plan needs to be developed. Summary of the Invention

[0004] This invention provides an intelligent health diet recommendation system for severely obese patients to address the problem that existing single, fixed health diet plans may not meet the dietary preferences of patients. The specific technical solution adopted is as follows: This invention proposes a smart dietary recommendation system for severely obese patients, the system comprising: Dietary Record and Pre-Recommended Food Acquisition Module: Acquires healthy dietary records of several severely obese patients; the healthy dietary records include the name and score of the food consumed at each meal during treatment, as well as the content of several nutrients contained in the food; severely obese patients who require targeted dietary plans are identified as target patients; acquires several pre-recommended foods for the target patients; Differentiating contribution index acquisition module: Based on the similarity relationship between the scores of severely obese patients on foods with similar nutrient content, the differentiating contribution index of each dietary record of severely obese patients to each nutrient is obtained. The discrimination analysis module: Based on the changes in the discrimination contribution index of severely obese patients to each nutrient across all dietary records, the reference weight of each dietary record for severely obese patients to each nutrient is obtained; based on the discrimination contribution index and reference weight of all dietary records for severely obese patients to each nutrient, the discrimination contribution degree of severely obese patients to each nutrient is obtained; based on the discrimination contribution degree of all severely obese patients to each nutrient and the discrimination contribution index of all dietary records for all severely obese patients to each nutrient, the discrimination degree of each nutrient is obtained. Representative food screening and scoring module: Based on the differences in nutrient content and the degree of differentiation of nutrients among foods, the taste similarity between foods is obtained; based on the scores of the pre-recommended foods and the taste similarity between the pre-recommended foods, the scores of the representative foods and the target patients on the representative foods are obtained. Recommended food module: Based on the similarity between the taste and scores of representative foods and pre-recommended foods, food recommendations are made for the target patients.

[0005] Furthermore, based on the similarity relationship between the ratings of foods with similar nutrient content by severely obese patients, the distinguishing contribution index of each dietary record of severely obese patients to each nutrient is obtained. The specific method for obtaining this index is as follows: In the formula, For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first The total number of dietary records for each severely obese patient; For the first The first severely obese patient The first in the food diary The content of each nutrient; For the first In addition to the first group of severely obese patients The first one outside of the diet record The first in the food diary The content of each nutrient; For the first The number of severely obese patients on the first A rating of each food record; For the first A severely obese patient, except for the first The first one outside of the diet record A rating of each food record; It is an exponential function with the natural constant as its base; It is an absolute value function.

[0006] Furthermore, the method for obtaining the reference weight of each dietary record for each nutrient for severely obese patients based on the variation of the differential contribution index of each nutrient to all dietary records across all dietary records includes: For any severely obese patient, the sequence formed by arranging the patient's distinguishing contribution index to any nutrient according to the order of the dietary records is denoted as the distinguishing sequence of that nutrient for that severely obese patient. Obtain all extreme values ​​of the discrimination sequence; for any discrimination contribution index in the discrimination sequence, record the maximum value of the extreme values ​​on both sides of the discrimination contribution index as the reference weight of the dietary record corresponding to the discrimination contribution index for that nutrient.

[0007] Furthermore, based on the distinguishing contribution index and reference weight of all dietary records of severely obese patients for each nutrient, the distinguishing contribution degree of severely obese patients for each nutrient is obtained. The specific method for obtaining this information is as follows: For any dietary record of any severely obese patient, the product of the reference weight of the dietary record for any nutrient and the distinguishing contribution index is recorded as the distinguishing index of the dietary record for that nutrient in the severely obese patient; the sum of the distinguishing indices of all dietary records of the severely obese patient for that nutrient is recorded as the degree of distinguishing contribution of the severely obese patient to that nutrient.

[0008] Furthermore, the degree of differentiation for each nutrient is obtained based on the differentiation contribution of all severely obese patients to each nutrient and the differentiation contribution index of all dietary records of all severely obese patients to each nutrient. The specific method for obtaining this information is as follows: In the formula, For the first The degree of differentiation among individual nutrients; The number of severely obese patients; For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first Among all the dietary records of the severely obese patient, the first The maximum value of the distinguishing contribution index of each nutrient; For the first The number of severely obese patients on the first The degree of contribution of each nutrient; This is the weight normalization function.

[0009] Furthermore, the method for obtaining the taste similarity between foods based on the differences in nutrient content and the degree of differentiation of nutrients is as follows: In the formula, For the first The food and the first The similarity of taste between individual foods; The total amount of nutrients; For the first The degree of differentiation among individual nutrients; For the first The first food The content of each nutrient; For the first The first food The content of each nutrient; It is a linear normalization function; It is an exponential function with the natural constant as its base; It is an absolute value function.

[0010] Furthermore, the specific method for obtaining the representative food and the target patient's rating of the representative food based on the rating of the pre-recommended food and the taste similarity between the pre-recommended foods includes: DBSCAN clustering was performed on all the pre-recommended foods to obtain several food clusters. The cluster distance metric was the difference in taste similarity between 1 and the pre-recommended foods. The representative probability of each pre-recommended food is obtained based on the average score of each pre-recommended food by all severely obese patients and the taste similarity between each pre-recommended food and other pre-recommended foods in its food cluster. For any food cluster, the food with the highest probability of being recommended in that food cluster is recorded as the representative food of that food cluster. Representative foods from all food clusters are pushed to the target patient, who then rates each representative food.

[0011] Furthermore, the specific method for obtaining the representative probability of each pre-recommended food based on the average score of each pre-recommended food by all severely obese patients and the taste similarity between each pre-recommended food and other pre-recommended foods in its food cluster includes: For any pre-recommended food, the product of the sum of the taste similarities between the pre-recommended food and all pre-recommended foods in its food cluster and the average score of the pre-recommended food by all severely obese patients is denoted as the representative probability of the pre-recommended food.

[0012] Furthermore, the specific methods for making food recommendations to the target patient based on the similarity between the representative food and the pre-recommended food in terms of taste and score include: Step 1: Among all representative foods rated by the target patient, select the highest-rated representative food as the reference food; Step 2: Calculate the recommendation index for each of the pre-recommended foods provided by the nutritionist: In the formula, For the first Recommendation index for each pre-recommended food; The number of foods with known scores for the target patient; For the first The recommended food and the known scores of the target patients The similarity of taste between foods; For the first The similarity in taste between each food and a reference food; For the known scores of the target patient The food rating; The target patients rated the reference foods; Step 3: Recommend the food with the highest recommendation index to the target patients; Step 4: Have the target patient rate the recommended food; Step 5: Perform steps 2 through 4 on the target patient each time they eat, until the patient recovers.

[0013] Furthermore, the specific method for obtaining several pre-recommended foods for the target patient includes: Acquire the target patient's physical health data, including weight and BMI, body fat percentage and muscle mass, as well as basal metabolic rate and blood glucose level; Based on the target patient's health data, determine the food calories required for each meal, and select several pre-recommended foods based on the food calories.

[0014] The beneficial effects of this invention are as follows: When recommending food to target patients, in order to better suit their taste preferences, it is necessary to differentiate foods based on the impact of nutrient content on taste. This invention analyzes the healthy dietary records of several severely obese patients to obtain the differentiation degree of each nutrient and determine the impact of each nutrient on food taste. To understand the dietary preferences of target patients, this invention obtains the taste similarity between foods by combining the differentiation degree of nutrients with the similarity of nutrient content, thereby differentiating foods with different tastes and selecting representative foods to understand the taste preferences of target patients, providing basic data for food recommendations. During the process of recommending food to target patients, this invention iteratively optimizes the recommended food results by using the target patients' ratings of representative foods and the similarity of other foods to representative foods, making the recommended food results more in line with the taste preferences of target patients. Thus, this invention continuously recommends food during the treatment process of target patients through recommended food and the target patients' ratings of the recommended food, improving the treatment experience of severely obese patients. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a structural block diagram of a smart recommendation system for healthy eating for severely obese patients, provided as an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 The diagram illustrates a flowchart of a smart dietary recommendation system for severely obese patients according to an embodiment of the present invention. The system includes: Dietary Records and Pre-Recommended Food Acquisition Module 101: Acquire healthy dietary records of several severely obese patients; designate severely obese patients who require targeted dietary plans as target patients; acquire several pre-recommended foods for target patients.

[0019] It should be noted that the purpose of this embodiment is to analyze the health dietary records of severely obese patients during treatment to determine the degree to which each nutrient differentiates food, thereby making food recommendations for the target patients, and optimizing the food recommendations based on the target patients' rating feedback, providing more personalized food recommendations for the target patients. Therefore, it is first necessary to obtain the health dietary records of severely obese patients during treatment.

[0020] Specifically, healthy eating records of several severely obese patients during treatment are obtained. These records include, but are not limited to, the time of eating, the name of the food, the score, and the content of nutrients contained in the food. These nutrients include: carbohydrates, amino acids, fats, organic acids, salts, alkaloids, cellulose, and volatile compounds.

[0021] It should be further explained that, since it is necessary to make food recommendations for patients with severe obesity, and the premise of the recommendations is to meet the patient's treatment needs, the food for patients with severe obesity should be initially selected first.

[0022] Specifically, a severely obese patient who requires a targeted dietary plan is designated as the target patient; Acquire the target patient's physical health data, including weight and BMI, body fat percentage and muscle mass, as well as basal metabolic rate and blood glucose level; The required food calories for each meal are determined based on the target patient's health data, and several pre-recommended foods are selected based on the food calories. The food calories are determined by a nutritionist using existing methods, which will not be elaborated on in this embodiment.

[0023] Module 102 for distinguishing contribution index: Based on the similarity relationship between the scores of foods with similar nutrient content by severely obese patients, the distinguishing contribution index of each dietary record of severely obese patients to each nutrient is obtained.

[0024] It's important to note that to recommend foods that better suit the target patients' tastes, it's necessary to differentiate the flavors of the foods in the pre-recommended diet. Among the foods recorded in the health diet records of all severely obese patients during treatment, the different nutrients contained in different foods lead to variations in taste, and different nutrients exhibit different degrees of influence on taste. In a short period, patients may show a greater preference for a particular flavor profile, i.e., they may rate that flavor profile higher. The nutrients common to this flavor profile are those with higher differentiation between food categories. Therefore, based on the content of each nutrient in different foods and each obese patient's food ratings, the nutrients with higher differentiation are identified.

[0025] It should be further clarified that in the healthy eating record of any severely obese patient, for a food with a high score, the more similar the content of various nutrients in other foods are to that food, the more likely they belong to the same flavor category. Such foods should all receive high scores from this severely obese patient. For any given nutrient, if that nutrient has a high degree of differentiation between food categories, then foods with similar nutrient content should also have similar scores.

[0026] Specifically, the first The first severely obese patient The first dietary record for the first The method for calculating the distinguishing contribution index of each nutrient is as follows: In the formula, For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first The total number of dietary records for each severely obese patient; For the first The first severely obese patient The first in the food diary The content of each nutrient; For the first In addition to the first group of severely obese patients The first one outside of the diet record The first in the food diary The content of each nutrient; For the first The number of severely obese patients on the first A rating of each food record; For the first A severely obese patient, except for the first The first one outside of the diet record A rating of each food record; It is an exponential function with the natural constant as its base; It is an absolute value function.

[0027] Module 103: Based on the changes in the differentiation contribution index of severely obese patients for each nutrient across all dietary records, the reference weight of each dietary record for severely obese patients for each nutrient is obtained; based on the differentiation contribution index and reference weight of all dietary records for severely obese patients for each nutrient, the degree of differentiation contribution of severely obese patients to each nutrient is obtained; based on the degree of differentiation contribution of all severely obese patients to each nutrient and the differentiation contribution index of all dietary records for all severely obese patients for each nutrient, the degree of differentiation for each nutrient is obtained.

[0028] It should be noted that dietary preferences in severely obese patients may change over time, especially during treatment. As their health improves or changes, they may gradually adapt to new eating habits, leading to changes in dietary preferences. Therefore, a patient's preference for a particular food category is not continuous during treatment; it is more likely to be a period of liking. Consequently, the phenomenon of similar ratings for the same type of food only exists within a specific timeframe. That is, a higher distinguishing contribution index for a particular nutrient only appears within a specific time period, and this index shows a clear upward and downward trend due to changes in patient preferences during this period. Because dietary records are kept chronologically, the distinguishing contribution index of a nutrient will show a clear upward and downward trend within this timeframe.

[0029] Specifically, for any severely obese patient, the sequence formed by arranging the patient's distinguishing contribution index to any nutrient according to the order of the dietary records is denoted as the distinguishing sequence of that nutrient for that severely obese patient. Obtain all extreme values ​​of the distinguishing sequence; for any distinguishing contribution index in the distinguishing sequence, record the maximum value among the extreme values ​​on both sides of the distinguishing contribution index as the reference weight of the dietary record corresponding to the distinguishing contribution index for that nutrient; for distinguishing contribution indices with extreme values ​​on both sides, and consisting of a maximum and a minimum value, obtain the reference weight as described above; if the distinguishing contribution index itself is an extreme value (maximum or minimum), then process it using the maximum value closest to the distinguishing contribution index, that is, if it is a maximum, process it using itself, and if it is a minimum, obtain the nearest maximum value, where nearest means the dietary record closest in the time sequence; if the distinguishing contribution index has an extreme value on only one side, that is, it is close to the boundary of the distinguishing sequence and therefore there is no extreme value on the other side, then similarly obtain the nearest maximum value and process it.

[0030] It should be noted that if the reference weight is high, it means that the time interval corresponding to the extreme value interval of the differentiation contribution index may fall within the time range during which the patient likes this type of food, indicating that the nutrient has a high differentiation ability. If the reference weight is consistently low, it means that the content of the nutrient has little impact on the differentiation of food categories, or that the patient has never preferred foods containing the nutrient during the treatment process.

[0031] It should be further noted that if dietary records with higher reference weights also have higher differentiation contribution indices, and the more dietary records with higher reference weights there are, the stronger the ability of the nutrient to differentiate food categories.

[0032] Specifically, the first The number of severely obese patients on the first The method for calculating the distinct contribution of each nutrient is as follows: In the formula, For the first The number of severely obese patients on the first The degree of contribution of each nutrient; For the first The total number of dietary records for each severely obese patient; For the first The first severely obese patient The first dietary record for the first Reference weights for each nutrient; For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first The first severely obese patient The first dietary record for the first The distinguishing index of each nutrient.

[0033] It should be further explained that, since some patients may not have a preference for a certain nutrient during treatment, their dietary records during treatment cannot be used as a basis for judging the food category of that nutrient. Therefore, the weight of the patient's judgment should be reduced.

[0034] Specifically, the first The method for calculating the degree of differentiation of individual nutrients is as follows: In the formula, For the first The degree of differentiation among individual nutrients; The number of severely obese patients; For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first Among all the dietary records of the severely obese patient, the first The maximum value of the distinguishing contribution index of each nutrient; For the first The number of severely obese patients on the first The degree of contribution of each nutrient; This is a weight normalization function, where the normalization object is the weight of all severely obese patients. Each nutrient .

[0035] It should be noted that if a patient's differential contribution index for a certain nutrient is consistently low, that is... The smaller size suggests that the patient may not have had a preference for this nutrient during treatment, hence the use of... The function reduces the weight of the patient's judgment on the degree of differentiation of the nutrient.

[0036] Representative food screening and scoring module 104: Based on the differences in nutrient content and the degree of differentiation of nutrients among foods, the similarity of taste between foods is obtained; based on the scores of the pre-recommended foods and the similarity of taste among the pre-recommended foods, the scores of representative foods and the target patients on representative foods are obtained.

[0037] It should be noted that when making food recommendations to target patients, it is necessary to understand their current food preferences in order to provide personalized food recommendation services. Since there are many different types of food, it is necessary for target patients to rate representative foods in order to understand their preferences.

[0038] It should be further clarified that representative foods are those that have a high degree of similarity in taste to some other foods and are liked by most patients. Therefore, the first step is to measure the taste similarity between foods. When measuring the taste similarity between foods, the more similar the two foods are in the content of nutrients that are highly distinguishable, the higher the taste similarity between the two foods.

[0039] Specifically, the first The food and the first The method for calculating the flavor similarity of food items is as follows: In the formula, For the first The food and the first The similarity of taste between individual foods; The total amount of nutrients; For the first The degree of differentiation among individual nutrients; For the first The first food The content of each nutrient; For the first The first food The content of each nutrient; It is a linear normalization function, and the normalization object is the degree of differentiation of all nutrients; It is an exponential function with the natural constant as its base; It is an absolute value function.

[0040] It should be noted that when selecting a representative food, the representative food represents a category of food, and the categories of food are divided according to taste, meaning that the foods in a single category are highly similar.

[0041] Specifically, DBSCAN clustering is performed on all pre-recommended foods to obtain several food clusters. The cluster distance metric is the difference between 1 and the taste similarity between the pre-recommended foods. DBSCAN clustering is a well-known technique, and the specific method will not be described here.

[0042] It should be noted that each food cluster is a food flavor category. For any pre-recommended food, if it has a high degree of similarity in taste to other pre-recommended foods in its food cluster, and the patient's rating of the pre-recommended food is high, then the pre-recommended food is more likely to be used as a representative food.

[0043] Specifically, the first The representative probability of each pre-recommended food item is calculated as follows: In the formula, For the first The representative probability of each pre-recommended food; For all severely obese patients, the first The average rating of the pre-recommended foods; For the first The number of recommended foods in the food cluster to which each recommended food belongs; For the first The first recommended food and its corresponding food cluster. Flavor similarity between the pre-recommended foods; For any food cluster, the food with the highest probability of being recommended in that food cluster is designated as the representative food of that food cluster.

[0044] It should be noted that the representative food represents a category of food. In order to understand the taste preferences of the target patients, the target patients need to rate all the representative foods.

[0045] Specifically, representative foods from all food clusters are pushed to the target patients, who then rate each representative food.

[0046] Recommended Food Module 105: Based on the similarity between the taste and scores of representative foods and pre-recommended foods, food recommendations are made for the target patients.

[0047] It's important to note that when recommending food to target patients, it's crucial to cater to their taste preferences as much as possible to enhance their comfort during treatment. The selected representative food only indicates that it suits the taste preferences of most people. Further refinement of the representative food is needed to determine which patient's preferred flavor is closest to the chosen food, based on patient ratings, to improve the overall treatment experience.

[0048] It should be further explained that when further determining the patient's food preferences, it is mainly based on the food that the target patient has already rated. For two foods with known high ratings, the patient's preference may fall between the two foods in terms of taste. Therefore, it is necessary to judge based on the taste similarity of a certain food to the two foods. If the ratio of the taste similarity of a certain food to the two foods is similar to the ratio of the ratings of the two foods, it means that the food is more suitable as a recommended food.

[0049] Specifically, food recommendations should be made for the target patients in the following manner: Step 1: Among all representative foods rated by the target patient, select the highest-rated representative food as the reference food; Step 2: Calculate the recommendation index for each of the pre-recommended foods provided by the nutritionist: In the formula, For the first Recommendation index for each pre-recommended food; The number of foods with known scores for the target patient; For the first The recommended food and the known scores of the target patients The similarity of taste between foods; For the first The similarity in taste between each food and a reference food; For the known scores of the target patient The food rating; The target patients rated the reference foods; Step 3: Recommend the food with the highest recommendation index to the target patients; Step 4: Have the target patient rate the recommended food; Step 5: Perform steps 2 to 4 as described above each time the target patient eats, until the patient recovers.

[0050] It should be noted that the recommended food is continuously optimized based on the target patient's food rating at each meal, making it more in line with the patient's taste preferences and improving the patient's treatment experience.

[0051] This embodiment adopts The model is used to represent the inverse proportional relationship and for normalization processing. As input to the model, implementers can set inverse proportional functions and normalization functions according to the actual situation.

[0052] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart dietary recommendation system for severely obese patients, characterized in that, The system includes: Dietary Record and Pre-Recommended Food Acquisition Module: Acquires healthy dietary records of several severely obese patients; the healthy dietary records include the name and score of the food consumed at each meal during treatment, as well as the content of several nutrients contained in the food; severely obese patients who require targeted dietary plans are identified as target patients; acquires several pre-recommended foods for the target patients; Differentiating contribution index acquisition module: Based on the similarity relationship between the scores of severely obese patients on foods with similar nutrient content, the differentiating contribution index of each dietary record of severely obese patients to each nutrient is obtained. The discrimination analysis module: Based on the changes in the discrimination contribution index of severely obese patients to each nutrient across all dietary records, the reference weight of each dietary record for severely obese patients to each nutrient is obtained; based on the discrimination contribution index and reference weight of all dietary records for severely obese patients to each nutrient, the discrimination contribution degree of severely obese patients to each nutrient is obtained; based on the discrimination contribution degree of all severely obese patients to each nutrient and the discrimination contribution index of all dietary records for all severely obese patients to each nutrient, the discrimination degree of each nutrient is obtained. Representative food screening and scoring module: Based on the differences in nutrient content and the degree of differentiation of nutrients among foods, the taste similarity between foods is obtained; based on the scores of the pre-recommended foods and the taste similarity between the pre-recommended foods, the scores of the representative foods and the target patients on the representative foods are obtained. Recommended food module: Based on the similarity between the taste and scores of representative foods and pre-recommended foods, food recommendations are made for the target patients.

2. The intelligent health diet recommendation system for severely obese patients according to claim 1, characterized in that, The method for obtaining the differentiation contribution index of each dietary record of severely obese patients to each nutrient is based on the similarity relationship between the scores of foods with similar nutrient content given by severely obese patients. In the formula, For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first The total number of dietary records for each severely obese patient; For the first The first severely obese patient The first in the food diary The content of each nutrient; For the first In addition to the first group of severely obese patients The first one outside of the diet record The first in the food diary The content of each nutrient; For the first The number of severely obese patients on the first A rating of each food record; For the first A severely obese patient, except for the first The first one outside of the diet record A rating of each food record; It is an exponential function with the natural constant as its base; It is an absolute value function.

3. The intelligent health diet recommendation system for severely obese patients according to claim 1, characterized in that, The method for obtaining the reference weight of each dietary record for each nutrient for severely obese patients based on the variation of the differential contribution index of each nutrient to all dietary records across all dietary records includes: For any severely obese patient, the sequence formed by arranging the patient's distinguishing contribution index to any nutrient according to the order of the dietary records is denoted as the distinguishing sequence of that nutrient for that severely obese patient. Obtain all extreme values ​​of the discrimination sequence; for any discrimination contribution index in the discrimination sequence, record the maximum value of the extreme values ​​on both sides of the discrimination contribution index as the reference weight of the dietary record corresponding to the discrimination contribution index for that nutrient.

4. The intelligent dietary recommendation system for severely obese patients according to claim 1, characterized in that, The method for determining the contribution of severely obese patients to each nutrient is as follows: The contribution index and reference weight of each nutrient are used to differentiate the nutrients based on all dietary records of severely obese patients. For any dietary record of any severely obese patient, the product of the reference weight of the dietary record for any nutrient and the distinguishing contribution index is recorded as the distinguishing index of the dietary record for that nutrient in the severely obese patient. The sum of the discrimination indices of all dietary records of the severely obese patient for this nutrient is recorded as the degree of discrimination contribution of the severely obese patient to this nutrient.

5. The intelligent health diet recommendation system for severely obese patients according to claim 1, characterized in that, The degree of differentiation for each nutrient is obtained by considering the contribution of all severely obese patients to the differentiation of each nutrient and the differentiation contribution index of all dietary records of all severely obese patients to the differentiation of each nutrient. The specific method for obtaining this information is as follows: In the formula, For the first The degree of differentiation among individual nutrients; The number of severely obese patients; For the first The first severely obese patient The first dietary record for the first The distinguishing contribution index of each nutrient; For the first Among all the dietary records of the severely obese patient, the first The maximum value of the distinguishing contribution index of each nutrient; For the first The number of severely obese patients on the first The degree of contribution of each nutrient; This is the weight normalization function.

6. The intelligent health diet recommendation system for severely obese patients according to claim 1, characterized in that, The similarity of taste between foods is obtained based on the differences in nutrient content and the degree of differentiation of nutrients. The specific method for obtaining this similarity is as follows: In the formula, For the first The food and the first The similarity of taste between individual foods; The total amount of nutrients; For the first The degree of differentiation among individual nutrients; For the first The first food The content of each nutrient; For the first The first food The content of each nutrient; It is a linear normalization function; It is an exponential function with the natural constant as its base; It is an absolute value function.

7. The intelligent health diet recommendation system for severely obese patients according to claim 1, characterized in that, The method for obtaining representative foods and target patient ratings for representative foods based on the scores of pre-recommended foods and the taste similarity between pre-recommended foods includes: DBSCAN clustering was performed on all the pre-recommended foods to obtain several food clusters. The cluster distance metric was the difference in taste similarity between 1 and the pre-recommended foods. The representative probability of each pre-recommended food is obtained based on the average score of each pre-recommended food by all severely obese patients and the taste similarity between each pre-recommended food and other pre-recommended foods in its food cluster. For any food cluster, the food with the highest probability of being recommended in that food cluster is recorded as the representative food of that food cluster. Representative foods from all food clusters are pushed to the target patient, who then rates each representative food.

8. The intelligent health diet recommendation system for severely obese patients according to claim 7, characterized in that, The method for obtaining the representative probability of each pre-recommended food based on the average score of each pre-recommended food given by all severely obese patients and the taste similarity between each pre-recommended food and other pre-recommended foods in its food cluster includes the following: For any pre-recommended food, the product of the sum of the taste similarities between the pre-recommended food and all pre-recommended foods in its food cluster and the average score of the pre-recommended food by all severely obese patients is denoted as the representative probability of the pre-recommended food.

9. A smart dietary recommendation system for severely obese patients according to claim 1, characterized in that, The method for making food recommendations to target patients based on the similarity between the taste and scores of representative foods and pre-recommended foods includes the following specific methods: Step 1: Among all representative foods rated by the target patient, select the highest-rated representative food as the reference food; Step 2: Calculate the recommendation index for each of the pre-recommended foods provided by the nutritionist: In the formula, For the first Recommendation index for each pre-recommended food; The number of foods with known scores for the target patient; For the first The recommended food and the known scores of the target patients The similarity of taste between foods; For the first The similarity in taste between each food and a reference food; For the known scores of the target patient The food rating; The target patients rated the reference foods; Step 3: Recommend the food with the highest recommendation index to the target patients; Step 4: Have the target patient rate the recommended food; Step 5: Perform steps 2 through 4 on the target patient each time they eat, until the patient recovers.

10. A smart dietary recommendation system for severely obese patients according to claim 1, characterized in that, The specific methods for obtaining several pre-recommended foods for the target patient include: Acquire the target patient's physical health data, including weight and BMI, body fat percentage and muscle mass, as well as basal metabolic rate and blood glucose level; Based on the target patient's health data, determine the food calories required for each meal, and select several pre-recommended foods based on the food calories.

Images