A method for constructing a family user feature database

Abstract

This invention relates to the field of family data analysis technology, and more particularly to a method for constructing a family user characteristic database, comprising: step S1, collecting health information, behavioral information, and identity information of family users; step S2, extracting and analyzing the daily characteristics of users based on the feature data; step S3, collecting unique information of family users based on the role type in the identity information; step S4, analyzing the unique characteristics of users based on the collected identity information and unique information; step S5, storing the feature data and unique characteristics of users, analyzing the family data characteristics based on the stored feature data, and classifying the family data characteristics to obtain family characteristic types; step S6, analyzing the saliency of family user characteristics; and step S7, constructing a family user characteristic database based on the saliency of family user characteristics. This invention achieves accurate analysis of family user characteristics.

Description

Technical Field

[0001] This invention relates to the field of family data analysis technology, and in particular to a method for constructing a family user characteristic database. Background Technology

[0002] Family member profiling involves collecting and storing unique biological, health, and behavioral data of family members, utilizing database technology and data analysis methods to achieve personalized management and services. With the development of the Internet of Things, artificial intelligence, and big data technologies, smart devices and biometric technologies have facilitated data collection. By building a profiling database, applications such as health monitoring, security verification, and personalized recommendations can be supported. However, it is crucial to prioritize data privacy and security protection, ensuring compliance with relevant laws and regulations.

[0003] Chinese Patent Publication No. CN112559586A discloses a method and system for obtaining household user characteristic data, including the following steps: Step S1: Obtain the identifier of the terminal service before and after network switching, where the network includes broadband and mobile networks; Step S2: Determine whether the broadband network is a household broadband network; Step S3: Perform data feature analysis on the user corresponding to the household broadband network identifier to determine whether the user is a household user member under the household broadband network identifier; Step S4: Store the ID of the household user member and its corresponding fixed network characteristic data and mobile network characteristic data under the household user member identifier corresponding to the household broadband network identifier, and establish a data set of all members under the household broadband network identifier. This invention achieves the classification and identification of data of different members in the household based on network information, but it does not achieve the extraction of characteristic data based on the daily health data, behavioral data, and unique data of different roles of household users, nor does it construct a household characteristic database based on the characteristic data. Therefore, it suffers from low efficiency in analyzing household user data and inaccurate analysis of household user characteristics. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a database of household user characteristics, so as to solve at least one of the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A method for constructing a database of household user characteristics includes:

[0007] Step S1: Collect health information, behavioral information, and identity information of household users;

[0008] Step S2: Extract feature data based on the collected health information, behavioral information, and identity information of the household users; and analyze the daily characteristics of the household users based on the collected health information, behavioral information, and identity information.

[0009] Step S3: Collect unique information about family users based on the role type in the family user's identity information;

[0010] Step S4: Analyze the unique characteristics of users based on the collected identity information and unique information of family users;

[0011] Step S5: Store the feature data and user-specific features, analyze the family data features based on the stored feature data, and classify the family data features to obtain the family feature type.

[0012] Step S6: Analyze the similarity of similar families based on family data characteristics and family characteristic types, and analyze the feature salience of family users based on similar family similarity, user daily characteristics, and user-specific characteristics;

[0013] Step S7: Construct a family user feature database based on the salience of family user features.

[0014] Further, step S2 includes:

[0015] Step S21: Number the household users to obtain the household user number and the household number;

[0016] Step S22: Extract feature data based on health information, behavioral information, and identity information;

[0017] Step S23: Analyze the user's daily characteristics based on health information, behavioral information, and identity information.

[0018] Further, in step S21, family users are numbered according to their identity information, and each family user is numbered in descending order of age. The family user number is set as i, and different families are numbered and set as j.

[0019] In step S22, the number of family users in the same household is counted as the number of family members, and the number of family members is set as NP(j); the frequency of each food preference in the same household is counted, and the food preference corresponding to the frequency of the most frequent food preference is taken as the family food preference. If the food preferences of each user in the same household are different, the family food preference is set to no preference; the frequency of each type of exercise preference in the same household is counted, and the type of exercise preference corresponding to the frequency of the most frequent exercise preference is taken as the family exercise preference. If the type of exercise preference of each user in the same household is different, the family exercise preference is set to no preference; the family meal time is set to be the same as the user's meal time.

[0020] In step S23, the daily characteristics of the family users are analyzed based on their height, weight, blood pressure, exercise time, sleep time, wake-up time and name to obtain the daily characteristics of the users D(j,i).

[0021] Furthermore, in step S3, unique information of the family user is collected based on the identity information. If the role type is child, the education information of the currently analyzed family user is collected. If the role type is father, the work information and living habits information of the currently analyzed family user are collected. If the role type is mother, the shopping information and childcare information of the currently analyzed family user are collected. If the role type is elderly, the disease information and medication information of the currently analyzed family user are collected.

[0022] Furthermore, in step S4, the user's unique characteristics are analyzed based on name, role type, and unique information to obtain the user's unique characteristics S(j,i).

[0023] Further, step S5 includes:

[0024] Step S51: Store the feature data and user-specific features;

[0025] Step S52: Analyze the characteristics of the household data based on the stored feature data;

[0026] Step S53: Count the number of similar families based on the characteristics of family data;

[0027] Step S54: Classify the family data features according to the number of similar families to obtain the family feature types.

[0028] Furthermore, in step S52, the family data features are analyzed based on the stored feature data to obtain the family data features F(j);

[0029] In step S53, families corresponding to the non-currently analyzed family data features that satisfy α≤F(j)·F(k)≤1 are extracted as similar families, and the number of similar families is counted as the number of similar families N(j), where F(k) represents the non-currently analyzed family data features, k represents the non-currently analyzed family number, and α represents the family extraction parameter;

[0030] In step S54, the family data features are classified according to the number of similar families to obtain family feature types, which include Category I and Category II.

[0031] Further, step S6 includes:

[0032] Step S61: Analyze the saliency of features based on users' daily characteristics and user-specific characteristics;

[0033] Step S62: Analyze the similarity of similar families based on the characteristics of family data;

[0034] Step S63: Improve the feature saliency analysis process based on the similarity of similar families;

[0035] Step S64: Improve the analysis process of similar family similarity based on family characteristic types.

[0036] Furthermore, in step S61, the feature saliency is analyzed based on the user's daily characteristics D(j,i) and user-specific characteristics S(j,i) to obtain the feature saliency G(j,i).

[0037] In step S62, the similarity of similar families is analyzed based on the daily characteristics D(j,i) of similar families and users to obtain the similarity W(j);

[0038] In step S63, the analysis process of feature saliency is improved based on the similarity W(j) of similar families. When W(j)≤w, the analysis process of feature saliency is improved, and the improved feature saliency is G1(j,i), where w represents the similarity threshold.

[0039] In step S64, the analysis process of similar family similarity is improved according to the family characteristic type. When the family characteristic type is of the same type, the analysis process of similar family similarity is improved, and the improved similar family similarity is W1(j).

[0040] Furthermore, in step S7, a family user feature database is constructed based on feature saliency. The collected health information, behavioral information, and identity information of family users are used as the basic data of the family user feature database, and user biometrics are used as the verification data for accessing the family user feature database. If G(j,i)>g, the collected unique information of the current family user is used as the feature data of the family user feature database; otherwise, the collected unique information of the current family user is used as the basic data of the family user feature database. Here, g represents the feature construction threshold.

[0041] The beneficial effects of this invention are as follows: By collecting and analyzing the health information, behavioral information, and identity information of family users, as well as unique data based on the role type of family users, it is possible to determine whether each data point is significant for each family user by combining the user's daily data characteristics and their unique data characteristics. This enables the analysis and extraction of data that reflects the characteristics of family users, so that the data in the constructed family user characteristic database can clearly reflect the user characteristics, providing effective data for subsequent family user analysis, thereby improving the efficiency of family user data analysis and the accuracy of family user characteristic analysis. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0043] Figure 1 This is a flowchart of the method for constructing a family user feature database in this embodiment.

[0044] Figure 2 This is a flowchart of the method for analyzing feature data and user daily characteristics in this embodiment.

[0045] Figure 3 This is a flowchart illustrating the analysis of household data characteristics in this embodiment.

[0046] Figure 4 This is a flowchart of the feature saliency analysis in this embodiment. Detailed Implementation

[0047] To more clearly illustrate the present invention, the following description, in conjunction with preferred embodiments and accompanying drawings, further explains the invention. Similar components in the drawings are indicated by the same reference numerals. Those skilled in the art should understand that the specific description below is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the present invention.

[0048] It should be noted that although the terms first, second, third, etc., may be used in the embodiments of this application for description, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of this application, first can also be referred to as second, and similarly, second can also be referred to as first.

[0049] Please see Figure 1 As shown, this is the method for constructing a family user feature database in this embodiment, including:

[0050] Step S1: Collect health information, behavioral information, and identity information of the family users. The health information includes, but is not limited to, data on the family users' physiological status such as height, weight, and blood pressure. Height is measured in meters, weight in kilograms, and blood pressure in millimeters of mercury. Blood pressure includes diastolic and systolic pressure. The behavioral information includes, but is not limited to, data on the family users' daily habits and preferences such as dietary information, exercise information, and sleep-wake cycle information. Dietary information includes food preferences and meal times. Exercise information includes exercise preference type, exercise time, and exercise intensity. Exercise time is measured in minutes. Exercise preference type includes, but is not limited to, running and basketball. Exercise intensity includes low, medium, and high intensity. Sleep-wake cycle information includes sleep time and wake-up time. The identity information includes, but is not limited to, data about the identity of family users, such as name, age, role type, and user biometrics. The role type includes, but is not limited to, data about the role of the family user in the family, such as father, mother, and child. The user biometrics include, but are not limited to, biometrics used to identify the identity of family users, such as fingerprints, facial features, and voiceprints. It should be noted that the family user information (including but not limited to family user device information, family user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) involved in this application are all information and data authorized by the subject or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the country and region.

[0051] Specifically, in step S1 of this embodiment, by collecting the health information (such as height, weight, blood pressure, etc.), behavioral information (such as diet, exercise, and daily routine), and identity information (such as name, age, and role type) of family users, a comprehensive data foundation is provided for subsequent feature extraction and analysis. The diversity of this data ensures that the database can cover multiple dimensions of family users, thereby providing rich information support for subsequent personalized management and services.

[0052] Specifically, this embodiment is applied to a cloud-based family data management server. By collecting data from different groups of people, it extracts the characteristic data of users of each role type, thereby achieving comprehensive extraction of the characteristic data of each member of the family.

[0053] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0054] Step S2: Extract feature data based on the collected health information, behavioral information, and identity information of the household users, and analyze the daily characteristics of the household users based on the collected health information, behavioral information, and identity information.

[0055] Specifically, in step S2 of this embodiment, by extracting and analyzing health information, behavioral information, and identity information, the daily characteristics of household users (such as exercise time, sleep time, and dietary preferences) can be identified. These daily characteristics are the core of household user behavior patterns and can help the database better reflect users' daily habits and preferences, thereby improving the accuracy and usability of the database.

[0056] Please see Figure 2 As shown, it is a method for analyzing feature data and users' daily characteristics, including:

[0057] Step S21: Number the household users to obtain the household user number and household number.

[0058] Specifically, in step S21 of this embodiment, family users are numbered according to their identity information. Each family user is numbered in descending order of age, and the family user number is set as i, i∈N. + Different families are assigned numbers, with family number j, where j∈N. + .

[0059] Please continue reading. Figure 2 As shown, the analysis method for the feature data and user daily characteristics also includes:

[0060] Step S22: Extract feature data based on health information, behavioral information, and identity information.

[0061] Specifically, in step S22 of this embodiment, the number of family users in the same household is counted as the number of family members, and the number of family members is set as NP(j); the frequency of each food preference in the same household is counted, and the food preference corresponding to the frequency of the most frequent food preference is taken as the family food preference. If the food preferences of each user in the same household are different, the family food preference is set to no preference; the frequency of each type of exercise preference in the same household is counted, and the type of exercise preference corresponding to the frequency of the most frequent exercise preference is taken as the family exercise preference. If the exercise preference types of each user in the same household are different, the family exercise preference is set to no preference; the family meal time is set to be the same as the user's meal time.

[0062] It is understood that this embodiment does not impose specific limitations on the extraction of feature data, as long as the data that can reflect the characteristics of the family can be extracted. For example, the extraction of data such as family consumption habits and the usage of equipment in the family can also be set.

[0063] Please continue reading. Figure 2 As shown, the analysis method for the feature data and user daily characteristics also includes:

[0064] Step S23: Analyze the user's daily characteristics based on health information, behavioral information, and identity information.

[0065] Specifically, in step S23 of this embodiment, the user's daily characteristics are analyzed based on the user's height, weight, blood pressure, exercise time, sleep time, wake-up time, and name. The user's daily characteristics are set as D(j,i), and then... Where M(j,i) represents weight, H(j,i) represents height, T1(j,i) represents exercise time, mh represents body mass parameter (20≤mh≤22), BP1(j,i) represents systolic blood pressure, BP2(j,i) represents diastolic blood pressure, ST1(j,i) represents sleep time, and ST2(j,i) represents wake-up time. It is understood that this embodiment does not impose specific limitations on the values ​​of the body mass parameter; those skilled in the art can freely set them, as long as they satisfy the analysis of the user's daily characteristics. The optimal value for the body mass parameter is: mh=21.

[0066] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0067] Step S3: Collect unique information about the family users based on their role type in the family user's identity information. This unique information includes, but is not limited to, data on different user concerns, such as education information, work information, lifestyle habits information, shopping information, parenting information, disease information, and medication information. The education information includes, but is not limited to, academic performance and hobbies. The work information includes, but is not limited to, work stress level and working hours, with working hours in hours and work stress level categorized as low, medium, and high. The lifestyle habits information includes, but is not limited to, smoking and drinking information, with smoking information being the daily amount of cigarettes smoked and drinking information including the daily amount and frequency of alcohol consumed. The shopping information includes, but is not limited to, shopping preferences and frequency of purchases. The parenting information includes, but is not limited to, the time spent with children and family activity arrangements. The disease information includes, but is not limited to, hypertension and arthritis. The medication information includes, but is not limited to, medication dosage and medication records.

[0068] Specifically, in step S3 of this embodiment, unique information of family users is collected based on identity information. If the role type is child, the education information of the currently analyzed family user is collected. If the role type is father, the work information and lifestyle information of the currently analyzed family user are collected. If the role type is mother, the shopping information and childcare information of the currently analyzed family user are collected. If the role type is elderly, the disease information and medication information of the currently analyzed family user are collected.

[0069] It is understood that this embodiment does not specifically limit the content of the collection of unique information. Those skilled in the art can set it freely. Family users can set the unique information according to the composition of family members and the allocation of family tasks. When setting the collection of unique information for children, attention should be paid to the fact that the unique data is related to the child's education and healthy growth data. When setting the collection of unique information for fathers or mothers, attention should be paid to the fact that the unique information is related to their life and health data and work stress data. When setting the collection of unique information for the elderly, attention should be paid to the fact that the unique information is related to the health and safety of the elderly.

[0070] Specifically, in step S3 of this embodiment, specific information related to the role of the family user (such as father, mother, child, elderly, etc.) is collected (such as education information, work information, shopping information, disease information, etc.). Targeted data collection can more accurately reflect the specific needs and behavioral patterns of different roles, thereby enhancing the personalized analysis capabilities of the database.

[0071] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0072] Step S4: Analyze the unique characteristics of users based on the collected identity information and unique information of family users.

[0073] Specifically, in step S4 of this embodiment, the user's unique characteristics are analyzed based on name, role type, and unique information. The user's unique characteristics are set as S(j,i). If the role type is child, S(j,i) = σ(s(j,i)); if the role type is father, S(j,i) = σ(s(j,i)). If the character type is mother, set S(j,i)=[f3(j,i)]·[T3(j,i)]; if the character type is elderly, set... Where s(j,i) represents academic performance, σ(s(j,i)) represents the standard deviation of academic performance, T2(j,i) represents working hours, A1(j,i) represents daily smoking amount, A2(j,i) represents daily alcohol consumption, f2(j,i) represents daily drinking frequency, f3(j,i) represents shopping frequency, T3(j,i) represents time spent with family, n(j,i) represents the number of diseases, A4(j,i) represents medication dosage, and E(j,i) represents age.

[0074] It is understood that this embodiment does not specifically limit the analysis method of user-specific features. Those skilled in the art can freely set it, as long as it meets the requirements for target feature extraction according to different role types. For example, the unique features of children should focus on children's education and growth data, the unique features of fathers should focus on fathers' work pressure and physical health data, the unique features of mothers should focus on mothers' daily shopping habits and childcare data, and the unique features of the elderly should focus on the elderly's physical health data.

[0075] Specifically, in step S4 of this embodiment, by analyzing the unique information of household users, the unique characteristics of users can be extracted. These unique characteristics are the key to distinguishing different users and can help the database more accurately identify and classify the characteristics of household users, thereby improving the accuracy of the database and the level of personalized services.

[0076] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0077] Step S5: Store the feature data and user-specific features, analyze the family data features based on the stored feature data, and classify the family data features to obtain the family feature types.

[0078] Specifically, in step S5 of this embodiment, by storing feature data and user-specific features, and analyzing and classifying family data features, it is possible to identify the feature types of different families, which can help the database better understand the similarities and differences between families, thereby improving the classification accuracy and analysis efficiency of the database.

[0079] Please see Figure 3 As shown, this is a method for analyzing the characteristics of household data, including:

[0080] Step S51: Store the feature data and user-specific features.

[0081] Step S52: Analyze the features of the household data based on the stored feature data.

[0082] Specifically, in step S52 of this embodiment, the family data features are analyzed based on the stored feature data, and the family data features are set as F(j), and F(j) = [NP(j), ET(j), P1(j), P2(j)], where ET(j) represents the family meal time, P1(j) represents the family food preference, and P2(j) represents the family exercise preference.

[0083] Specifically, in this embodiment, when analyzing family data characteristics, different numbers can be set for different family food preferences and family exercise preferences, and the numbers can be matched one-to-one with different preferences. For example, 0 represents no preference, and 1 represents family food preference as pasta or family exercise preference as running.

[0084] Specifically, in step S52 of this embodiment, the characteristics of family data are analyzed by analyzing the stored feature data, so as to realize the comprehensive analysis of the feature data of different families, and use the same set of data to describe the data characteristics of family feature data, thereby improving the efficiency of family user data analysis and improving the accuracy of family user feature analysis.

[0085] Please continue reading. Figure 3 As shown, the method for analyzing the characteristics of household data also includes:

[0086] Step S53: Count the number of similar families based on the characteristics of family data.

[0087] Specifically, in step S53 of this embodiment, families corresponding to the non-currently analyzed family data features that satisfy α≤F(j)·F(k)≤1 are extracted as similar families, and the number of similar families is counted as the number of similar families. The number of similar families is set as N(j), where F(k) represents the non-currently analyzed family data feature, k represents the non-currently analyzed family number, and k∈N +And k≠j, α represents the family extraction parameter, 0.9≤α<1. It is understood that this embodiment does not impose specific limitations on the value of the family extraction parameter. Those skilled in the art can set it freely, as long as it satisfies the extraction of similar families. The optimal value of the family extraction parameter is: α=0.95.

[0088] Please continue reading. Figure 3 As shown, the method for analyzing the characteristics of household data also includes:

[0089] Step S54: Classify the family data features according to the number of similar families to obtain the family feature types.

[0090] Specifically, in step S54 of this embodiment, the family data features are classified according to the number of similar families. If N(j) / j max <β, the family characteristic type of the family being analyzed is determined to be of type one, if N(j) / j max If ≥β, the family characteristic type of the currently analyzed family is determined to be Class II, where j max β represents the maximum value of the family ID, and β represents the family classification threshold, where 0.05 ≤ β ≤ 0.08. It is understood that this embodiment does not specifically limit the optimal value of the family classification threshold; those skilled in the art can freely set it, as long as it satisfies the judgment of family characteristic types. The optimal value of the family classification threshold is: β = 0.06.

[0091] Specifically, in step S54 of this embodiment, by analyzing similar families, it is determined whether there are many families similar to the currently analyzed family, thereby extracting families with strong specific characteristics, thus improving the efficiency of family user data analysis and the accuracy of family user characteristic analysis.

[0092] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0093] Step S6: Analyze the similarity of similar families based on family data characteristics and family characteristic types, and analyze the feature salience of family users based on similar family similarity, user daily characteristics, and user-specific characteristics.

[0094] Specifically, in step S6 of this embodiment, by analyzing the saliency of features of household users, it is possible to identify which features are most significant for household users, which can help the database extract the most representative features for users more accurately, thereby improving the accuracy of the database and its personalized recommendation capabilities.

[0095] Please see Figure 4 As shown, this is a method for analyzing feature significance, including:

[0096] Step S61: Analyze the saliency of features based on the user's daily characteristics and user-specific characteristics.

[0097] Specifically, in step S61 of this embodiment, feature saliency is analyzed based on user's daily characteristics and user-specific characteristics, and the feature saliency is set as G(j,i). Q(j,i) represents the user characteristic parameter, Q(j,i)=D(j,i)·S(j,i), and u represents the non-currently analyzed household user ID, u∈N. + And u≠i,i max This represents the maximum value of the family user ID.

[0098] Please continue reading. Figure 4 As shown, the method for analyzing feature saliency further includes:

[0099] Step S62: Analyze the similarity of similar families based on the characteristics of family data.

[0100] Specifically, in step S62 of this embodiment, the similarity of similar families is analyzed based on similar families and users' daily characteristics, and the similarity of similar families is set as W(j). Where D(K,i) represents the daily characteristics of users in similar families whose role type is the same as that of the user in the currently analyzed family, and K represents the family number of the similar family.

[0101] Please continue reading. Figure 4 As shown, the method for analyzing feature saliency further includes:

[0102] Step S63: Improve the analysis process of feature saliency based on similarity of similar families.

[0103] Specifically, in step S63 of this embodiment, the analysis process of improving feature saliency based on similarity of similar families is as follows: if W(j) > w, the analysis process of improving feature saliency is not improved; otherwise, the analysis process of improving feature saliency is improved. The improved feature saliency is G1(j,i), and G1(j,i) = G(j,i) × e W(j) Where w represents the similarity threshold, 0.4≤w≤0.6. It is understood that this embodiment does not impose specific limitations on the value of the similarity threshold, and those skilled in the art can set it freely, as long as it satisfies the improvement of the feature significance analysis process. The optimal value of the similarity threshold is: w=0.5.

[0104] Please continue reading. Figure 4 As shown, the method for analyzing feature saliency further includes:

[0105] Step S64: Improve the analysis process of similar family similarity based on family characteristic types.

[0106] Specifically, in step S64 of this embodiment, the analysis process of similarity between similar families is improved according to the family characteristic type. If the family characteristic type is of type one, the analysis process of similarity between similar families is improved, and the improved similarity between similar families is W1(j), which is set as W1(j) = [W(j) + R(j)] / 2. If the family characteristic type is of type two, the analysis process of similarity between similar families is not improved, where R(j) represents the unique characteristic similarity. S(K,i) represents the user-specific characteristics of family users in similar families whose role type is the same as that of the user in the currently analyzed family.

[0107] Please continue reading. Figure 1 As shown, the method for constructing the household user characteristic database further includes:

[0108] Step S7: Construct a family user feature database based on the salience of family user features.

[0109] Specifically, in step S7 of this embodiment, a family user feature database is constructed based on feature saliency. The collected health information, behavioral information, and identity information of family users are used as the basic data for the family user feature database, and user biometrics are used as verification data for accessing the database. If G(j,i) > g, the collected unique information of the current family user is used as the feature data of the family user feature database; otherwise, the collected unique information of the current family user is used as the basic data. Here, g represents the feature construction threshold, 0.7 ≤ g ≤ 0.9. It is understood that this embodiment does not specifically limit the value of the feature construction threshold; those skilled in the art can freely set it, as long as it satisfies the construction of the family user feature database. The optimal value for the feature construction threshold is: g = 0.8.

[0110] Specifically, the feature data in the family user feature database is feature data that reflects the differences between family users and other users. Its data content can better reflect the characteristics of users and provide necessary analytical data for understanding user needs and optimizing the user service experience.

[0111] Specifically, in step S7 of this embodiment, a family user feature database is constructed based on feature saliency. This ensures that the data in the database accurately reflects the characteristics and needs of family users. By using salient features as the core data of the database, the database's usability and personalized service level can be improved, thereby providing strong data support for subsequent applications such as health monitoring, security verification, and personalized recommendations.

[0112] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. A method for constructing a database of household user characteristics, characterized in that, include: Step S1: Collect health information, behavioral information, and identity information of household users; Step S2: Extract feature data based on the collected health information, behavioral information, and identity information of the household users; and analyze the daily characteristics of the household users based on the collected health information, behavioral information, and identity information. Step S3: Collect unique information about family users based on the role type in the family user's identity information; In step S3, unique information of family users is collected based on identity information. If the role type is child, the education information of the family user being analyzed is collected. If the role type is father, the work information and living habits information of the family user being analyzed are collected. If the role type is mother, the shopping information and childcare information of the family user being analyzed are collected. If the role type is elderly, the disease information and medication information of the family user being analyzed are collected. Step S4: Analyze the unique characteristics of users based on the collected identity information and unique information of family users; In step S4, user-specific characteristics are analyzed based on name, role type, and unique information to obtain user-specific characteristics S(j,i). If the role type is child, S(j,i) is set to σ(s(j,i)); if the role type is father, S(j,i) is set to σ(s(j,i)). If the character type is mother, set S(j,i)=[f3(j,i)]·[T3(j,i)]; if the character type is elderly, set... Where s(j,i) represents academic performance, σ(s(j,i)) represents the standard deviation of academic performance, T2(j,i) represents working hours, A1(j,i) represents daily smoking amount, A2(j,i) represents daily alcohol consumption, f2(j,i) represents daily drinking frequency, f3(j,i) represents shopping frequency, T3(j,i) represents time spent with family, n(j,i) represents number of diseases, A4(j,i) represents medication dosage, and E(j,i) represents age. Step S5: Store the feature data and user-specific features, analyze the family data features based on the stored feature data, and classify the family data features to obtain the family feature type. Step S5 includes: Step S51: Store the feature data and user-specific features; Step S52: Analyze the characteristics of the household data based on the stored feature data; Step S53: Count the number of similar families based on the characteristics of family data; Step S54: Classify the family data features according to the number of similar families to obtain the family feature types; In step S52, the family data features are analyzed based on the stored feature data to obtain the family data features F(j); In step S53, families corresponding to the non-currently analyzed family data features that satisfy α≤F(j)·F(k)≤1 are extracted as similar families, and the number of similar families is counted as the number of similar families N(j), where F(k) represents the non-currently analyzed family data features, k represents the non-currently analyzed family number, and α represents the family extraction parameter; In step S54, the family data features are classified according to the number of similar families to obtain family feature types, which include Category I and Category II. Step S6: Analyze the similarity of similar families based on family data characteristics and family characteristic types, and analyze the feature salience of family users based on similar family similarity, user daily characteristics, and user-specific characteristics; Step S6 includes: Step S61: Analyze the saliency of features based on users' daily characteristics and user-specific characteristics; Step S62: Analyze the similarity of similar families based on the characteristics of family data; Step S63: Improve the feature saliency analysis process based on the similarity of similar families; Step S64: Improve the analysis process of similar family similarity based on family characteristic types; In step S61, the feature saliency is analyzed based on the user's daily characteristics D(j,i) and user-specific characteristics S(j,i) to obtain the feature saliency G(j,i), and then set... Q(j,i) represents the user characteristic parameter, Q(j,i)=D(j,i)·S(j,i), and u represents the non-currently analyzed household user ID, u∈N. + And u≠i,i max This represents the maximum value of the household user ID; In step S62, the similarity of similar families is analyzed based on the daily characteristics D(j,i) of similar families and users to obtain the similarity W(j) of similar families, and then set... Where D(K,i) represents the daily characteristics of users in similar families whose role type is the same as that of the user in the family being analyzed, and K represents the family ID of the similar families; In step S63, the analysis process of improving feature significance is carried out based on the similarity of similar families W(j). When W(j) ≤ w, the analysis process of improving feature significance is carried out, and the improved feature significance is G1(j,i). G1(j,i) is set to G(j,i) × e W (j) w represents the similarity threshold; In step S64, the analysis process for similarity of similar families is improved based on family characteristic types. When the family characteristic type is of the same category, the analysis process for similarity of similar families is improved, and the improved similarity of similar families is denoted as W1(j). W1(j) is set to = [W(j) + R(j)] / 2, where R(j) represents the unique characteristic similarity. S(K,i) represents the user-specific characteristics of family users in similar families whose role type is the same as that of the user in the currently analyzed family; Step S7: Construct a family user feature database based on the salience of family user features; In step S7, a family user feature database is constructed based on feature saliency. The collected health information, behavioral information, and identity information of family users are used as the basic data of the family user feature database, and user biometrics are used as the verification data for accessing the family user feature database. If G(j,i)>g, the collected unique information of the current family user is used as the feature data of the family user feature database; otherwise, the collected unique information of the current family user is used as the basic data of the family user feature database. Here, g represents the feature construction threshold.

2. The method for constructing a family user characteristic database according to claim 1, characterized in that, Step S2 includes: Step S21: Number the household users to obtain the household user number and the household number; Step S22: Extract feature data based on health information, behavioral information, and identity information; Step S23: Analyze the user's daily characteristics based on health information, behavioral information, and identity information.

3. The method for constructing a family user characteristic database according to claim 2, characterized in that, In step S21, family users are numbered according to their identity information. Each family user is numbered in descending order of age. The family user number is set as i. Different families are also numbered and the family number is set as j. In step S22, the number of family users in the same family is counted as the family size, and the family size is set as NP(j); the frequency of each food preference in the same family is counted, and the food preference corresponding to the most frequent food preference is taken as the family food preference. If the food preferences of each user in the same family are different, the family food preference is set to no preference. The frequency of each type of exercise preference within the same household is counted. The type of exercise preference with the highest frequency is taken as the household exercise preference. If the exercise preference types of each user in the same household are different, the household exercise preference is set to no preference. The household meal times are set to be the same as the users' meal times. In step S23, the daily characteristics of the family users are analyzed based on their height, weight, blood pressure, exercise time, sleep time, wake-up time and name to obtain the daily characteristics of the users D(j,i).

Images