Nursing product pushing method and system based on address features

Abstract

The invention discloses a nursing product pushing method and system based on address features, and relates to the technical field of artificial intelligence. According to the nursing product pushing method and system based on the address features, real-time address data of a user is obtained through a geographic information system, and the address region features of the user are extracted in combination with a region division algorithm. And deep matching is carried out on the geographic features of the user and the nursing product demand data through a clustering algorithm, proper nursing products are screened, and a recommendation list is generated. And the pushing priority of the nursing products is evaluated according to the matching strength, and the pushing sequence and frequency are adjusted in combination with the user behavior feedback data, so that the pushing accuracy and the user satisfaction are improved. According to the method, the individuation and timeliness of nursing product recommendation are effectively improved, the interference of irrelevant pushing is reduced, the accuracy and practicability of a traditional pushing mode are remarkably improved, the user experience is enhanced, and the market demand matching efficiency of nursing products is optimized.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for pushing nursing products based on address features. Background Art

[0002] In today's society, with the improvement of residents' health awareness, the demand for nursing products has increased significantly. Especially in terms of the nursing needs of the elderly, special patients and people with limited mobility, the market demand for accurate and personalized nursing product push is increasing. At the same time, the popularization of smart devices and data analysis technology has provided the possibility for personalized recommendation of nursing products. More and more service platforms use users' basic information, health status and geographical location to try to push nursing products that meet their needs to users at the right time and place. However, traditional push methods mostly rely on users' basic data, lack of sufficient consideration for users' real-time location, geographical environment or timeliness of needs, which affects the accuracy and practicality of push.

[0003] The push model of traditional care products lacks in-depth mining of user living environment and geographic location information, resulting in generalized push content and difficulty in meeting personalized care needs. Secondly, most existing push methods are based on users' health data or purchase records, and lack effective analysis methods for the dynamic changes in their geographic location or product demand preferences in specific regions, resulting in recommended care products that do not match actual usage scenarios. In addition, existing systems usually lack precise control over the timing of push, and are unable to make corresponding adjustments based on the user's real-time location or differences in demand during specific time periods, reducing the timeliness of push and user satisfaction. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a method and system for pushing care products based on address features, which solves the problems of the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for pushing nursing products based on address features, comprising the following steps: S1. obtaining real-time address data of users through a geographic information system, and extracting regional features of user addresses in combination with a regional division algorithm; S2. based on regional features of user addresses, deeply correlating and matching regional features of user addresses with geographic regional demand data of nursing products through a clustering algorithm, screening nursing products, and generating a list of recommended nursing products; S3. evaluating the priority of pushing nursing products and obtaining a push index of nursing products based on the matching strength between regional features of user addresses and geographic regional demand data of nursing products; S4. prioritizing the list of recommended nursing products based on the push index of nursing products; S5. monitoring user behavior feedback data, and regulating the order and frequency of pushing nursing products.

[0006] Furthermore, the specific process of extracting user address area characteristics through the regional division algorithm is as follows: extract the user's geographic coordinates according to the geographic information system, spatially cluster the user's geographic coordinates through the regional division algorithm, and divide the user's geographic location into corresponding geographic area units; based on the division results, extract the socioeconomic characteristics and medical resource distribution characteristics of each geographic area, and construct a regional feature vector; match the user's geographic area according to the regional feature vector to generate the user's address area characteristics.

[0007] Furthermore, the specific process of deeply associating and matching user address area characteristics with the geographical area demand data of nursing products through clustering algorithms is as follows: based on user address area characteristics and the geographical area demand data of nursing products, construct a user demand feature matrix and a product demand feature matrix; jointly cluster user address area characteristics and nursing product demand characteristics through clustering algorithms, associate and match users and nursing products in the geographical area demand dimension, analyze the similarity between user geographical characteristics and nursing product demand data, and assign users and nursing products with similar demand patterns to the same cluster.

[0008] Furthermore, the specific process of screening nursing products and generating a recommended list of nursing products is as follows: based on the joint clustering results, determine the address area characteristics of each user and the corresponding type of nursing product demand; based on the cluster group to which each user belongs, screen out nursing products that meet the needs of the group and evaluate the matching degree of each nursing product; prioritize the screened nursing products and generate a recommended list of nursing products.

[0009] Furthermore, according to the matching strength between the user address area characteristics and the nursing product geographical area demand data, the specific process of evaluating the nursing product push priority is as follows: Set the user address area characteristics as , the geographical demand characteristics of care products are ,in and Respectively represent users Address area characteristics and care products Demand feature vector; Calculate the matching strength between each care product and the user ,The matching strength formula is as follows: ;in, Represents a computing user and care products Cosine similarity function on the geographic area demand dimension.

[0010] Furthermore, the specific process of obtaining the nursing product push index is as follows: calculate the matching strength between each nursing product and all users, select the best matching product corresponding to each user, and sort all nursing products according to the matching strength to obtain the nursing product push index of each nursing product. , where the formula for the nursing product push index is: ;in, For users The weighted coefficient represents the relative importance of the user's demand for care products. For users and care products The matching strength.

[0011] Furthermore, the specific process of prioritizing the care product recommendation list is as follows: based on the care product push index, the care products are sorted in descending order from large to small; for care products with the same push index, secondary sorting is performed according to user behavior feedback data.

[0012] Furthermore, the specific process of monitoring user behavior feedback data and regulating the push order and frequency of care products is as follows: based on the user's behavior feedback data on the pushed care products, including click volume, purchase rate, and stay time, calculate the behavior feedback index of each care product; based on the behavior feedback index of each care product, compare it with the set behavior feedback index threshold range; when the behavior feedback index is greater than or equal to the set feedback index threshold range, increase its push frequency; when the behavior feedback index is less than the set feedback index threshold range, reduce its push frequency and adjust the push order.

[0013] The nursing product push system based on address features includes the following modules: an address feature extraction module, a cluster matching module, a priority evaluation module, a sorting module, and a behavior feedback control module; the address feature extraction module is used to obtain the user's real-time address data through a geographic information system, and extract the user's address area features in combination with a regional division algorithm; the cluster matching module is used to perform deep correlation matching between the user's address area features and the nursing product geographic area demand data based on the user's address area features through a clustering algorithm, screen nursing products, and generate a nursing product recommendation list; the priority evaluation module is used to evaluate the nursing product push priority and obtain the nursing product push index based on the matching strength between the user's address area features and the nursing product geographic area demand data; the sorting module is used to prioritize the nursing product recommendation list based on the nursing product push index; the behavior feedback control module is used to monitor the user's behavior feedback data and regulate the push order and frequency of nursing products.

[0014] The present invention has the following beneficial effects:

[0015] (1) This method of pushing care products based on address features achieves accurate product recommendations by combining the user's address area features with the geographical area demand data of care products. It improves the matching degree of care products, avoids invalid or irrelevant pushes, and optimizes the user experience. By matching users and products through clustering algorithms, it is possible to quickly identify user groups with similar demand patterns in big data, ensure that the pushed care products meet the demand characteristics of specific regions, and improve the relevance and usage rate of products. By calculating the matching strength between users and care products, reasonably evaluating the push priority, and sorting according to the care product push index, the push content is more in line with the actual needs of users, improving the accuracy and effectiveness of the push.

[0016] (2) This address-based care product push system monitors user behavior feedback data and adjusts the order and frequency of care product push in real time, so that the recommendation system can dynamically adjust according to the user's actual feedback, avoiding excessive push or push of inappropriate products, thereby improving user satisfaction and system adaptability. Combining geographic information and user behavior analysis, it can more quickly identify and respond to changes in user needs. The system can optimize the recommendation strategy in a timely manner, reduce user waiting time, and improve user trust and satisfaction with the system.

[0017] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of the method for pushing care products based on address features of the present invention.

[0019] Figure 2 This is a flow chart of the care product push system based on address features of the present invention. DETAILED DESCRIPTION

[0020] The embodiment of the present application solves the problem that the traditional push mode ignores the user's geographical location and the timeliness of demand through a nursing product push method and system based on address features, thereby achieving more accurate and personalized nursing product recommendations.

[0021] The overall idea of ​​the problem in the embodiment of this application is as follows:

[0022] The user's real-time address data is obtained through the geographic information system, and the user's address area characteristics are extracted in combination with the regional division algorithm.

[0023] Based on the user's address area characteristics, a clustering algorithm is used to deeply correlate and match the user's address area characteristics with the geographical area demand data of nursing products, screen nursing products, and generate a recommended list of nursing products.

[0024] According to the matching strength between the user's address area characteristics and the nursing product geographical area demand data, the nursing product push priority is evaluated and the nursing product push index is obtained.

[0025] Based on the nursing product push index, the nursing product recommendation list is prioritized.

[0026] Monitor user behavior feedback data and adjust the order and frequency of pushing care products.

[0027] See also Figure 1 , an embodiment of the present invention provides a technical solution: a method for pushing nursing products based on address features, comprising the following steps: S1. obtaining real-time address data of users through a geographic information system, and extracting regional features of user addresses in combination with a regional division algorithm; S2. based on regional features of user addresses, deeply correlating and matching regional features of user addresses with geographic regional demand data of nursing products through a clustering algorithm, screening nursing products, and generating a recommended list of nursing products; S3. evaluating the push priority of nursing products and obtaining a push index of nursing products based on the matching strength between regional features of user addresses and geographic regional demand data of nursing products; S4. prioritizing the recommended list of nursing products based on the push index of nursing products; S5. monitoring user behavior feedback data, and regulating the push order and frequency of nursing products.

[0028] In this implementation scheme, S1. Geographic Information System is a technology used to capture, store, analyze and manage geospatial data. It can help obtain the user's real-time address data (such as GPS coordinates or other geolocation information). Regional division algorithm: The function of this algorithm is to spatially divide the user's geographic location data to generate different geographic area units. Through division, characteristic information of each area can be extracted, such as population density, traffic conditions, climate conditions, etc. User address area characteristics: The characteristics extracted from the regional division algorithm include the geographical location of the user's area, environmental factors, and socioeconomic characteristics of the area. S2. Clustering algorithm: Clustering algorithm is an unsupervised learning algorithm used to group elements in a data set according to certain characteristics, so that the similarity of elements in the same group is high, while the similarity between different groups is low. Here, the clustering algorithm is used to analyze the relationship between the user's geographic area characteristics and the care product demand characteristics. Care product geographic area demand data: refers to information such as the demand, preference, and usage of care products in different geographic areas. For example, some areas may have a high demand for specific types of care products. Deep association matching: refers to a comprehensive analysis of the user's geographic characteristics and care product needs through a clustering algorithm to find out the similarities between the user and the care product needs. Based on this analysis, the most suitable care products for the user can be screened out and a care product recommendation list can be generated. S3. Matching strength: Matching strength refers to the similarity or correlation between the user's geographical area characteristics and the care product demand characteristics. Similarity measurement methods can be used to quantify this matching degree. Push priority is to rank the importance of each care product according to the matching strength. The higher the matching strength of the care product, the higher its push priority. Care product push index: The care product push index is a comprehensive indicator used to measure the push priority of a care product in front of a specific user. The higher the push index, the higher the push priority of the product. S4. Sort the recommended care products by the calculated care product push index. The higher the push index of the product, the higher the priority of the product at the front of the recommendation list, thereby ensuring that the user sees the product that best meets his needs first. S5. User behavior feedback data: User behavior feedback data includes the interaction data between users and recommended products, such as click-through rate, purchase rate, browsing time, feedback evaluation, etc. Based on these behavioral data, the user's acceptance and interest in the pushed products can be analyzed. Push order and frequency regulation: Based on the monitored user behavior data, the push order and frequency of products can be dynamically adjusted. For example, if the push effect of a certain care product is poor, you can reduce its push frequency or optimize the push content.

[0029] Specifically, the specific process of extracting user address area characteristics through the regional division algorithm is as follows: extract the user's geographic coordinates according to the geographic information system, spatially cluster the user's geographic coordinates through the regional division algorithm, and divide the user's geographic location into corresponding geographic area units; based on the division results, extract the socioeconomic characteristics and medical resource distribution characteristics of each geographic area, and construct a regional feature vector; match the user's geographic area according to the regional feature vector to generate the user's address area characteristics.

[0030] In this implementation scheme, a geographic information system (GIS) is used to obtain the user's real-time geographic coordinates (such as longitude and latitude). These coordinates are an accurate representation of the user's current location. The user's geographic location is the basis for accurately pushing care products. Through geographic coordinates, the system can know the user's actual location and further analyze the characteristics of the area where the user is located (such as environment, socio-economics, and medical resources) to prepare for subsequent push. The regional division algorithm groups geographic coordinate data (user location) by spatial distance or similarity to form different geographic area units. Mapping the geographic locations of all users into a specific area can provide aggregated data for subsequent analysis. For example, users in a certain area may face similar socio-economic and medical resource needs, and regional division helps to find user groups with similar needs. Based on the distribution of users in each geographic area, the socio-economic characteristics (such as income level, population density, education level, etc.) and medical resource characteristics (such as the number of hospitals and clinics, accessibility of medical services, etc.) of the area are extracted. These characteristics help reflect the care needs of residents in the area. The above-extracted regional characteristics (such as socio-economics, medical resources, etc.) are converted into a "regional feature vector", which is a set of digital values ​​used to describe the overall characteristics of the area. Each feature (such as income level, number of medical facilities, etc.) is converted into a part of the vector through a certain quantitative method. The regional feature vector represents the complex information of the region (social, economic, medical, etc.) in a simplified way, so that subsequent calculations (such as matching users with regional features, recommendation push, etc.) can be carried out efficiently. The feature vector provides the system with a standardized and unified way to represent regional characteristics, which is convenient for comparing and analyzing different regions. By matching the user's geographical location with the existing regional feature vector, the user's geographical area characteristics are determined, and the geographical area that best meets their needs is identified. By matching regional characteristics, the system can accurately identify the specific needs of the user's region and provide personalized recommendations. This process ensures the regional adaptability and demand accuracy of the pushed products.

[0031] Specifically, the specific process of deeply associating and matching user address area characteristics with the geographical area demand data of nursing products through clustering algorithms is as follows: based on user address area characteristics and the geographical area demand data of nursing products, construct a user demand feature matrix and a product demand feature matrix; jointly cluster user address area characteristics and nursing product demand characteristics through clustering algorithms, associate and match users and nursing products in the geographical area demand dimension, analyze the similarity between user geographical characteristics and nursing product demand data, and assign users and nursing products with similar demand patterns to the same cluster.

[0032] In this implementation scheme, the user demand feature matrix: the geographical characteristics of each user (income level, health status, distribution of medical resources) are converted into numerical feature vectors and constructed into a matrix form, with each row representing a user and each column representing a feature. Product demand feature matrix: the demand characteristics of each care product in different geographical regions (demand intensity, frequency of use) are converted into numerical feature vectors and constructed into a matrix form, with each row representing a product and each column representing a demand feature. By quantifying the geographical demand characteristics of users and products, a standardized data basis is provided for subsequent clustering analysis, enabling the system to match and recommend under a unified framework. The user demand feature matrix and the product demand feature matrix are merged, and each user and each product constitutes a high-dimensional feature vector through its geographical characteristics and demand characteristics. The clustering algorithm (K-means) clusters these vectors and classifies users and products with similar needs into one category. By identifying the similarities between user and product demand characteristics, the clustering algorithm can group users and care products with similar demand patterns, ensuring that the products matched to each user meet their geographical region and demand characteristics, thereby improving the relevance and accuracy of the recommendation. Based on the clustering results, the system selects care products suitable for the group according to the user's cluster group and the group's demand characteristics, and matches them with the corresponding users. Through this matching method, the system can ensure that the recommended care products not only meet the basic needs of users, but also adapt to the specific needs of their region, avoid pushing inappropriate products, and improve the accuracy of the recommendation system. In the joint clustering process, the system calculates the similarity between each user and the care product, analyzes the differences in the feature vectors of users and products in different clusters, and identifies which products are most in line with which users' needs. This similarity analysis can help the system accurately identify the best matching relationship between users and products, thereby optimizing the push strategy, improving user satisfaction and the effectiveness of care product recommendations.

[0033] Specifically, the specific process of screening nursing products and generating a recommended list of nursing products is as follows: based on the joint clustering results, determine the address area characteristics of each user and the corresponding type of nursing product demand; based on the cluster group to which each user belongs, screen out nursing products that meet the needs of the group and evaluate the matching degree of each nursing product; prioritize the screened nursing products and generate a recommended list of nursing products.

[0034] In this embodiment, after the joint cluster analysis is completed, each user is assigned to a specific cluster group according to the characteristics of his address area (geographic location, socioeconomic conditions, distribution of medical resources, etc.). Each cluster group represents a class of users with similar demand characteristics (high-demand care products in the same area). The system determines the type of care products related to the group based on the demand characteristics of the cluster group. For example, a cluster group may need more care products for the elderly group, while another group may prefer care products for specific diseases. The clustering results are used to clarify the type of care product demand for each user's group. This enables the system to analyze the user's demand characteristics as a whole and push personalized products, rather than simply recommending based on a single user's characteristics. The care products that are usually needed by each user are screened out according to the cluster group to which each user belongs. For example, for the elderly group, the system will give priority to screening out care products that meet the needs of the elderly, such as assistive devices, elderly care beds, etc. The matching degree between each user and the care product is quantitatively evaluated, and commonly used methods include scoring models based on geographical needs, product types, and historical needs. The matching degree evaluation formula can be comprehensively scored based on factors such as geographical area, product function requirements, and user health status. For example, if a user's area has a strong demand for a certain care product, and the user's health condition requires such a product, the matching degree is relatively high. Products that match user characteristics are screened through group demand. Care products that meet user needs are accurately screened to ensure that the recommended care products are more in line with user needs, thereby improving the effectiveness of the recommendation. According to the matching score between each care product and the user, the system sorts the screened care products. Products with high matching scores are ranked first, and products with low scores are ranked last. The core purpose of this sorting process is to ensure that the system can recommend the care products that best meet the needs and have the highest matching degree to users. Through priority sorting, the products in the recommendation list will be accurately arranged according to the user's needs, increasing the user's acceptance and satisfaction with the recommendation results.

[0035] Specifically, according to the matching strength between the user's address area characteristics and the nursing product geographical area demand data, the specific process of evaluating the priority of nursing product push is as follows: Set the user's address area characteristics as , the geographical demand characteristics of care products are ,in and Respectively represent users Address area characteristics and care products Demand feature vector; Calculate the matching strength between each care product and the user ,The matching strength formula is as follows: ;in, Represents a computing user and care products Cosine similarity function on the geographic area demand dimension.

[0036] In this implementation scheme, the matching strength is calculated as follows: based on the user's geographical area characteristics and the geographical demand characteristics of the care products, the system calculates the matching strength between each care product and the user. The matching strength reflects the degree of fit between the user's needs and the characteristics of the care product in terms of geographical area needs. By calculating this matching strength, the system can determine which care products best meet the needs of the user's area. For each care product, the system calculates its matching strength with all users and sorts the products according to the matching strength. Based on the relative importance of each user's demand for the care product, the system assigns a push index to each product. The push index indicates the priority of each care product relative to the user's needs, helping the system determine which products should be pushed to the user first.

[0037] Specifically, the specific process of obtaining the nursing product push index is as follows: calculate the matching strength between each nursing product and all users, select the best matching product corresponding to each user, and sort all nursing products according to the matching strength to obtain the nursing product push index of each nursing product. , where the formula for the nursing product push index is: ;in, For users The weighted coefficient represents the relative importance of the user's demand for care products. For users and care products The matching strength.

[0038] In this implementation scheme, the calculation of matching strength: the matching strength of each care product with all users is calculated. The matching strength reflects the degree of fit between each care product and user needs, taking into account the similarity between geographical area characteristics and care product needs. Select the best matching product: for each user, the system selects the care product that best matches the user's needs. This process ensures that the pushed product each user receives is the most relevant product in terms of their needs. Sorting care products: all care products are sorted according to the matching strength with the user, with products with higher matching strength at the front. The system assigns a push index to each care product, which reflects the priority of the care product for all users. The role of the weighting coefficient: in the calculation of the push index, the user's weighting coefficient plays a key role. This coefficient indicates the importance of the user's demand for a certain care product. The weighting coefficient can be adjusted according to the user's specific needs or historical behavior, so that products that are more demanding for specific users have a higher push priority.

[0039] Specifically, the specific process of prioritizing the care product recommendation list is as follows: based on the care product push index, the care products are sorted in descending order from large to small; for care products with the same push index, secondary sorting is performed according to user behavior feedback data.

[0040] In this implementation scheme, preliminary sorting: sorting is based on the push index of each care product. The push index reflects the matching strength of the care product and the user's needs. Care products with a higher push index are considered to be more relevant to the user's needs and are therefore recommended first. Therefore, the system will sort the care products in descending order according to the push index from large to small, ensuring that the products that best meet the user's needs are ranked first. Secondary sorting: For care products with the same push index, the system performs secondary sorting by analyzing the user's behavioral feedback data. Behavioral feedback data includes information such as the user's interaction with the care product, purchase history, and click-through rate. These data can help the system further determine which products are more attractive or in demand for the user, thereby adjusting the order of recommendations.

[0041] Specifically, the specific process of monitoring user behavior feedback data and regulating the push order and frequency of care products is as follows: based on the user's behavior feedback data on the pushed care products, including click volume, purchase rate, and stay time, calculate the behavior feedback index of each care product; based on the behavior feedback index of each care product, compare it with the set behavior feedback index threshold range; when the behavior feedback index is greater than or equal to the set feedback index threshold range, increase its push frequency; when the behavior feedback index is less than the set feedback index threshold range, reduce its push frequency and adjust the push order.

[0042] In this implementation plan, behavioral feedback data collection: The system monitors the user's interactive data on the pushed care products, mainly including: Click volume: The number of times a user clicks on a care product, reflecting the initial interest in the product. Purchase rate: The user's actual purchase behavior of the pushed product directly reflects the attractiveness and practicality of the product. Stay time: The user's stay time when browsing the product shows the user's attention to the product and the depth of interest. The formula for calculating the behavioral feedback index of each care product is as follows: in: : Behavioral feedback index of the ath care product. :No. The number of clicks on the ath care product by a user. :No. The purchase rate of the ath care product by users. :No. The length of time a user stays on the ath care product. :No. The number of times a user collects the ath care product. : Indicates the timestamp when product a is pushed. : Represents the weighted coefficients of click volume, purchase rate, stay time and number of favorites. The weight factor of each user can be adjusted according to the priority and importance of their behavior. m: Represents the total number of users. The behavior data of all users will affect the calculation of the push index. : represents the time weighted attenuation factor, is the time decay coefficient, which reflects the impact of time on the push index. As time goes by, the effectiveness of push may decrease, so the decay function is used.

[0043] See also Figure 2 The nursing product push system based on address features includes the following modules: address feature extraction module, cluster matching module, priority evaluation module, sorting module, and behavior feedback control module; the address feature extraction module is used to obtain the user's real-time address data through the geographic information system, and extract the user's address area features in combination with the regional division algorithm; the cluster matching module is used to perform deep correlation matching between the user's address area features and the nursing product geographic area demand data based on the user's address area features through the clustering algorithm, screen nursing products, and generate a nursing product recommendation list; the priority evaluation module is used to evaluate the nursing product push priority and obtain the nursing product push index based on the matching strength between the user's address area features and the nursing product geographic area demand data; the sorting module is used to prioritize the nursing product recommendation list based on the nursing product push index; the behavior feedback control module is used to monitor the user's behavior feedback data and regulate the push order and frequency of nursing products.

[0044] In this implementation scheme, the address feature extraction module: obtains the user's real-time address data through the geographic information system, and extracts the user's geographic area characteristics to provide basic data for subsequent matching. Cluster matching module: uses a clustering algorithm to match the user's address area characteristics with the geographical needs of care products, and screens out care products that meet user needs. Priority evaluation module: evaluates and calculates the push priority of each care product based on the matching degree between the user's geographic area characteristics and the care product needs. Sorting module: Sorts care products according to their push priority to ensure that the products that best meet the needs are displayed first. Behavioral feedback regulation module: monitors user behavioral feedback data, adjusts the push order and frequency of care products to optimize the push effect.

[0045] In summary, this application has at least the following effects:

[0046] The method and system for pushing nursing products based on address features provide nursing product recommendations that better meet the actual needs of users by combining the user's geographic location and regional needs, thereby improving the accuracy of push notifications. Dynamically adjust the push order and frequency based on user behavior feedback to avoid excessive push notifications, thereby improving user satisfaction and the actual use value of the product. Utilize the deep matching of geographic information and regional demand data to provide personalized nursing product recommendations for different users to meet diverse nursing needs. Through real-time monitoring and push strategy adjustment, ensure that the push notifications of nursing products are more timely and meet the user's time needs. Through precise matching and efficient sorting, reduce unnecessary push notifications, optimize resource allocation, and improve the overall efficiency of the system.

[0047] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0049] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0050] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0051] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0052] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for pushing care products based on address features, characterized in that: The following steps are involved: S1. Obtain the user's real-time address data through the geographic information system, and extract the user's address area characteristics in combination with the regional division algorithm; S2. Based on the user's address area characteristics, the clustering algorithm is used to deeply correlate and match the user's address area characteristics with the geographical area demand data of nursing products, screen nursing products, and generate a recommended list of nursing products; S3. Evaluate the priority of nursing product push and obtain the nursing product push index based on the matching strength between the user's address area characteristics and the nursing product geographical area demand data; S4. Prioritize the recommended list of nursing products based on the nursing product push index; S5. Monitor user behavior feedback data and adjust the order and frequency of pushing care products.

2. The method for pushing care products based on address features according to claim 1, characterized in that: The specific process of extracting user address area features through the area division algorithm is as follows: Extract the user's geographic coordinates from the geographic information system, perform spatial clustering on the user's geographic coordinates through a regional division algorithm, and divide the user's geographic location into corresponding geographic area units; Based on the division results, the socioeconomic characteristics and medical resource distribution characteristics of each geographical area are extracted to construct a regional feature vector; The user's geographical area is matched according to the regional feature vector to generate the user address regional feature.

3. The method for pushing care products based on address features according to claim 2, characterized in that: The specific process of deep correlation and matching between user address area characteristics and geographic area demand data of care products through clustering algorithm is as follows: Based on the user address area characteristics and the geographical area demand data of nursing products, construct the user demand feature matrix and the product demand feature matrix; The user address area characteristics and care product demand characteristics are jointly clustered through a clustering algorithm, and users and care products are associated and matched in the dimension of geographical area demand. The similarity between user geographical characteristics and care product demand data is analyzed, and users and care products with similar demand patterns are assigned to the same cluster.

4. The method for pushing care products based on address features according to claim 3, characterized in that: The specific process of screening care products and generating a recommended list of care products is as follows: According to the joint clustering results, determine the address area characteristics of each user and the corresponding type of care product demand; According to the cluster group of each user, select the care products that meet the needs of the group and evaluate the matching degree of each care product; The screened care products are prioritized and a recommended care product list is generated.

5. The method for pushing care products based on address features according to claim 4, characterized in that: The specific process of evaluating the priority of nursing product push based on the matching strength between the user's address area characteristics and the geographical area demand data of nursing products is as follows: Set the user address area characteristics to , the geographical demand characteristics of care products are ,in and Respectively represent users Address area characteristics and care products The demand feature vector of Calculate the matching strength between each care product and the user ,The matching strength formula is as follows: ;in, Represents a computing user and care products Cosine similarity function on the geographic area demand dimension.

6. The method for pushing care products based on address features according to claim 5, characterized in that: The specific process of obtaining the nursing product push index is as follows: Calculate the matching strength between each care product and all users, select the best matching product for each user, and sort all care products according to the matching strength to obtain the care product push index for each care product , where the formula for the nursing product push index is: ;in, For users The weighted coefficient represents the relative importance of the user's demand for care products. For users and care products The matching strength.

7. The method for pushing care products based on address features according to claim 6, characterized in that: The specific process for prioritizing the recommended list of care products is as follows: Based on the nursing product push index, the nursing products are sorted in descending order from largest to smallest; For care products with the same push index, secondary sorting is performed based on user behavior feedback data.

8. The method for pushing care products based on address features according to claim 7, characterized in that: The specific process of monitoring user behavior feedback data and adjusting the order and frequency of pushing care products is as follows: Calculate the behavioral feedback index of each care product based on the user's behavioral feedback data on the pushed care products, including click volume, purchase rate, and stay time; Based on the behavioral feedback index of each care product, it is compared with the set behavioral feedback index threshold range. When the behavioral feedback index is greater than or equal to the set feedback index threshold range, its push frequency is increased. When the behavioral feedback index is less than the set feedback index threshold range, its push frequency is reduced and the push order is adjusted.

9. A system for pushing care products based on address features, using a method for pushing care products based on address features as claimed in any one of claims 1 to 8, characterized in that: It includes the following modules: address feature extraction module, cluster matching module, priority evaluation module, sorting module, and behavior feedback control module; The address feature extraction module is used to obtain the user's real-time address data through the geographic information system, and extract the user's address area features in combination with the area division algorithm; The cluster matching module is used to perform deep correlation matching between the user address area characteristics and the nursing product geographical area demand data based on the user address area characteristics through a clustering algorithm, screen nursing products, and generate a nursing product recommendation list; The priority evaluation module is used to evaluate the priority of nursing product push and obtain the nursing product push index according to the matching strength between the user address area characteristics and the nursing product geographical area demand data; The sorting module is used to prioritize the nursing product recommendation list based on the nursing product push index; The behavior feedback control module is used to monitor user behavior feedback data and control the push order and frequency of care products.

Images