Collaborative elderly care service method and system based on artificial intelligence and big data analysis

Abstract

The present invention relates to the field of artificial intelligence technology, and in particular to a collaborative elderly care service method and system based on artificial intelligence and big data analysis. The method collects feature data of multiple elderly people to form a feature data set; performs secondary clustering on the feature data set to obtain multiple initial clustering areas and multiple discrete feature data points by performing primary density clustering on the feature data set; performs similarity matching on the feature data in each clustering area with the historical service data corresponding to each elderly care service type, and recommends the elderly care service type corresponding to each elderly person based on the matching results. Secondary clustering based on the local discrete degree corresponding to the discrete feature data can improve the clustering accuracy of clustering, avoid the discrete feature data being identified as abnormal data due to parameter settings, and thus solve the problem of being unable to accurately recommend elderly care services for the elderly corresponding to the isolated data.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a collaborative elderly care service method and system based on artificial intelligence and big data analysis. Background Art

[0002] As the aging of society becomes increasingly serious, the traditional elderly care model cannot fully meet the increasing elderly care needs. The use of big data and artificial intelligence provides new paths and new options for smart elderly care services. At present, the elderly care services based on big data analysis mainly use intelligent electronic devices to obtain the elderly’s daily behavior, health, and home life data, and use classification algorithms such as random forests to conduct a comprehensive analysis of all aspects of the elderly’s data, so as to provide collaborative elderly care services based on the analysis results.

[0003] Technical personnel in this field have found that the existing technology has the following problems: However, when all aspects of the elderly's data are special compared to other people, the mapping will be an isolated data in the classification result, and the isolated data will affect the accuracy of collaborative elderly care services, and thus will lead to the inability to obtain accurate elderly care service recommendations. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a collaborative elderly care service method and system based on artificial intelligence and big data analysis. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides a collaborative elderly care service method based on artificial intelligence and big data analysis, the method comprising the following specific steps:

[0006] Collecting characteristic data of a plurality of elderly people to form a characteristic data set, wherein the characteristic data includes health data, daily life data and personal hobby data;

[0007] Density clustering is performed on the feature data set to obtain a plurality of initial clustering regions and a plurality of discrete feature data points, and the local discreteness of each discrete feature data point relative to each initial clustering region is obtained in combination with the distance between the discrete feature data point and each feature data point in the initial clustering region; the discrete feature data points are re-divided into the initial clustering regions according to the local discreteness to obtain a plurality of clustering regions after secondary clustering;

[0008] The characteristic data in each of the clustering areas is matched with the historical service data corresponding to each elderly care service type respectively, and the elderly care service type corresponding to each elderly person is recommended according to the matching result.

[0009] Furthermore, the method for determining the local discreteness of each discrete feature data point relative to each initial clustering region comprises:

[0010] Calculating the actual clustering density of the current initial clustering area according to the positional relationship between all feature data points in the initial clustering area;

[0011] The local clustering density of the current discrete feature data point relative to the current initial clustering region is calculated, and the local discrete degree of the current discrete feature data point relative to the current initial clustering region is obtained from the ratio between the local clustering density and the actual clustering density.

[0012] Furthermore, the method for obtaining the actual cluster density includes:

[0013] Calculate the Euclidean distances between any feature data point and other feature data points in the current initial clustering area, obtain the Euclidean distance set corresponding to the feature data point, and take the maximum Euclidean distance in the Euclidean distance set as the target distance of the feature data point;

[0014] The target distances of all feature data points in the current initial clustering region are added to obtain the sum of the first distances, and the actual clustering density of the current initial clustering region is obtained by combining the sum of the first distances with the total number of the feature data points contained in the current initial clustering region.

[0015] Furthermore, the method for calculating the local clustering density of the current discrete feature data point relative to the current initial clustering area includes:

[0016] The distance between the current discrete feature data point and each feature data point in the current initial clustering area is calculated to obtain the sum of the second distances, and the local clustering density of the current discrete feature data point relative to the current initial clustering area is obtained by combining the sum of the second distances with the total number of the feature data points contained in the current initial clustering area.

[0017] Furthermore, the local clustering density is negatively correlated with the sum of the second distances, and the local clustering density is positively correlated with the total number.

[0018] Furthermore, the density clustering is a DBSCAN algorithm.

[0019] Furthermore, an embodiment of the present invention also provides a collaborative elderly care service system based on artificial intelligence and big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that when the processor executes the computer program, it implements the steps of any one of the above-mentioned collaborative elderly care service methods based on artificial intelligence and big data analysis.

[0020] The embodiments of the present invention have at least the following beneficial effects: secondary clustering based on the local discrete degree corresponding to discrete feature data can improve the clustering accuracy of clustering, avoid discrete feature data being identified as abnormal data due to parameter settings, and thus solve the problem of being unable to accurately recommend elderly care services for the elderly corresponding to isolated data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A flowchart of the steps of a collaborative elderly care service method based on artificial intelligence and big data analysis is provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of a collaborative elderly care service method and system based on artificial intelligence and big data analysis proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0024] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0025] The following is a detailed description of a specific scheme of a collaborative elderly care service method and system based on artificial intelligence and big data analysis provided by the present invention in conjunction with the accompanying drawings.

[0026] See also Figure 1 , which shows a flowchart of a collaborative elderly care service method based on artificial intelligence and big data analysis provided by an embodiment of the present invention, the method comprising the following steps:

[0027] Step S001, collecting characteristic data of multiple elderly people to form a characteristic data set, the characteristic data including health data, daily life data and personal hobby data.

[0028] Specifically, the health data of the elderly are obtained through health monitoring equipment, such as sleep quality, heart rate, blood pressure, blood sugar, body temperature, respiratory rate, etc.; the daily life data of the elderly are captured through community specialist statistics or surveillance cameras. Daily life data refers to food, clothing, housing and transportation, such as daily grocery shopping, special grocery delivery, walking, frequency of relatives' visits, communications, geographic location, etc.; similarly, the personal hobby data of the elderly are obtained through community specialist statistics or surveillance cameras, such as various sports, chess, cards, dancing, etc. The health data, daily life data and personal hobby data constitute the characteristic data of the elderly. One elderly person corresponds to one characteristic data, and the characteristic data of multiple elderly people are obtained to constitute a characteristic data set.

[0029] In the process of acquiring feature data, some feature data may be missing due to various situations, and then data filling is performed for the missing data of the elderly's feature data. Adaptive data missing filling is performed according to the feature data distribution under the missing data dimension. Among them, commonly used data filling methods include: Lagrange interpolation method, random interpolation method, mean median interpolation method, Bayesian prediction, etc.

[0030] Step S002, density clustering is performed on the feature data set to obtain multiple initial clustering regions and multiple discrete feature data points, and the local discreteness of each discrete feature data point is obtained in combination with the distance between the discrete feature data point and each feature data point in the initial clustering region; the discrete feature data points are re-divided into the initial clustering regions according to the local discreteness to obtain multiple clustering regions after secondary clustering.

[0031] Specifically, the feature data in the feature data set are density clustered to obtain the initial clustering area. The density clustering adopts the DBSCAN algorithm, which is as follows: take any feature data as the cluster center, set the initial clustering radius, and obtain multiple sub-clustering areas according to the initial clustering radius; set the Minpts threshold, when the number of feature data points in the sub-clustering area is less than the Minpts threshold, the sub-clustering area is considered to be a boundary area, when the number of feature data points in the sub-clustering area is greater than or equal to the Minpts threshold, continue to traverse other sub-clustering areas in its neighborhood until the boundary area, and then cluster these sub-clustering areas into an initial clustering area; traverse all feature data in the feature data set to complete the initial clustering.

[0032] Since the characteristic data of the elderly is more special than that of other people, after the initial clustering, not only k initial clustering regions will be obtained, but also multiple discrete characteristic data points will be obtained. The discrete characteristic data points are data points that do not belong to the initial clustering regions, also known as outliers. In order to allow these discrete characteristic data points to participate in clustering, the embodiment of the present invention performs adaptive secondary clustering based on the distribution relationship between the initial clustering regions and the discrete points after the initial clustering. The process of secondary clustering is as follows:

[0033] (1) Calculate the actual clustering density of each initial clustering area based on the positional relationship between all feature data points in the initial clustering area.

[0034] Specifically, taking the kth initial clustering region as an example, the Euclidean distances between the pth feature data point and other feature data points in the region are calculated to obtain the Euclidean distance set corresponding to the pth feature data point, and the maximum Euclidean distance in the Euclidean distance set is taken as the target distance of the pth feature data point; similarly, the target distance corresponding to each feature data point in the kth initial clustering region is obtained to obtain the sum of the first distances, and the actual clustering density of the initial clustering region is obtained by combining the sum of the first distances and the total number of feature data points contained in the kth initial clustering region: Among them, i is the i-th other feature data point; Num k is the total number of feature data points contained in the kth initial clustering area; is the sum of the first distances.

[0035] (2) The local discreteness of each discrete feature data point relative to each initial clustering region is obtained from the distance between the discrete feature data point and each feature data point in the initial clustering region.

[0036] Specifically, taking the kth initial clustering region as an example, the local discrete degree of the ath discrete feature data point relative to it is calculated respectively: the distance between the ath discrete feature data point and each feature data point in the kth initial clustering region is calculated to obtain the sum of the second distances, and the local clustering density corresponding to the ath discrete feature data point is obtained by combining the sum of the second distances with the total number of feature data points contained in the kth initial clustering region; the ratio between the local clustering density and the actual clustering density of the kth initial clustering region is calculated, and the local discrete degree corresponding to the ath discrete feature data point is obtained according to the ratio. Right now where ρ ′ a is the local clustering density corresponding to the a-th discrete feature data point; ρ k is the actual clustering density of the kth initial clustering area.

[0037] The local discreteness calculation method described above can be used to obtain the local discreteness of all discrete feature data points relative to the kth initial clustering region, and further the local discreteness of all discrete feature data points relative to each initial clustering region.

[0038] (3) According to the local discreteness, the discrete feature data points are divided into corresponding initial clustering areas to achieve secondary clustering and obtain multiple clustering areas.

[0039] Specifically, since a discrete feature data point has a local discrete degree relative to each initial clustering area, that is, a discrete feature data point has a local discrete degree set relative to all initial clustering areas, the minimum local discrete degree in the local discrete degree set is used as the target local discrete degree of the corresponding discrete feature data point, and then the discrete feature data point is divided into the initial clustering area corresponding to the target local discrete degree; similarly, all discrete feature data points are divided to realize secondary clustering to obtain multiple clustering areas again.

[0040] Step S003, performing similarity matching between the feature data in each cluster area and the historical service data corresponding to each elderly care service type, and recommending the elderly care service type corresponding to each elderly person according to the matching result.

[0041] Specifically, the characteristic data in the clustering area after the secondary clustering is matched with the service data corresponding to the elderly care service type, so as to provide targeted elderly care service support for the elderly corresponding to each clustering area. The similarity matching can be performed by using a twin network to perform similarity matching between the characteristic data corresponding to the clustering area after the secondary clustering and the service history data of the elderly care service end, wherein the input of the twin network is: the characteristic data of the clustering area and the service history data of the elderly care service end, the loss function is: the cosine distance loss function, and the output of the twin network is: the characteristic data contained in the service history data and its matching clustering area. Targeted elderly care services are provided to the elderly or a relevant elderly care service system is specified based on the service history data corresponding to the characteristic data of the clustering area.

[0042] In summary, the embodiment of the present invention provides a collaborative elderly care service method based on artificial intelligence and big data analysis, which collects feature data of multiple elderly people to form a feature data set, wherein the feature data includes health data, daily life data and personal hobby data; the feature data set is density clustered to obtain multiple initial clustering areas and multiple discrete feature data points, and the local discreteness of each discrete feature data point relative to each initial clustering area is obtained by combining the distance between the discrete feature data point and each feature data point in the initial clustering area; the discrete feature data points are re-divided into the initial clustering area according to the local discreteness to obtain multiple clustering areas after secondary clustering; the feature data in each clustering area is matched with the historical service data corresponding to each elderly care service type for similarity, and the elderly care service type corresponding to each elderly person is recommended according to the matching results. Secondary clustering based on the local discreteness corresponding to the discrete feature data can improve the clustering accuracy of clustering, avoid the discrete feature data from being identified as abnormal data due to parameter settings, and thus solve the problem of being unable to accurately recommend elderly care services for the elderly corresponding to isolated data.

[0043] Preferably, in order to make the initial density clustering more accurate, the Minpts threshold in the density clustering process is adjusted according to the clustering results after the secondary clustering. The adjustment method is: counting the first number of discrete feature data points obtained by re-dividing each clustering area after the secondary clustering; calculating the sum of the first number and the Minpts threshold of each clustering area respectively, taking the average of all the sums as the new Minpts threshold, and taking the new Minpts threshold as the parameter for subsequent data clustering.

[0044] Based on the same inventive concept as the above method, an embodiment of the present invention also provides a collaborative elderly care service system based on artificial intelligence and big data analysis, including a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that when the processor executes the computer program, it implements the steps of any one of the above-mentioned collaborative elderly care service methods based on artificial intelligence and big data analysis.

[0045] It should be noted that the sequence of the embodiments of the present invention described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0046] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0047] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A collaborative elderly care service method based on artificial intelligence and big data analysis, characterized in that: The method includes: Collecting characteristic data of a plurality of elderly people to form a characteristic data set, wherein the characteristic data includes health data, daily life data and personal hobby data; Density clustering is performed on the feature data set to obtain a plurality of initial clustering regions and a plurality of discrete feature data points, and the local discreteness of each discrete feature data point relative to each initial clustering region is obtained in combination with the distance between the discrete feature data point and each feature data point in the initial clustering region; the discrete feature data points are re-divided into the initial clustering regions according to the local discreteness to obtain a plurality of clustering regions after secondary clustering; Performing similarity matching between the feature data in each clustering area and the historical service data corresponding to each elderly care service type, and recommending the elderly care service type corresponding to each elderly person according to the matching results; The method for determining the local discrete degree of each discrete feature data point relative to each initial clustering region comprises: Calculating the actual clustering density of the current initial clustering area according to the positional relationship between all feature data points in the initial clustering area; Calculating the local clustering density of the current discrete feature data point relative to the current initial clustering region, and obtaining the local discrete degree of the current discrete feature data point relative to the current initial clustering region from the ratio between the local clustering density and the actual clustering density; The method for obtaining the actual clustering density includes: Calculate the Euclidean distances between any feature data point and other feature data points in the current initial clustering area, obtain the Euclidean distance set corresponding to the feature data point, and take the maximum Euclidean distance in the Euclidean distance set as the target distance of the feature data point; Adding the target distances of all feature data points in the current initial clustering region to obtain a sum of first distances, and combining the sum of the first distances with the total number of feature data points included in the current initial clustering region to obtain the actual clustering density of the current initial clustering region; The method for calculating the local clustering density of the current discrete feature data point relative to the current initial clustering area includes: Calculating the distance between the current discrete feature data point and each feature data point in the current initial clustering region to obtain the sum of the second distances, and combining the sum of the second distances with the total number of the feature data points included in the current initial clustering region to obtain the local clustering density of the current discrete feature data point relative to the current initial clustering region; The local clustering density is negatively correlated with the sum of the second distances, and the local clustering density is positively correlated with the total number.

2. The method according to claim 1, characterized in that The density clustering is a DBSCAN algorithm.

3. A collaborative elderly care service system based on artificial intelligence and big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 2 are implemented.

Images