Intelligent matching method and system for health and wellness services based on deep learning and knowledge graph

Abstract

The method and system for intelligent matching of health care services based on deep learning and knowledge graph include the following steps: user demand acquisition, demand attribute acquisition, service and demand matching, service combination, service recommendation and service reselection. Through matching in three dimensions of nature, time and space, the above method, system and robot not only consider the matching of service types, but also consider the time and proximity of services, making the matched services real and available. In addition, the comprehensive priority of the service in the nature dimension is considered, so that high-efficiency and high-quality services can be matched first. Through service combination, the services of multiple institutions can be used to meet the needs of the same user. Through priority, the needs of users with high priority can be met first.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a health and wellness service intelligent matching method and system based on deep learning and a knowledge graph. BACKGROUND

[0002] In the process of implementing the present application, the inventors found that at least the following problems exist in the prior art: the service supply of health and wellness institutions forms an island and cannot be matched with the needs of users in a timely and effective manner.

[0003] Therefore, the prior art still needs to be improved and developed. SUMMARY

[0004] Therefore, it is necessary to provide a health and wellness service intelligent matching method and system based on deep learning and a knowledge graph to solve the problem that the service supply of health and wellness institutions can be matched with the needs of users in a coordinated manner, thereby improving the utilization rate and efficiency of health and wellness services and better meeting the health and wellness needs of users.

[0005] In a first aspect, an embodiment of the present application provides an artificial intelligence method, which comprises:

[0006] a user demand obtaining step of obtaining new demands in a user demand knowledge graph;

[0007] a demand attribute obtaining step of obtaining the nature attribute, time attribute and space attribute of the new demands of the user;

[0008] a service and demand matching step of calculating the matching degree of the new demands and the services, the matching degree of the new demands and the services = g (the comprehensive priority of the services, the nature matching degree, the time matching degree and the space matching degree); the time matching degree is the matching degree between the time period of the services and the time period of the demands, the space matching degree is the matching degree between the area faced by the services and the area of the expected services of the user, and the nature matching degree includes the matching degree between the type of the demands and the type of the services and the matching degree between the content of the demands and the content of the services;

[0009] a service combination step of adding the services with a matching degree greater than a preset matching degree to a candidate service set, combining any two services in the candidate service set to obtain the services after combination, and if the matching degree between the services after combination and the demands is greater than the existing maximum matching degree, adding the services after combination to the candidate service set, deleting the two services constituting the combination from the set, and then repeating the step until the matching degree between the services after combination and the demands is not greater than the existing maximum matching degree;

[0010] The service recommendation step is: finding a preset number of services with the largest matching degree and not being occupied from the candidate service set and sending the information of the services and the information of the institutions to which the services belong to the user, so that the user selects a service from the services;

[0011] The service reselection step is: if multiple user demands select the same service, assigning the service to the user demand with the largest comprehensive priority, updating the state of the service to an occupied state, sending a prompt that the service has been preempted to the user demand that does not obtain the service assignment, and returning to the service recommendation step to re-execute.

[0012] Preferably, the method further comprises:

[0013] The user demand prediction deep learning model training step is: obtaining the selection of a preset number of users for a preset user demand type as the demand of each user, obtaining the data of the user, obtaining an initial deep learning model, taking the data of the user as the input of the model, taking the demand of the user as the expected output of the model, training the model, and obtaining a user demand prediction deep learning model.

[0014] The user demand prediction deep learning model testing step is: obtaining the selection of a second preset number of users for a preset user demand type as the demand of each user, obtaining the data of the user, obtaining the user demand prediction deep learning model trained, taking the data of the user as the input of the model, taking the demand of the user as the expected output of the model, and testing the model; the second preset number of users do not belong to the first preset number of users.

[0015] The test judgment step is: if the prediction accuracy of the test is less than a preset accuracy, executing 7; if the prediction accuracy of the test is greater than or equal to the preset accuracy, taking the latest user demand prediction deep learning model as the user demand prediction deep learning model, and executing the model prediction step;

[0016] The retraining step is: obtaining the selection of a new third preset number of users for a preset user demand type as the demand of each user, obtaining the data of the user, obtaining the latest user demand prediction deep learning model, taking the data of the user as the input of the model, taking the demand of the user as the expected output of the model, training the model, and then returning to the test judgment step for execution; the new third preset number of users do not belong to the users that have been used in training and testing;

[0017] Model prediction step: obtaining data of a user whose demand needs to be predicted, obtaining a user demand prediction deep learning model, taking the data of the user as input of the model, and taking output calculated by the model as the predicted demand of the user;

[0018] User feedback step: sending the predicted demand to the user, obtaining feedback of the user on whether the prediction is correct, if the feedback is that the prediction is correct, taking the predicted demand as the demand of the user; if the feedback is that the prediction is incorrect, obtaining selection of a preset user demand type by the user as the demand of the user, obtaining data of the user, obtaining a latest user demand prediction deep learning model, taking the data of the user as input of the model, taking the demand of the user as expected output of the model, and incrementally training the model.

[0019] Preferably, the method further comprises:

[0020] Comprehensive priority calculation step of service: calculating an institution basic priority of each institution according to qualification of the institution, obtaining an evaluation of each service of the institution, calculating a priority of the service, and calculating a comprehensive priority of the service according to the institution basic priority and the priority of the service.

[0021] Institution service knowledge graph ontology construction step: constructing an ontology of an institution service knowledge graph according to various types of institutions, various types of services and relationships therebetween; in the ontology of the institution service knowledge graph, a service entity has a current attribute and a historical attribute; the current attribute and the historical attribute both include an intrinsic attribute, a time attribute and a space attribute; the intrinsic attribute includes a basic attribute and a state attribute; the basic attribute includes a type, a content, a comprehensive priority, a user type and a charging standard; the state attribute includes whether the service is idle or occupied; the time attribute includes a time period during which the service can be provided; and the space attribute includes an area to which the service is oriented; and the historical attribute is a record of a past current attribute.

[0022] Institution service knowledge graph formation step: accepting registration of an institution and a service thereof, and automatically adding the institution and the service to the institution service knowledge graph according to the ontology.

[0023] Preferably, the method further comprises:

[0024] Tree structure of service entities: each service entity can include multiple sub-service entities; the tree of service entities is a tree; if the relationship between an agency and a service is an ownership relationship, the relationship between the agency and all sub-services of the service is an ownership relationship; if the relationship between an agency and a service is a partial ownership relationship, the relationship between the agency and some sub-services of the service is an ownership relationship or a partial ownership relationship; if the relationship between service A and service B is a certain relationship, the relationship between all sub-services of service A and all sub-services of service B is the certain relationship; if the relationship between service A and service B is a certain partial relationship, the relationship between some sub-services of service A and some sub-services of service B is the certain relationship or the certain partial relationship;

[0025] Tree structure of agency entities: each agency entity can include multiple sub-agency entities; the tree of agency entities is a tree or a tree; if the relationship between an agency and a service is an ownership relationship, the relationship between some or all sub-agencies of the agency and the service is an ownership relationship; if the relationship between an agency and a service is a partial ownership relationship, the relationship between some or all sub-agencies of the agency and the service is a partial ownership relationship; if the relationship between agency A and agency B is a certain relationship, the relationship between some or all sub-agencies of agency A and some or all sub-agencies of agency B is the certain relationship.

[0026] In a second aspect, embodiments of the present application provide an artificial intelligence system, the system comprising:

[0027] A user demand acquisition module: acquiring new demands in a user demand knowledge graph;

[0028] A demand attribute acquisition module: acquiring the nature attribute, time attribute, and space attribute of the new demand of the user;

[0029] A service and demand matching module: calculating the matching degree of the new demand and the service, the matching degree of the new demand and the service = g (comprehensive priority of the service, nature matching degree, time matching degree, space matching degree); the time matching degree is the matching degree between the time period of the service and the time period of the demand, the space matching degree is the matching degree between the area faced by the service and the area of the user's expected service, and the nature matching degree includes the matching degree between the type of the demand and the type of the service and the matching degree between the content of the demand and the content of the service;

[0030] a service combination module: adding services with a matching degree greater than a preset matching degree to a candidate service set, combining any two services in the candidate service set to obtain a combined service, and if the matching degree between the combined service and the demand is greater than an existing maximum matching degree, adding the combined service to the candidate service set and deleting the two services that constitute the combination from the set, and then repeating the module until the matching degree between the combined service and the demand is not greater than the existing maximum matching degree;

[0031] a service recommendation module: finding a preset number of services with the maximum matching degree and not occupied from the candidate service set and recommending the services to the user, and extracting information of the services and information of institutions to which the services belong from the institutional service knowledge graph and sending the information to the user, so that the user selects a service from the information;

[0032] a service reselection module: if multiple user demands select the same service at the same time, assigning the service to a user demand with a high comprehensive priority, updating a state of the service to an occupied state, sending a prompt that the service has been preempted to a user demand that does not obtain the service assignment, and returning to the service recommendation module to re-execute.

[0033] Preferably, the system further comprises:

[0034] a user demand prediction deep learning model training module: obtaining a selection of a preset user demand type by each of a first preset number of users as a demand of the user, obtaining data of the user, obtaining an initial deep learning model, taking the data of the user as an input of the model, taking the demand of the user as an expected output of the model, training the model, and obtaining a user demand prediction deep learning model;

[0035] a user demand prediction deep learning model testing module: obtaining a selection of a preset user demand type by each of a second preset number of users as a demand of the user, obtaining data of the user, obtaining the user demand prediction deep learning model trained, taking the data of the user as an input of the model, taking the demand of the user as an expected output of the model, and testing the model; the second preset number of users do not belong to the first preset number of users;

[0036] a test judgment module: if a prediction accuracy of the test is less than a preset accuracy, executing 7; if the prediction accuracy of the test is greater than or equal to the preset accuracy, taking the latest user demand prediction deep learning model as the user demand prediction deep learning model, and executing the model prediction module;

[0037] A retraining module: obtaining the selection of each user of a new third preset number of preset user demand types as the demand of the user, obtaining the data of the user, obtaining the latest user demand prediction deep learning model, taking the data of the user as the input of the model, taking the demand of the user as the expected output of the model, training the model, and then returning to the test judgment module for execution; the new third preset number of each user does not belong to the user who has been used in training and testing;

[0038] A model prediction module: obtaining the data of a user whose demand needs to be predicted, obtaining a user demand prediction deep learning model, taking the data of the user as the input of the model, and taking the output calculated by the model as the predicted demand of the user;

[0039] A user feedback module for the model: sending the predicted demand to the user, obtaining the feedback of the user on whether the prediction is correct, if the feedback is that the prediction is correct, taking the predicted demand as the demand of the user; if the feedback is that the prediction is incorrect, obtaining the selection of the user of a preset user demand type as the demand of the user, obtaining the data of the user, obtaining the latest user demand prediction deep learning model, taking the data of the user as the input of the model, taking the demand of the user as the expected output of the model, and incrementally training the model.

[0040] Preferably, the system further comprises:

[0041] A comprehensive priority calculation module of a service: calculating an institutional basic priority of each institution according to the qualification of the institution, obtaining the evaluation of each service of the institution, calculating the priority of the service, and calculating the comprehensive priority of the service according to the institutional basic priority and the priority of the service;

[0042] An institutional service knowledge graph ontology construction module: constructing the ontology of the institutional service knowledge graph according to various types of institutions, various types of services, and the relationship therebetween; in the ontology of the institutional service knowledge graph, a service entity has a current attribute and a historical attribute; the current attribute and the historical attribute both include an intrinsic attribute, a time attribute, and a space attribute; the intrinsic attribute includes a basic attribute and a state attribute; the basic attribute includes a type, a content, a comprehensive priority, a user type, and a charging standard; the state attribute includes whether the service is idle or occupied; the time attribute includes a time period during which the service can be provided; the space attribute includes an area to which the service is oriented; and the historical attribute is a record of the past current attribute;

[0043] An institutional service knowledge graph formation module: accepting the registration of an institution and its service, and automatically adding the institution and its service to the institutional service knowledge graph according to the ontology.

[0044] Preferably, the system further comprises:

[0045] Tree structure of service entities: each service entity can include multiple child service entities; the tree of service entities is a tree; if the relationship between an agency and a service is an owning relationship, the relationship between the agency and all child services of the service is an owning relationship; if the relationship between an agency and a service is a partial owning relationship, the relationship between the agency and some child services of the service is an owning relationship or the relationship between the agency and some child services of the service is a partial owning relationship; if the relationship between an A service and a B service is a certain relationship, the relationship between all child services of the A service and all child services of the B service is the certain relationship; if the relationship between an A service and a B service is a certain partial relationship, the relationship between some child services of the A service and some child services of the B service is the certain relationship or the relationship between some child services of the A service and some child services of the B service is the certain partial relationship.

[0046] Tree structure of agency entities: each agency entity can include multiple child agency entities; the tree of agency entities is a tree or a tree; if the relationship between an agency and a service is an owning relationship, the relationship between some or all child agencies of the agency and the service is an owning relationship; if the relationship between an agency and a service is a partial owning relationship, the relationship between some or all child agencies of the agency and the service is a partial owning relationship; if the relationship between an A agency and a B agency is a certain relationship, the relationship between some or all child agencies of the A agency and some or all child agencies of the B agency is the certain relationship.

[0047] In a third aspect, an embodiment of the present application provides an artificial intelligence device, and the system comprises the device of any one of the modules in the second aspect.

[0048] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the steps of the method in any one of the first aspect.

[0049] In a fifth aspect, an embodiment of the present application provides a robot system, which comprises a memory, a processor, and an artificial intelligence robot program stored in the memory and executable on the processor, and the processor executes the program to implement the steps of the method in any one of the first aspect.

[0050] The method and system for intelligent matching of health and wellness services based on deep learning and knowledge graph provided by the embodiment include: a user demand obtaining step; a demand attribute obtaining step; a service and demand matching step; a service combination step; a service recommendation step; and a service reselection step. The method, system and robot match in three dimensions of nature, time and space, consider not only the matching of the type of service, but also the time of service and nearby service, so that the matched service is real and available, and the comprehensive priority of the service is considered in the nature dimension, so that high-efficiency and high-quality service can be matched preferentially; through service combination, the services of multiple institutions can be used to serve the demand of a same user; and through priority, the demand of a user with high priority can be satisfied preferentially. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 A module diagram of an artificial intelligence system is provided for the embodiment of the application.

[0052] Figure 2 A module diagram of an artificial intelligence system is provided for the embodiment of the application.

[0053] Figure 3 A module diagram of an artificial intelligence system is provided for the embodiment of the application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the application will be described in detail below with reference to the embodiments of the application.

[0055] I. Basic embodiment of the application

[0056] In a first aspect, the embodiment of the application provides an artificial intelligence method, which includes: a user demand obtaining step; a demand attribute obtaining step; a service and demand matching step; a service combination step; a service recommendation step; and a service reselection step. Technical effects: through matching in three dimensions of nature, time and space, not only the matching of the type of service is considered, but also the time of service and nearby service are considered, so that the matched service is real and available, and the comprehensive priority of the service is considered in the nature dimension, so that high-efficiency and high-quality service can be matched preferentially; through service combination, the services of multiple institutions can be used to serve the demand of a same user; and through priority, the demand of a user with high priority can be satisfied preferentially.

[0057] In a preferred embodiment, the method further includes: a user demand prediction deep learning model training step; a user demand prediction deep learning model testing step; a testing judgment step; a retraining step; a model prediction step; and a user feedback step. Technical effects: through continuous improvement of the demand prediction model according to the feedback of the user in use, the model can continuously evolve itself.

[0058] In a preferred embodiment, the method further comprises: a service comprehensive priority calculation step; an institutional service knowledge graph ontology construction step; an institutional service knowledge graph formation step. Technical effect: services are described in the knowledge graph from the dimensions of nature, time, and space, laying a technical foundation for accurate matching of services.

[0059] In a preferred embodiment, the method further comprises: a tree structure of service entities; a tree structure of institutional entities. Technical effect: according to the characteristics of institutional subdivision and service subdivision, the services are respectively organized into a tree and a tree, laying a technical foundation for the cooperation of multiple service combinations to provide services for the same demand.

[0060] In a second aspect, the embodiments of the present application provide an artificial intelligence system, as shown in Figure 1 The system comprises: a user demand acquisition module; a demand attribute acquisition module; a service and demand matching module; a service combination module; a service recommendation module; a service reselection module.

[0061] In a preferred embodiment, as shown in Figure 2 The system further comprises: a user demand prediction deep learning model training module; a user demand prediction deep learning model testing module; a testing judgment module; a retraining module; a model prediction module; a user feedback module for the model.

[0062] In a preferred embodiment, as shown in Figure 3 The system further comprises: a service comprehensive priority calculation module; an institutional service knowledge graph ontology construction module; an institutional service knowledge graph formation module.

[0063] In a preferred embodiment, the system further comprises: a tree structure of service entities; a tree structure of institutional entities.

[0064] In a third aspect, the embodiments of the present application provide an artificial intelligence device, which comprises the modules of the system of any one of the embodiments of the second aspect.

[0065] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, which stores a computer program, and the program is executed by a processor to realize the steps of the method of any one of the embodiments of the first aspect.

[0066] In a fifth aspect, the embodiments of the present application provide a robot system, which comprises a memory, a processor, and an artificial intelligence robot program stored in the memory and executable on the processor, and the processor executes the program to realize the steps of the method of any one of the embodiments of the first aspect.

[0067] II. Preferred embodiments of the present application

[0068] (I) Key issues

[0069] How to achieve precise and efficient matching of user differentiated needs and diversified supply of medical and health care service enterprises to improve users' sense of acquisition, happiness and security, and promote the benign competition and mutual benefit of medical and health care service enterprises?

[0070] (II) Key technologies

[0071] Through big data and artificial intelligence technology, precise and efficient matching of user differentiated needs and diversified supply of medical and health care service enterprises is achieved.

[0072] (III) Technical focus

[0073] The precise and efficient matching mode of user differentiated needs and diversified supply of medical and health care service enterprises can improve users' sense of acquisition, happiness and security, and promote the benign competition and mutual benefit of medical and health care service enterprises.

[0074] (IV) Technical summary scheme

[0075] Collect user big data such as basic information (personal resume, family information, etc.), pension information (pension history, pension status, etc.), health information (medical record, physical sign, etc.), daily information (behavior, consumption, etc.), basic information of medical and health care service institutions (enterprise type, enterprise qualification, etc.), service information (service type, service content, etc.), service record (service history, service evaluation, etc.), build user demand prediction deep learning model according to user big data, predict user differentiated needs in real time through the model, build medical and health care service matching expert system model according to medical and health care expert knowledge, and recommend diversified services of medical and health care service enterprises to users in time according to the expert system model. Experts check the results of prediction and recommendation, collect user feedback, and optimize and improve the model and system according to the user feedback and the results of the check.

[0076] (V) Technical detailed scheme

[0077] User demand prediction steps

[0078] 1. Obtain user big data. User big data includes data of each user.

[0079] 2. User big data includes user's basic information (personal resume, family information, etc.), pension information (pension history, pension status, etc.), health information (medical record, physical sign, etc.), daily information (behavior, consumption, etc.).

[0080] 3 Obtain the preset user demand type. The user demand type includes medical demand, rehabilitation demand, and old-age demand. Each demand type includes more specific subtypes, and each demand subtype includes more specific grandchild types, and so on, until the most basic indivisible demand type. For example, rehabilitation demand can be further divided into demand for rehabilitation in an institution, demand for rehabilitation in a community, and demand for rehabilitation at home, which can be further subdivided.

[0081] 4 Obtain the selection of each user of a preset user demand type as the demand of the user, obtain the data of the user, obtain an initial deep learning model, input the data of the user into the model, input the demand of the user as the expected output of the model, train the model, and obtain a user demand prediction deep learning model.

[0082] 5 Obtain the selection of each user of a preset user demand type as the demand of the user, obtain the data of the user, obtain the trained user demand prediction deep learning model, input the data of the user into the model, input the demand of the user as the expected output of the model, and test the model. Each user in the second preset number does not belong to each user in the first preset number.

[0083] 6 If the prediction accuracy of the test is less than the preset accuracy, perform 7; if the prediction accuracy of the test is greater than or equal to the preset accuracy, use the latest user demand prediction deep learning model as the user demand prediction deep learning model, and perform 8.

[0084] 7 Obtain the selection of each user of a preset user demand type as the demand of the user, obtain the data of the user, obtain the latest user demand prediction deep learning model, input the data of the user into the model, input the demand of the user as the expected output of the model, train the model, and then return to 6 for execution. Each user in the new third preset number does not belong to the users that have been used in training and testing.

[0085] 8 Obtain the data of a user whose demand needs to be predicted, obtain the user demand prediction deep learning model, input the data of the user into the model, and obtain the output calculated by the model as the predicted demand of the user.

[0086] 9 send the predicted demand to the user, obtain the user's feedback on whether the prediction is correct, if the feedback is that the prediction is correct, take the predicted demand as the user's demand; if the feedback is that the prediction is incorrect, obtain the user's selection of the preset user demand type as the user's demand, and obtain the user's data, obtain the latest user demand prediction deep learning model, take the user's data as the input of the model, take the user's demand as the expected output of the model, and incrementally train the model.

[0087] Institution service description step

[0088] 1 Obtain institution big data. The institution big data includes data of each institution. The institution includes a hospital, a community, a nursing home, a health and wellness enterprise, etc. The institution big data includes basic information of medical and health service institutions (enterprise type, enterprise qualification, etc.), service information (service type, service content, etc.), service records (service history, service evaluation, etc.).

[0089] 2 Obtain preset institution service types. The institution service types include medical services, rehabilitation services, and nursing services, each service type includes more specific subtypes, and each service subtype includes more specific service grandchild types, and so on, until the most basic indivisible service type. For example, rehabilitation services can be further divided into rehabilitation services in institutions, rehabilitation services in communities, and rehabilitation services at home, and can be further subdivided. The preset institution service types correspond one-to-one to the preset user demand types.

[0090] 3 Calculate the institution basic priority of each institution according to its qualification, obtain the evaluation of each service of the institution (including the measurement of efficiency and the evaluation of quality), calculate the priority of the service, and calculate the comprehensive priority of the service according to the institution basic priority and the priority of the service.

[0091] Institution service knowledge graph construction step

[0092] 1 Construct an institution service knowledge graph according to each service of each institution and its comprehensive priority. The knowledge graph has institution entities and service entities. The relationships between institutions include cooperative relationships, competitive relationships, and competitive cooperative relationships, etc., the relationships between institutions and services include ownership relationships and partial ownership relationships, etc., and the relationships between services include precedence relationships and cooperative relationships, etc. The attributes of services include service types, service comprehensive priorities, service contents, etc. Services of the same type can have different differentiated services.

[0093] 2 Service entity has current attribute, history attribute. Both current attribute and history attribute include nature attribute, time attribute and space attribute. Nature attribute includes basic attribute and state attribute. Basic attribute includes type, content, user type, charging standard. State attribute includes whether service is idle, occupied, if occupied, which user is served. Time attribute includes time period when service can be provided. Space attribute includes area where service is provided. History attribute is record of past current attribute.

[0094] 3 Each institution entity can include multiple sub-institution entities. Each service entity can include multiple sub-service entities. Tree of service entities is tree of trees. If relationship between institution and service is ownership, then relationship between institution and all sub-services of the service is ownership. If relationship between institution and service is partial ownership, then relationship between institution and some sub-services of the service is ownership or partial ownership. If relationship between service A and service B is a certain relationship, then relationship between all sub-services of A and all sub-services of B is the certain relationship. If relationship between service A and service B is a certain partial relationship, then relationship between some sub-services of A and some sub-services of B is the certain relationship or some sub-services of A and some sub-services of B are the certain partial relationship. Tree of institution entities is tree of trees or tree of trees of trees. If relationship between institution and service is ownership, then relationship between some or all sub-institutions of the institution and the service is ownership. If relationship between institution and service is partial ownership, then relationship between some or all sub-institutions of the institution and the service is partial ownership. If relationship between institution A and institution B is a certain relationship, then relationship between some or all sub-institutions of A and some or all sub-institutions of B is the certain relationship.

[0095] User demand description step

[0096] 1 Basic information (personal resume, family information, etc.), pension information (pension history, pension status, etc.), health information (medical record, physical sign, etc.), daily information (behavior, consumption, etc.).

[0097] 2 Calculate basic priority of demand of each user according to basic information of the user (for example, if user is a laborer, basic priority is high), evaluate urgency of each demand of the user according to health status of the user, calculate priority of the demand, calculate comprehensive priority of the demand according to basic priority of institution and priority of the demand.

[0098] User demand knowledge graph construction step

[0099] 1. Construct a user demand knowledge graph according to each demand of each user and its comprehensive priority. The knowledge graph has user entities and demand entities. The relationship between users includes family relationship, neighborhood relationship, friendship, etc. The relationship between users and demands includes having relationship and partial having relationship. The relationship between demands includes precedence relationship and association relationship. The attributes of demands include demand type, comprehensive priority, content, etc. Demands of the same type can have different personalized contents. Each user entity can include multiple sub-user entities. For example, a community user entity can include multiple family user entities, and a family user entity can include multiple individual user entities.

[0100] 2. Demand entities have current attributes and historical attributes. Both current attributes and historical attributes include nature attributes, time attributes, and space attributes. Nature attributes include basic attributes and state attributes. Basic attributes include type, content, targeted organization type, and payment standard. State attributes include whether the demand is unmet or met, and if met, which organization provides the service. Time attributes include the expected time period of the demand. Space attributes include the area where the demand expects to receive service, such as the user's home. Historical attributes are records of past current attributes.

[0101] 3. Each demand entity can include multiple sub-demand entities. The tree structure of demand entities is a with-tree. If the relationship between a user and a demand is having relationship, the user has a having relationship with all sub-demands of the demand. If the relationship between a user and a demand is partial having relationship, the user has a having relationship with some sub-demands of the demand or a partial having relationship with some sub-demands. If the relationship between demand A and demand B is a certain relationship, all sub-demands of A and all sub-demands of B have that relationship. If the relationship between demand A and demand B is a certain partial relationship, some sub-demands of A and some sub-demands of B have that relationship or some sub-demands of A and some sub-demands of B have that partial relationship. The tree structure of user entities is a or-tree. If the relationship between a user and a demand is having relationship, some or all sub-users of the user have a having relationship with the demand. If the relationship between a user and a demand is partial having relationship, some or all sub-users of the user have a partial having relationship with the demand. If the relationship between user A and user B is a certain relationship, some or all sub-users of A and some or all sub-users of B have that relationship.

[0102] Knowledge graph formation steps:

[0103] 1. Construct the ontology of the user demand knowledge graph according to various types of users, various types of demands, and the relationships between them.

[0104] 2 Accept the registration of users and their needs, and automatically add the user needs knowledge graph according to the ontology.

[0105] 3 Construct the ontology of the institution service knowledge graph according to various institutions, various services and the relationship between them.

[0106] 4 Accept the registration of institutions and their services, and automatically add the institution service knowledge graph according to the ontology.

[0107] Demand and service matching step

[0108] 1 Obtain the new demand in the user needs knowledge graph; obtain the demand type, time attribute and space attribute of the new demand of the user, and find the most matched service in the institution service knowledge graph. The matching degree of the new demand and the service = g (the comprehensive priority of the service, the nature matching degree, the time matching degree and the space matching degree), g is a preset function, and the higher the comprehensive priority of the service, the higher the nature matching degree, the higher the time matching degree, the higher the space matching degree, and the higher the matching degree of the new demand and the service. (1) The time matching degree is the matching degree between the time period of the service and the time period of the demand, the space matching degree is the matching degree between the area faced by the service and the area of the user's expected service, and the nature matching degree includes the matching degree of the type of the demand and the type of the service, and the matching degree of the content of the demand and the content of the service. In the matching, first, the services irrelevant in type are excluded through the matching degree of the type of the demand and the type of the service, and then the matching degree of the new demand and the service is calculated. (2) In the matching, the services with a matching degree greater than a preset matching degree are added to a candidate service set, any two services in the candidate service set are combined to obtain a service after combination, if the matching degree of the service after combination and the demand is greater than the existing maximum matching degree, the service after combination is added to the candidate service set, and the two services constituting the combination are deleted from the set, and then the step is repeated until the matching degree of the service after combination and the demand is not greater than the existing maximum matching degree. (3) From the candidate service set, find a preset number of services with the maximum matching degree and not occupied, recommend to the user, and extract the information of the services and the information of the institutions to which the services belong from the institution service knowledge graph and send to the user, so that the user selects a service from them. (4) If multiple user needs select the same service at the same time, the service is allocated to the user needs with a high comprehensive priority, the state of the service is updated to an occupied state, a prompt that the service has been preempted is sent to the user needs that do not obtain the service allocation, and the step (3) is executed again.

[0109] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the patent scope of the present application. It should be pointed out that, then, for ordinary skilled persons in the art, without departing from the concept of the present application, several modifications and improvements can be made, which all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A health and wellness service intelligent matching method based on deep learning and knowledge graph, characterized in that, The method comprises: a user demand obtaining step of obtaining a new demand in a user demand knowledge graph; a demand attribute obtaining step of obtaining a nature attribute, a time attribute and a space attribute of the new demand of the user; a service and demand matching step of calculating a matching degree of the new demand and a service, the matching degree of the new demand and the service being g (a comprehensive priority of the service, a nature matching degree, a time matching degree and a space matching degree); the time matching degree being a matching degree between a time period of the service and a time period of the demand, the space matching degree being a matching degree of an area faced by the service and an area of the service expected by the user, and the nature matching degree including a matching degree of a type of the demand and a type of the service and a matching degree of content of the demand and content of the service; a service combination step of adding a service with a matching degree greater than a preset matching degree to a candidate service set, combining any two services in the candidate service set to obtain a service after combination, and if a matching degree of the service after combination and the demand is greater than an existing maximum matching degree, adding the service after combination to the candidate service set and deleting the two services constituting the combination from the set, and then repeating the step until the matching degree of the service after combination and the demand is not greater than the existing maximum matching degree; a service recommendation step of finding a preset number of services with the greatest matching degree and not occupied from the candidate service set and recommending the services to the user, and extracting information of the services and information of an institution to which the services belong from an institutional service knowledge graph and sending the information to the user, so that the user selects a service from the information; a service reselection step of, if multiple user demands select the same service, allocating the service to a user demand with a greater comprehensive priority, updating a state of the service to an occupied state, sending a prompt that the service has been preempted to a user demand that does not obtain the service allocation, and returning to the service recommendation step to be executed again.

2. The method of claim 1, wherein, The method further comprises: a comprehensive priority calculation step of calculating an institutional basic priority of each institution according to a qualification of the institution, obtaining an evaluation of each service of the institution, calculating a priority of the service, and calculating a comprehensive priority of the service according to the institutional basic priority and the priority of the service; an institutional service knowledge graph ontology construction step of constructing an ontology of the institutional service knowledge graph according to various types of institutions, various types of services and relationships therebetween; in the ontology of the institutional service knowledge graph, a service entity has a current attribute and a historical attribute; the current attribute and the historical attribute both include a nature attribute, a time attribute and a space attribute; the nature attribute includes a basic attribute and a state attribute; the basic attribute includes a type, a content, a comprehensive priority, a user type faced by the service and a charging standard; the state attribute includes whether the service is idle or occupied; the time attribute includes a time period in which the service can be provided; and the space attribute includes an area faced by the service; and the historical attribute is a record of a past current attribute; an institutional service knowledge graph formation step of accepting registration of institutions and services thereof and automatically adding the institutions and the services thereof to the institutional service knowledge graph according to the ontology.

3. The method of claim 2, wherein, The method further comprises: Tree structure of service entities: each service entity can include multiple sub-service entities; the tree of service entities is a tree or a tree; if the relationship between an agency and a service is an ownership relationship, the agency has an ownership relationship with all sub-services of the service; if the relationship between an agency and a service is a partial ownership relationship, the agency has an ownership relationship with some sub-services of the service or a partial ownership relationship with some sub-services of the service; if the relationship between A service and B service is a certain relationship, all sub-services of A service and all sub-services of B service are the relationship; if the relationship between A service and B service is a certain partial relationship, some sub-services of A service and some sub-services of B service are the relationship or some sub-services of A service and some sub-services of B service are the partial relationship; Tree structure of agency entities: each agency entity can include multiple sub-agency entities; the tree of agency entities is a tree or a tree; if the relationship between an agency and a service is an ownership relationship, the agency has an ownership relationship with some or all sub-agencies of the service; if the relationship between an agency and a service is a partial ownership relationship, the agency has a partial ownership relationship with some or all sub-agencies of the service; if the relationship between A agency and B agency is a certain relationship, some or all sub-agencies of A agency and some or all sub-agencies of B agency are the relationship.

4. A health and wellness service intelligent matching system based on deep learning and a knowledge graph, characterized in that, The system comprises: a user demand acquisition module: acquiring new demands in a user demand knowledge graph; a demand attribute acquisition module: acquiring the nature attribute, time attribute and space attribute of the new demands of the user; a service and demand matching module: calculating the matching degree of the new demands and the services, the matching degree of the new demands and the services = g (comprehensive priority of the services, nature matching degree, time matching degree, space matching degree); the time matching degree is the matching degree between the time period of the service and the time period of the demand, the space matching degree is the matching degree between the area faced by the service and the area of the expected service of the user, and the nature matching degree includes the matching degree between the type of the demand and the type of the service and the matching degree between the content of the demand and the content of the service; a service combination module: adding services with a matching degree greater than a preset matching degree to a candidate service set, combining any two services in the candidate service set to obtain a combined service, if the matching degree between the combined service and the demand is greater than the existing maximum matching degree, adding the combined service to the candidate service set and deleting the two services constituting the combination from the set, then repeating the module until the matching degree between the combined service and the demand is not greater than the existing maximum matching degree; a service recommendation module: finding a preset number of services with the maximum matching degree and not occupied from the candidate service set and recommending the services to the user, and extracting the information of the services and the information of the agencies to which the services belong from the agency service knowledge graph and sending the information to the user, so that the user selects a service from the information; The service reselection module: if multiple user demands select the same service, the service is allocated to the user demand with higher comprehensive priority, the state of the service is updated to occupied, a prompt that the service has been preempted is sent to the user demand that does not obtain the service allocation, and the service recommendation module is returned to re-execute.

5. The system of claim 4, wherein, The system further comprises: The service comprehensive priority calculation module: the basic priority of each institution is calculated according to the qualification of each institution, the evaluation of each service of the institution is obtained, the priority of the service is calculated, and the comprehensive priority of the service is calculated according to the basic priority of the institution and the priority of the service; The institution service knowledge graph ontology construction module: the ontology of the institution service knowledge graph is constructed according to various institutions, various services and the relationship therebetween; in the ontology of the institution service knowledge graph, the service entity has current attributes and historical attributes; the current attributes and the historical attributes both include essential attributes, time attributes and space attributes; the essential attributes include basic attributes and state attributes; the basic attributes include type, content, comprehensive priority, user type and charging standard; the state attributes include whether the service is idle or occupied; the time attributes include the time period during which the service can be provided; and the space attributes include the area to which the service is oriented; the historical attributes are the record of the past current attributes; The institution service knowledge graph formation module: the registration of institutions and their services is accepted, and the institution service knowledge graph is automatically added according to the ontology.

6. The system of claim 5, wherein, The system further comprises: The tree structure of the service entity: each service entity can include multiple sub-service entities; the tree formed by the service entities is a tree; if the relationship between the institution and the service is an owning relationship, the relationship between the institution and all sub-services of the service is also an owning relationship; if the relationship between the institution and the service is a partial owning relationship, the relationship between the institution and some sub-services of the service is an owning relationship or a partial owning relationship; if the relationship between service A and service B is a certain relationship, the relationship between all sub-services of service A and all sub-services of service B is also the certain relationship; if the relationship between service A and service B is a certain partial relationship, the relationship between some sub-services of service A and some sub-services of service B is the certain relationship or the certain partial relationship; The tree structure of the institution entity: each institution entity can include multiple sub-institution entities; the tree formed by the institution entities is a tree or a tree; if the relationship between the institution and the service is an owning relationship, the relationship between the institution and some or all sub-institutions of the service is also an owning relationship; if the relationship between the institution and the service is a partial owning relationship, the relationship between the institution and some or all sub-institutions of the service is also a partial owning relationship; if the relationship between institution A and institution B is a certain relationship, the relationship between some or all sub-institutions of institution A and some or all sub-institutions of institution B is also the certain relationship.

7. A robotic system comprising a memory, a processor, and an artificial intelligence robot program stored on the memory and executable on the processor, wherein, The processor executes the program to realize the steps of the method of any one of claims 1-3.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that The program is executed by the processor to realize the steps of the method of any one of claims 1-3.

Images