Insurance product pushing method and device, storage medium and electronic equipment

Abstract

The invention provides an insurance product pushing method and device, a storage medium and electronic device.The method comprises the steps that user attribute information of a target user account is obtained, dynamic behavior information of the target user account is obtained, and environment state information of the position where user equipment is located is obtained, the target user account logs in the user equipment; adopting the user attribute information, the dynamic behavior information and the environment state information to construct a user situation portrait of the target user account; and pushing a target insurance product to the target user account based on the user situation portrait. The technical problem that the insurance product pushing efficiency is low in the related technology is solved, the personalized matching degree of the insurance product and the user is improved, and the insurance product pushing efficiency is also improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to a method and device for pushing insurance products, a storage medium, and an electronic device. Background Art

[0002] In the related art, with the development of Internet and big data technologies, the insurance industry is facing unprecedented opportunities and challenges. The traditional recommendation methods for insurance products mainly rely on the experience of salespersons and limited information provided by users, which is not only inefficient but also prone to deviation, and it is difficult to meet the personalized needs of users. In addition, due to the large variety of insurance types in the market, users often need to spend a lot of time to select suitable insurance products for themselves, and even make unreasonable decisions in the absence of professional knowledge.

[0003] In the related art, the recommendation methods based on historical data have limitations and are difficult to capture the immediate needs and dynamically changing situations of users, resulting in inaccurate recommended insurance data, which is likely to mislead users or even cause information harassment, affecting the user experience.

[0004] In view of the above problems existing in the related art, no efficient and accurate solution has been found yet. Summary of the Invention

[0005] The present invention provides a method and device for pushing insurance products, a storage medium, and an electronic device to solve the technical problems in the related art.

[0006] According to an embodiment of the present invention, a method for pushing insurance products is provided, including: obtaining user attribute information of a target user account, obtaining dynamic behavior information of the target user account, and obtaining environmental status information of the location where the user device is located, wherein the target user account is logged in on the user device; constructing a user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information; and pushing a target insurance product to the target user account based on the user context portrait.

[0007] Optionally, obtaining the dynamic behavior information of the target user account includes: obtaining query data input by the target user account on a search interface; obtaining the online browsing record of the target user account; obtaining the online public opinion data of the target user account; obtaining the travel information of the user corresponding to the target user account; and / or, obtaining the environmental status information of the location where the user device is located includes: obtaining the geological features of the location where the user device is located; obtaining the weather type at the location where the user device is located in a future time; obtaining the traffic status information at the location where the user device is located in a future time; obtaining the disease prevalence status at the location where the user device is located; obtaining the penetration status of insurance products at the location where the user device is located.

[0008] Optionally, constructing the user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information includes: performing data preprocessing on the user attribute information and the dynamic behavior information to obtain multi-source fusion data; inputting the multi-source fusion data into a pre-trained context awareness model to output the immediate demand information of the target user account; predicting the future behavior trend of the target user account based on the immediate demand information and the environmental status information, where the user context portrait includes the immediate demand information and the future behavior trend.

[0009] Optionally, pushing a target insurance product to the target user account based on the user context portrait includes: obtaining a pre-constructed insurance knowledge graph; constructing a dynamic rule library based on the insurance knowledge graph, where the dynamic rule library is used to represent the recommendation rules and priorities of insurance products, and the insurance knowledge graph is used to store the knowledge base related to insurance products; searching in the insurance knowledge graph for the target insurance product that best matches the user context portrait based on the dynamic rule library; pushing the target insurance product to the target user account.

[0010] Optionally, before obtaining the pre-constructed insurance knowledge graph, the method further includes: obtaining insurance standard materials and disease standard materials; configuring multiple types of entity objects by using the insurance standard materials and the disease standard materials, where the types of the entity objects include: insurance companies, insurance products, diseases; configuring multiple types of entity relationships, where the types of the entity relationships include: ownership, coverage, applicability, complications; constructing an insurance knowledge graph with the entity objects as nodes and the entity relationships as the relationship edges between adjacent nodes.

[0011] Optionally, constructing the dynamic rule library based on the insurance knowledge graph includes: reading the user fixed events of the target user account based on the user context portrait; creating several general rules based on the user fixed events; reading the user temporary events and environmental trend events of the target user account according to the user context portrait; creating several temporary rules based on the user temporary events and the environmental trend events; creating the binding relationships between the general rules and the temporary rules and the corresponding nodes or corresponding relationship edges in the insurance knowledge graph to obtain the dynamic rule library.

[0012] Optionally, creating the binding relationships between the general rules and the temporary rules and the corresponding nodes or relationship edges in the insurance knowledge graph includes: for each target rule, determining the logical conditions of the target rule; wherein, the target rules include the general rules and the temporary rules, and the logical conditions include the user's fixed events, user's temporary events, and environmental trend events; searching for the recommended objects of the logical conditions in a preset mapping relationship library; locating the target nodes in the insurance knowledge graph that match the recommended objects; creating a first binding relationship between the target rule and the target nodes in the insurance knowledge graph, and / or creating a second binding relationship between the target rule and the relationship edges connected to the target nodes.

[0013] According to another embodiment of the present invention, there is provided a pushing device for insurance products, including: an acquisition module, configured to acquire the user attribute information of a target user account, acquire the dynamic behavior information of the target user account, and acquire the environmental status information of the location where the user device is located, wherein the target user account is logged in on the user device; a construction module, configured to construct a user situation portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information; a pushing module, configured to push a target insurance product to the target user account based on the user situation portrait.

[0014] Optionally, the acquisition module includes: a first acquisition unit, configured to acquire the query data input by the target user account on the search interface; a second acquisition unit, configured to acquire the online browsing records of the target user account; a third acquisition unit, configured to acquire the online public opinion data of the target user account; a fourth acquisition unit, configured to acquire the itinerary information of the user corresponding to the target user account; and / or, the acquisition module includes: a fifth acquisition unit, configured to acquire the geological features of the location where the user device is located; a sixth acquisition unit, configured to acquire the weather type at the location where the user device is located in the future; a seventh acquisition unit, configured to acquire the traffic status information at the location where the user device is located in the future; an eighth acquisition unit, configured to acquire the disease prevalence status at the location where the user device is located; a ninth acquisition unit, configured to acquire the penetration status of insurance products at the location where the user device is located.

[0015] Optionally, the construction module includes: a processing unit, configured to perform data preprocessing on the user attribute information and the dynamic behavior information to obtain multi-source fusion data; an output unit, configured to input the multi-source fusion data into a pre-trained situation awareness model and output the immediate demand information of the target user account; a prediction unit, configured to predict the future behavior trend of the target user account based on the immediate demand information and the environmental status information, wherein the user situation portrait includes the immediate demand information and the future behavior trend.

[0016] Optionally, the pushing module includes: a first obtaining unit configured to obtain a pre-constructed insurance knowledge graph; a first constructing unit configured to construct a dynamic rule base based on the insurance knowledge graph, where the dynamic rule base is used to represent the recommendation rules and priorities of insurance products, and the insurance knowledge graph is used to store a knowledge base related to insurance products; a searching unit configured to search for a target insurance product that best matches the user context portrait in the insurance knowledge graph based on the dynamic rule base; and a pushing unit configured to push the target insurance product to the target user account.

[0017] Optionally, the pushing module further includes: a second obtaining unit configured to obtain insurance standard materials and disease standard materials before the first obtaining unit obtains the pre-constructed insurance knowledge graph; a first configuring unit configured to configure multiple types of entity objects by using the insurance standard materials and the disease standard materials, where the types of the entity objects include: insurance companies, insurance products, and diseases; a second configuring unit configured to configure multiple types of entity relationships, where the types of the entity relationships include: attribution, coverage, applicability, and complication; and a second constructing unit configured to construct an insurance knowledge graph by using the entity objects as nodes and the entity relationships as relationship edges between adjacent nodes.

[0018] Optionally, the first constructing unit includes: a first reading subunit configured to read user fixed events of the target user account based on the user context portrait; a first creating subunit configured to create a number of general rules based on the user fixed events; a second reading subunit configured to read user temporary events and environmental trend events of the target user account according to the user context portrait; a second creating subunit configured to create a number of temporary rules based on the user temporary events and the environmental trend events; and a third creating subunit configured to create a binding relationship between the general rules and the temporary rules and corresponding nodes or corresponding relationship edges in the insurance knowledge graph to obtain a dynamic rule base.

[0019] Optionally, the third creating subunit is further configured to: for each target rule, determine a logical condition of the target rule; where the target rules include the general rules and the temporary rules, and the logical conditions include the user fixed events, user temporary events, and environmental trend events; search for a recommended object of the logical condition in a preset mapping relationship library; locate a target node in the insurance knowledge graph that matches the recommended object; and create a first binding relationship between the target rule and the target node in the insurance knowledge graph, and / or create a second binding relationship between the target rule and relationship edges connected to the target node.

[0020] According to another aspect of the embodiments of the present application, a storage medium is further provided. The storage medium includes a stored program that, when running, executes the above steps.

[0021] According to another aspect of the embodiments of the present application, an electronic device is further provided, including a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus; where: the memory is used to store a computer program; the processor is used to execute the steps in the above method by running the program stored on the memory.

[0022] According to another embodiment of the present invention, a storage medium is further provided. A computer program is stored in the storage medium, where the computer program is set to execute the steps in any one of the above device embodiments when running.

[0023] Through the embodiments of the present invention, user attribute information of a target user account, dynamic behavior information of the target user account, and environmental status information of the location where the user device is located are obtained. Among them, the target user account is logged in to the user device; the user context portrait of the target user account is constructed by using the user attribute information, the dynamic behavior information, and the environmental status information; and a target insurance product is pushed to the target user account based on the user context portrait. By constructing the user context portrait of the target user account and pushing the target insurance product based on this, the technical problem of low efficiency in pushing insurance products in the related art is solved, the personalized matching degree between the insurance product and the user is improved, and the pushing efficiency of the insurance product is also improved. Description of the Drawings

[0024] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of this application, and the schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0025] Figure 1 is a hardware structure block diagram of a server according to an embodiment of the present application;

[0026] Figure 2 is a flowchart of a method for pushing an insurance product according to an embodiment of the present invention;

[0027] Figure 3 is a technical principle diagram of personalized recommendation in an embodiment of the present invention;

[0028] Figure 4 is a structure block diagram of a device for pushing an insurance product according to an embodiment of the present invention. Detailed Embodiments

[0029] To enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application. It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other.

[0030] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0031] Embodiment 1

[0032] The method embodiment provided by the first embodiment of this application can be executed on a server, a computer, a mobile phone, or a similar computing device. Taking running on a server as an example, Figure 1 is a hardware structure block diagram of a server in an embodiment of this application. As Figure 1 shown, the server may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the above-mentioned server may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned server. For example, the server may further include more or fewer components than those shown in Figure 1 the figure, or have a different configuration from that shown in Figure 1 the figure.

[0033] The memory 104 can be used to store server programs, for example, software programs and modules of application software, such as the server program corresponding to the push method of an insurance product in an embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the server program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories can be connected to the server through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and combinations thereof.

[0034] The transmission device 106 is used to receive or send data via a network. Specific examples of the above network may include a wireless network provided by a communication provider of the server. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, abbreviated as NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0035] In this embodiment, a push method of an insurance product is provided. Figure 2 It is a flowchart of a push method of an insurance product according to an embodiment of the present invention, as Figure 2 shown, and the process includes the following steps:

[0036] Step S202, obtaining user attribute information of a target user account, obtaining dynamic behavior information of the target user account, and obtaining environmental status information of the location where the user device is located, where the target user account is logged in on the user device;

[0037] In this embodiment, the user attribute information includes static personal information, such as age, occupation, medical insurance information, income, health status, disease history, etc. Optionally, the dynamic behavior information is to capture the dynamic behavior data of the user in real time through Internet of Things (IoT) devices and mobile applications, such as location, online behavior trajectory, activity pattern, social network interaction, input information of the search interface, etc. The environmental status information refers to the external environmental information of the location where the user is located, such as weather, address characteristics, traffic conditions, disease epidemic status, etc.

[0038] For example, the location information of the user can reveal the risk level of their living environment. For example, users living in areas with frequent natural disasters (such as flood-prone areas or earthquake-prone areas) may need more property insurance or disaster insurance; real-time location data and dynamic behavior information can be used for dynamic risk assessment. For example, when the user is driving, the system can recommend short-term traffic accident insurance or additional car insurance services. The user's dynamic behavior information such as search records and browsing history (such as frequently searching for "health insurance" or "car insurance") can reflect their potential insurance needs; by analyzing the user's attention to different insurance products, the system can recommend products that match their interests (such as travel insurance, health insurance, etc.). The activity patterns (such as exercise frequency and travel habits) in the user's itinerary information can reflect their health risks and accident risks; by analyzing the user's working hours and locations, the system can recommend occupation-related insurance. The online public opinion data (such as discussing insurance topics and sharing claims experiences) of the user on social networks can reflect their concerns or needs regarding insurance; by analyzing the user's emotional state (such as concerns about health and safety), the system can recommend corresponding insurance products (such as medical insurance, critical illness insurance, or accident insurance). External environmental data such as weather, traffic conditions, and high disease incidence seasons can reflect the immediate risks faced by the user, and the system can recommend corresponding types of insurance.

[0039] Step S204: Construct a user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information.

[0040] The user context portrait in this embodiment includes a user interest portrait, a user environment portrait, a real-time scenario portrait, and a future scenario portrait, including the user's fixed biological attribute information (age, occupation, underlying diseases, etc.), the real-time interests actively explored by the user, the user's real-time behaviors, and the future probability events brought about by the fixed biological attribute information, and the unknown environmental trends of the user.

[0041] Step S206: Push target insurance products to the target user account based on the user context portrait.

[0042] In this embodiment, the target insurance products can be pushed on a human-computer interaction interface such as an application program or a session window, which can be the human-computer session window of an online insurance store, an online medical assistant, a chat software, a search software, etc.

[0043] Through the above steps, obtain the user attribute information of the target user account, obtain the dynamic behavior information of the target user account, and obtain the environmental status information of the location where the user device is located. Among them, the target user account is logged in on the user device; construct the user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information; push the target insurance product to the target user account based on the user context portrait. By constructing the user context portrait of the target user account and pushing the target insurance product based on this, the technical problem of low efficiency in pushing insurance products in the related art is solved, the personalized matching degree between the insurance product and the user is improved, and the pushing efficiency of the insurance product is also improved.

[0044] In an implementation manner of this embodiment, obtaining the dynamic behavior information of the target user account includes: obtaining the query data input by the target user account on the search interface; obtaining the online browsing record of the target user account; obtaining the online public opinion data of the target user account; obtaining the itinerary information of the user corresponding to the target user account.

[0045] Among them, the itinerary record includes activity patterns, activity trajectories, locations, etc., and the online public opinion data includes behavior data such as online comments, reply comments, likes, collections, forwards, etc.

[0046] In an implementation manner of this embodiment, obtaining the environmental status information of the location where the user device is located includes: obtaining the geological characteristics of the location where the user device is located; obtaining the weather type at the location where the user device is located at a future time; obtaining the traffic status information at the location where the user device is located at a future time; obtaining the disease prevalence status at the location where the user device is located; obtaining the penetration status of insurance products at the location where the user device is located.

[0047] Among them, the disease prevalence status is only the current high-risk disease types, such as influenza, hand, foot and mouth disease, etc., and the penetration status of insurance products refers to the insurance rate of a certain type or a certain insurance product.

[0048] In an example of this embodiment, constructing the user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information includes: performing data preprocessing on the user attribute information and the dynamic behavior information to obtain multi-source fusion data; inputting the multi-source fusion data into a pre-trained context awareness model to output the immediate demand information of the target user account; predicting the future behavior trend of the target user account based on the immediate demand information and the environmental status information, where the user context portrait includes the immediate demand information and the future behavior trend.

[0049] After obtaining multi-source data (user attribute information, dynamic behavior information, environmental status information), operations such as cleaning, denoising, and normalizing the multi-source data can be performed to ensure data quality. For example, removing outliers from health monitoring data or geocoding location data. Extract meaningful features from the original data and integrate information from different data sources through multi-source data fusion techniques (such as feature-level fusion and decision-level fusion). For example, combining health data with activity patterns to assess the health risks of users. Process real-time data through stream processing techniques (such as Apache Kafka and Apache Flink) to ensure that the recommendation system can quickly respond to the immediate needs of users. Based on the fused data, build a user context model and generate personalized insurance recommendation results by combining collaborative filtering, content recommendation, or hybrid recommendation algorithms. Adopt technologies such as differential privacy and federated learning to protect user privacy and ensure compliance in data usage.

[0050] In one implementation manner of this embodiment, pushing a target insurance product to the target user account based on the user context portrait includes:

[0051] S11, obtaining a pre-constructed insurance knowledge graph;

[0052] In one example, before obtaining the pre-constructed insurance knowledge graph, it further includes: obtaining insurance standard materials and disease standard materials; configuring multiple types of entity objects using the insurance standard materials and the disease standard materials, where the types of the entity objects include: insurance companies, insurance products, and diseases; configuring multiple types of entity relationships, where the types of the entity relationships include: attribution, coverage, applicability, and complication; constructing an insurance knowledge graph with the entity objects as nodes and the entity relationships as the relationship edges between adjacent nodes.

[0053] Extract entities (such as users, insurance products, diseases, services) and relationships (such as "users are concerned about diseases" and "insurance products cover diseases") from multi-source data (such as obtaining insurance standard materials and disease standard materials from insurance company official websites, health commission websites, and disease quality guidance manuals); store the insurance knowledge graph using a graph database (such as Neo4j and TigerGraph) and support efficient graph traversal and semantic queries.

[0054] In this embodiment, the insurance knowledge graph is constructed based on a graph structure, where the entity objects represent the nodes in the insurance knowledge graph, and the relationship edges represent the associations between the entity nodes.

[0055] For example, the nodes include:

[0056] Insurance companies: such as "China Life" and "Ping An Insurance".

[0057] Insurance products: such as "Critical illness insurance", "Medical insurance".

[0058] Diseases: such as "Diabetes", "Hypertension".

[0059] Users: such as "Age", "Occupation", "Health status", "Hobbies", etc.

[0060] For example, the relationship edges include:

[0061] Insurance company - Insurance product: such as "Provides".

[0062] Insurance product - Disease: such as "Covers".

[0063] User - Insurance product: such as "Suitable for".

[0064] Disease - Disease: such as "Complications".

[0065] Neo4j can be used to construct an insurance knowledge graph. Neo4j is a graph database management system suitable for processing complex node and relationship networks. The construction steps include:

[0066] 1) Create nodes: Create nodes for each entity and assign unique identifiers and attributes.

[0067] CREATE(c:Company{name:"China Life Insurance"})

[0068] CREATE(p:Product{name:"Critical illness insurance"})

[0069] CREATE(d:Disease{name:"Diabetes"})

[0070] CREATE(u:User{age:30,occupation:"Engineer"})

[0071] 2) Create relationships: Define the relationships between nodes.

[0072] MATCH(c:Company{name:"China Life Insurance"}),(p:Product{name:"Critical illness insurance"})

[0073] CREATE(c)-[:PROVIDES]->(p)

[0074] MATCH(p:Product{name:"Critical illness insurance"}),(d:Disease{name:"Diabetes"})

[0075] CREATE(p)-[:COVERS]->(d)

[0076] MATCH(u:User{age:30}),(p:Product{name:"Critical Illness Insurance"})

[0077] CREATE(u)-[:SUITABLE_FOR]->(p)

[0078] 3. Correspondence between entity nodes and relationship edges

[0079] Entity nodes: Represent specific objects or concepts, such as insurance companies, insurance products, diseases, and users.

[0080] Relationship edges: Represent the specific associations between these objects, such as "provide", "cover", "suitable for", etc.

[0081] S12. Construct a dynamic rule base based on the insurance knowledge graph, where the dynamic rule base is used to represent the recommendation rules and priorities of insurance products, and the insurance knowledge graph is used to store the knowledge base related to insurance products;

[0082] The dynamic rule base in this embodiment is a rule mechanism that can be automatically adjusted according to external environmental changes, user behavior data, market trends, etc. The dynamic rules are generated based on machine learning models, real-time data analysis, or user feedback. It can be updated in real time or regularly to adapt to the changing situational requirements. The dynamic rules are used to adapt to market changes in real time, automatically adjust the recommendation strategy, and dynamically adjust the risk assessment rules to ensure that the recommendation results match the latest needs of users.

[0083] In one example, constructing a dynamic rule base based on the insurance knowledge graph includes: reading the user's fixed events of the target user account based on the user situation portrait; creating several general rules based on the user's fixed events; reading the user's temporary events and environmental trend events of the target user account according to the user situation portrait; creating several temporary rules based on the user's temporary events and the environmental trend events; creating the binding relationships between the general rules and the temporary rules and the corresponding nodes or corresponding relationship edges in the insurance knowledge graph to obtain the dynamic rule base.

[0084] Optionally, creating the binding relationships between the general rules and the temporary rules and the corresponding nodes or relationship edges in the insurance knowledge graph includes: for each target rule, determining the logical conditions of the target rule; wherein, the target rules include the general rules and the temporary rules, and the logical conditions include the user's fixed events, user's temporary events, and environmental trend events; searching for the recommended objects of the logical conditions in a preset mapping relationship library; locating the target nodes in the insurance knowledge graph that match the recommended objects; creating a first binding relationship between the target rule and the target nodes in the insurance knowledge graph, and / or creating a second binding relationship between the target rule and the relationship edges connected to the target nodes.

[0085] The association is achieved by binding all general rules and temporary rules to entity nodes and relationship edges in the insurance knowledge graph.

[0086] In this embodiment, a rule (including a general rule and a temporary rule) is a logical expression used to describe the recommended objects (such as insurance products, insurance companies, etc.) triggered by the occurrence of a certain logical condition. For example: if the user's age is greater than 50 years old, then recommend "critical illness insurance"; if the user has a history of diabetes, then recommend "medical insurance"; if an insurance product covers a certain disease, then add it to the recommended list.

[0087] Each target rule can include logical conditions and actions (recommended objects). Logical conditions: based on the node attributes or path relationships in the knowledge graph; actions (recommended objects): operations triggered according to the conditions, such as recommending a certain insurance product.

[0088] When associating a rule with the knowledge graph, 1) the rule can be directly associated with a certain entity node in the knowledge graph. For example: rule condition: the user's age > 50 years old. Associated node: the "user" node in the knowledge graph. Associated attribute: the "age" attribute of the user node. 2) The rule can be associated with the relationship edges in the knowledge graph (i.e., the relationship chains between nodes). For example: rule condition: the user has diabetes. Associated path: user -> [has] -> disease -> [covers] -> insurance product.

[0089] The relational rule engine (such as Drools) is integrated with the knowledge graph (such as Neo4j) through an API; the rule engine calls the query interface of the knowledge graph to obtain relevant nodes or paths and performs reasoning according to the rule conditions.

[0090] In the process of converting the trend hotspots corresponding to user temporary events and environmental trend events into temporary rules, it includes identification, abstraction, rule generation, execution, and optimization.

[0091] First, the system needs to be able to identify trends and hot issues in market dynamics or user feedback. The identification methods include: 1) obtaining the latest insurance product information, policy changes, etc. from official channels (such as insurance company announcements, industry reports); 2) using web crawlers or APIs to scrape hot topics on platforms such as social media and news websites; 3) collecting users' behavioral data (such as search records, click behaviors, purchase records); 4) analyzing users' comments, complaints or suggestions to extract high-frequency keywords or themes; 5) using natural language processing (NLP) technology to perform sentiment analysis and topic extraction on text data.

[0092] Abstract the trend hotspots into the conditions and actions of rules, and transform the abstract logical expressions into executable temporary rules. For example: IF user age > 50 AND disease = diabetes THEN recommend special medical insurance for diabetes; if the user is located in a flood-prone area, then recommend "flood insurance"; if the current time is the high-incidence season of influenza, then recommend "medical insurance", etc.

[0093] Bind the conditions in the rules to the nodes or paths in the knowledge graph.

[0094] Store the temporary rules in the rule library and mark them as "temporary rules" with an expiration date set.

[0095] During the validity period, use methods such as A / B testing to compare the effects of temporary rules and regular rules. Such as recommendation success rate, user satisfaction.

[0096] If the temporary rule is no longer applicable (such as the hot issue subsides), then archive or delete it. If the effect of the temporary rule is significant, it can be converted into a general rule.

[0097] In this embodiment, when generating a dynamic rule library, general rules and temporary rules can be generated based on a machine learning model. Using the user context portrait as input data, training a machine learning model (such as a decision tree, random forest) based on the real-time data of the user context portrait (such as user behavior, market trend) to generate a dynamic rule library; using a rule engine (such as Drools, Jess) to execute the rules and optimize the rule logic according to the semantic associations of the knowledge graph.

[0098] S13, search in the insurance knowledge graph for the target insurance product that best matches the user context portrait based on the dynamic rule library;

[0099] Mining the association between user needs and insurance products through graph traversal algorithms (such as PageRank, shortest path) of the insurance knowledge graph; combining the logical reasoning ability of the rule engine to generate personalized recommendation results. For example: User is concerned about a certain disease → The knowledge graph is associated with relevant health insurance → The rule engine filters suitable products according to conditions such as user age and income; supporting multi-hop reasoning, for example: User is concerned about a disease → The knowledge graph is associated with preventive measures → Further associated with health management services → The rule engine generates a comprehensive recommendation.

[0100] The dynamic rule base of this embodiment is the core of the rule engine. It supports fast, flexible, and transparent decision-making and recommendation by storing and executing predefined rules. It can generate personalized recommendation results based on the behavioral data in the user scenario portrait and market trends in combination with the insurance knowledge graph, and adapt to changing needs through a dynamic update mechanism. The combination of the dynamic rule base with technologies such as the insurance knowledge graph and machine learning further improves the intelligence level and reasoning ability of the system.

[0101] S14, pushing the target insurance product to the target user account.

[0102] The dynamic rule base of this embodiment is updated according to a cycle, monitoring market changes and user behavior through real-time data processing technologies (such as Apache Kafka, Apache Flink). Using reinforcement learning or online learning models to dynamically adjust the rule base. For example, when a certain type of insurance product becomes popular in a specific region, the system automatically generates new rules to give priority to recommending this type of product.

[0103] It is also possible to optimize the dynamic rule base and the insurance knowledge graph based on user feedback, collecting user feedback on recommendation results (such as clicks, purchases, ratings) for optimizing rules and the insurance knowledge graph. Optimizing models and rules under the premise of protecting user privacy through technologies such as federated learning or differential privacy.

[0104] This solution proposes an intelligent insurance product recommendation system that combines a rule engine, a knowledge graph, and real-time context awareness technology. The system aims to provide more accurate and personalized insurance product recommendation services by integrating multi-source information. Figure 3 It is the technical principle diagram of personalized recommendation in the embodiments of the present invention, including: real-time context awareness of multi-source information fusion, deep fusion of the rule engine and the knowledge graph, improvement of personalized user experience, continuous learning and self-optimization.

[0105] In the process of real-time context awareness through multi-source information fusion, various data sources such as IoT devices, mobile applications, and third-party APIs (such as weather forecasts and traffic information) are integrated. ETL (Extract, Transform, Load) tools are used to clean, transform, and load data in these different formats to ensure that all data can be uniformly processed. A message queue system based on Apache Kafka or a similar one is built to efficiently transmit and manage real-time data streams, and stream processing engines such as Spark Streaming or Flink are used to implement real-time data analysis. A large number of user behavior samples are collected and typical features are labeled, and deep learning algorithms (such as RNN (Recurrent Neural Network) and LSTM (Long Short-Term Memory)) are used to train the context awareness model to identify users' immediate needs and predict their future behavior trends. The descriptions of users' immediate needs and behavior predictions generated for real-time contexts after processing are sent to the recommendation engine module as an important basis for personalized recommendations.

[0106] In the process of deep integration of the dynamic rule base and the knowledge graph, knowledge graph construction: Data such as the latest insurance company information, insurance product manuals, and disease classification standards are obtained from official channels, and a knowledge graph containing entity nodes and their relationship edges is constructed using Neo4j or other graph database management systems. Rule definition platform development: A graphical rule editing interface is developed to allow administrators to intuitively create, modify, and delete rules, and each rule is associated with specific nodes or paths in the knowledge graph. Semantic reasoning engine integration: The SPARQL (SPARQL Protocol and RDF Query Language) query language and OWL (Web Ontology Language) ontology are introduced to support complex logical reasoning operations. When a user queries, the knowledge graph is automatically traversed to find eligible products, and the final recommendation list is generated in combination with the rule base. Rule dynamic update mechanism: Regular scanning tasks are set up to monitor market dynamics and user feedback, and relevant rules are adjusted in a timely manner. For newly emerging trends or hot issues, corresponding temporary rules are quickly added to ensure the flexibility and response speed of the system.

[0107] In the process of enhancing the personalized user experience, a pre-trained model based on the Transformer architecture (such as BERT (Bidirectional Encoder Representations from Transformers)) is trained to have powerful natural language understanding and generation capabilities, and a dialogue management module specifically for the insurance field is constructed, including functions such as intent recognition, slot filling, and context tracking. A simple and intuitive front-end page is designed to support various interaction methods such as text input and speech recognition, and detailed insurance product introductions, clause explanations, etc. are provided to help users understand the recommended results. A large number of real user conversation records are collected, correct and incorrect intent recognition cases are marked, and the intent recognition model is continuously iteratively improved using reinforcement learning algorithms to improve its accuracy and robustness.

[0108] In the process of continuous learning and self-optimization, a convenient user feedback collection system is established to encourage users to provide opinions such as ratings and comments, and satisfaction surveys are regularly launched to deeply understand the real experience and improvement suggestions of users. An automated workflow is developed to directly transfer user feedback to the person in charge of each module, and the recommendation rules are adjusted, the knowledge graph structure is optimized, and the NLP model is improved according to the feedback content. An online learning algorithm is implemented so that the system can gradually introduce new training data without affecting normal services, and an A / B test environment is set up to compare the performance of the new and old models to ensure that each update can bring performance improvement.

[0109] Adopting the solution of this embodiment, combining dynamic rules with a dynamic knowledge graph realizes efficient and accurate recommendation, collects, processes, and applies the real-time behavior data of users to generate a user context portrait to ensure a high degree of relevance of the recommended results. Through continuous learning and optimization, it is ensured that the system can operate stably in the long term and continuously improve the recommendation quality and service level over time.

[0110] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in various embodiments of the present invention.

[0111] Embodiment 2

[0112] In this embodiment, a pushing device for insurance products is further provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be repeated. As used hereinafter, the term "module" can be a combination of software and hardware that can achieve a predefined function. Although the devices described in the following embodiments are preferably implemented in software, implementations in hardware, or a combination of software and hardware, can also be conceived.

[0113] Figure 4 is a structural block diagram of a pushing device for insurance products according to an embodiment of the present invention. As Figure 4 shown, it includes:

[0114] An acquisition module 40, configured to acquire user attribute information of a target user account, acquire dynamic behavior information of the target user account, and acquire environmental status information of the location where the user device is located, where the target user account is logged in to the user device;

[0115] A construction module 42, configured to construct a user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information;

[0116] A pushing module 44, configured to push a target insurance product to the target user account based on the user context portrait.

[0117] Optionally, the acquisition module includes: a first acquisition unit, configured to acquire query data input by the target user account on the search interface; a second acquisition unit, configured to acquire the online browsing records of the target user account; a third acquisition unit, configured to acquire the online public opinion data of the target user account; a fourth acquisition unit, configured to acquire the itinerary information of the user corresponding to the target user account; and / or, the acquisition module includes: a fifth acquisition unit, configured to acquire the geological features of the location where the user device is located; a sixth acquisition unit, configured to acquire the weather type at the location where the user device is located in the future; a seventh acquisition unit, configured to acquire the traffic status information at the location where the user device is located in the future; an eighth acquisition unit, configured to acquire the disease prevalence status at the location where the user device is located; a ninth acquisition unit, configured to acquire the penetration status of insurance products at the location where the user device is located.

[0118] Optionally, the construction module includes: a processing unit, configured to perform data preprocessing on the user attribute information and the dynamic behavior information to obtain multi-source fusion data; an output unit, configured to input the multi-source fusion data into a pre-trained context awareness model and output the immediate demand information of the target user account; a prediction unit, configured to predict the future behavior trend of the target user account based on the immediate demand information and the environmental status information, where the user context portrait includes the immediate demand information and the future behavior trend.

[0119] Optionally, the pushing module includes: a first obtaining unit configured to obtain a pre-constructed insurance knowledge graph; a first constructing unit configured to construct a dynamic rule base based on the insurance knowledge graph, where the dynamic rule base is used to represent the recommendation rules and priorities of insurance products, and the insurance knowledge graph is used to store a knowledge base related to insurance products; a searching unit configured to search for a target insurance product that best matches the user situation portrait in the insurance knowledge graph based on the dynamic rule base; and a pushing unit configured to push the target insurance product to the target user account.

[0120] Optionally, the pushing module further includes: a second obtaining unit configured to obtain insurance standard materials and disease standard materials before the first obtaining unit obtains the pre-constructed insurance knowledge graph; a first configuring unit configured to configure multiple types of entity objects by using the insurance standard materials and the disease standard materials, where the types of the entity objects include: insurance companies, insurance products, and diseases; a second configuring unit configured to configure multiple types of entity relationships, where the types of the entity relationships include: belonging to, covering, applying to, and complication; and a second constructing unit configured to construct an insurance knowledge graph by using the entity objects as nodes and the entity relationships as relationship edges between adjacent nodes.

[0121] Optionally, the first constructing unit includes: a first reading subunit configured to read user fixed events of the target user account based on the user situation portrait; a first creating subunit configured to create a number of general rules based on the user fixed events; a second reading subunit configured to read user temporary events and environmental trend events of the target user account according to the user situation portrait; a second creating subunit configured to create a number of temporary rules based on the user temporary events and the environmental trend events; and a third creating subunit configured to create a binding relationship between the general rules and the temporary rules and corresponding nodes or corresponding relationship edges in the insurance knowledge graph to obtain a dynamic rule base.

[0122] Optionally, the third creating subunit is further configured to: for each target rule, determine the logical condition of the target rule; where the target rules include the general rules and the temporary rules, and the logical conditions include the user fixed events, user temporary events, and environmental trend events; search for a recommended object of the logical condition in a preset mapping relationship library; locate a target node in the insurance knowledge graph that matches the recommended object; and create a first binding relationship between the target rule and the target node in the insurance knowledge graph, and / or create a second binding relationship between the target rule and relationship edges connected to the target node.

[0123] It should be noted that the above-mentioned modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above-mentioned modules are all located in the same processor; or, the above-mentioned modules are respectively located in different processors in any combination form.

[0124] Embodiment 3

[0125] An embodiment of the present invention also provides a storage medium, in which a computer program is stored. Wherein, the computer program is set to execute the steps in any one of the above method embodiments when running.

[0126] Optionally, in this embodiment, the above storage medium can be set to store a computer program for execution:

[0127] S1, obtain the user attribute information of the target user account, obtain the dynamic behavior information of the target user account, and obtain the environmental status information of the location where the user device is located, wherein the target user account is logged in on the user device;

[0128] S2, construct a user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information;

[0129] S3, push a target insurance product to the target user account based on the user context portrait.

[0130] Optionally, in this embodiment, the above storage medium may include but is not limited to: various media such as a USB flash drive, a read-only memory (ROM for short), a random access memory (RAM for short), a mobile hard disk, a magnetic disk, or an optical disc that can store computer programs.

[0131] An embodiment of the present invention also provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is set to run the computer program to execute the steps in any one of the above method embodiments.

[0132] Optionally, the above electronic device may further include a transmission device and an input / output device. Wherein, the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0133] Optionally, in this embodiment, the above processor can be set to execute the following steps through a computer program:

[0134] S1. Obtain the user attribute information of the target user account, obtain the dynamic behavior information of the target user account, and obtain the environmental status information of the location where the user device is located, where the target user account is logged in to the user device;

[0135] S2. Construct a user context portrait of the target user account by using the user attribute information, the dynamic behavior information, and the environmental status information;

[0136] S3. Push a target insurance product to the target user account based on the user context portrait.

[0137] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation manners, and will not be elaborated herein.

[0138] The serial numbers of the embodiments of the present application above are only for description and do not represent the advantages and disadvantages of the embodiments.

[0139] In the above embodiments of the present application, the descriptions of the respective embodiments have their own emphases. For parts not elaborated in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0140] In the several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0141] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0142] In addition, the functional units in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0143] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a controller, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), mobile hard disks, magnetic disks, or optical discs.

[0144] The above are only the preferred embodiments of this application. It should be noted that for those of ordinary skill in the art, without departing from the principle of this application, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of this application.

Claims

1. A method for pushing insurance products, characterized in that: include: Acquire user attribute information of a target user account, acquire dynamic behavior information of the target user account, and acquire environmental status information of a location of a user device, wherein the target user account is logged in on the user device; Constructing a user context profile of the target user account using the user attribute information, the dynamic behavior information, and the environmental status information; Push target insurance products to the target user account based on the user situation profile.

2. The method according to claim 1, characterized in that Obtaining dynamic behavior information of the target user account includes: obtaining query data entered by the target user account in the search interface; obtaining online browsing history of the target user account; obtaining online public opinion data of the target user account; obtaining itinerary information of the user corresponding to the target user account; and / or, Obtaining environmental status information of the location of the user device includes: obtaining geological features of the location of the user device; obtaining weather types at the location of the user device in the future; obtaining traffic status information at the location of the user device in the future; obtaining disease epidemic status at the location of the user device; and obtaining penetration status of insurance products at the location of the user device.

3. The method according to claim 1, characterized in that Using the user attribute information, the dynamic behavior information, and the environmental status information to construct a user context profile of the target user account includes: Performing data preprocessing on the user attribute information and the dynamic behavior information to obtain multi-source fusion data; Inputting the multi-source fusion data into a pre-trained context-aware model, and outputting the instant demand information of the target user account; The future behavior trend of the target user account is predicted based on the immediate demand information and the environmental status information, wherein the user situation portrait includes the immediate demand information and the future behavior trend.

4. The method according to claim 1, characterized in that: Pushing a target insurance product to the target user account based on the user situation profile includes: Get pre-built insurance knowledge graphs; Building a dynamic rule base based on the insurance knowledge graph, wherein the dynamic rule base is used to characterize recommendation rules and priorities of insurance products, wherein the insurance knowledge graph is used to store a knowledge base related to insurance products; Searching the insurance knowledge graph for a target insurance product that best matches the user context profile based on the dynamic rule base; The target insurance product is pushed to the target user account.

5. The method according to claim 4, characterized in that Before obtaining the pre-built insurance knowledge graph, the method further includes: Obtain insurance standard information and disease standard information; The insurance standard data and the disease standard data are used to configure multiple types of entity objects, wherein the types of the entity objects include: insurance companies, insurance products, and diseases; Configuring multiple types of entity relationships, wherein the types of entity relationships include: attribution, coverage, applicability, and complications; An insurance knowledge graph is constructed with the entity objects as nodes and the entity relationships as relationship edges between adjacent nodes.

6. The method according to claim 4, characterized in that Building a dynamic rule base based on the insurance knowledge graph includes: Reading fixed user events of the target user account based on the user situation portrait; Creating a plurality of general rules based on the user fixed events; Reading user temporary events and environmental trend events of the target user account according to the user situation profile; Creating a plurality of temporary rules based on the user temporary event and the environmental trend event; A binding relationship between the general rules and the temporary rules and corresponding nodes or corresponding relationship edges in the insurance knowledge graph is created to obtain a dynamic rule base.

7. The method according to claim 6, characterized in that Creating a binding relationship between the general rule and the temporary rule and corresponding nodes or corresponding relationship edges in the insurance knowledge graph includes: For each target rule, determining the logical condition of the target rule; wherein the target rule includes the general rule and the temporary rule, and the logical condition includes the user fixed event, the user temporary event, and the environmental trend event; Searching for a recommended object of the logical condition in a preset mapping relationship library; Locating a target node matching the recommended object in the insurance knowledge graph; A first binding relationship is created between the target rule and the target node in the insurance knowledge graph, and / or a second binding relationship is created between the target rule and the relationship edge connecting the target node.

8. A pushing device for insurance products, characterized in that: include: an acquisition module, used to acquire user attribute information of a target user account, acquire dynamic behavior information of the target user account, and acquire environmental status information of a location of a user device, wherein the target user account is logged in on the user device; A construction module, configured to construct a user context profile of the target user account using the user attribute information, the dynamic behavior information, and the environmental status information; A push module is used to push the target insurance product to the target user account based on the user situation portrait.

9. A storage medium, characterized in that: A computer program is stored in the storage medium, wherein the computer program is configured to execute the steps of the method for pushing an insurance product in any one of claims 1 to 7 when running.

10. An electronic device comprising a processor, a communication interface, a memory and a communication bus, wherein: The processor, the communication interface, and the memory communicate with each other via a communication bus; wherein: Memory, used to store computer programs; A processor, configured to execute the steps of the insurance product push method according to any one of claims 1 to 7 by running a program stored in a memory.

Images