CN113191911A - Insurance recommendation method, system, equipment and medium based on user information - Google Patents
Insurance recommendation method, system, equipment and medium based on user information Download PDFInfo
- Publication number
- CN113191911A CN113191911A CN202110740664.1A CN202110740664A CN113191911A CN 113191911 A CN113191911 A CN 113191911A CN 202110740664 A CN202110740664 A CN 202110740664A CN 113191911 A CN113191911 A CN 113191911A
- Authority
- CN
- China
- Prior art keywords
- user
- insurance
- information
- similarity
- insurance product
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/08—Insurance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Public Health (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Medical Informatics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Development Economics (AREA)
- Pathology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention provides an insurance recommendation method, system, equipment and medium based on user information, which comprises the following steps: acquiring a face image of a user, and identifying the face image to obtain identity information of the user; acquiring personal information of a user according to the identity information, and constructing a user portrait by using the personal information; evaluating the user representation to generate a weighting factor for the user with respect to risk; generating an insurance product-attribute matrix and a user-insurance product scoring matrix by using a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of the insurance product, and calculating the user-insurance product scoring matrix to obtain a second similarity; carrying out weighting calculation on the first similarity and the second similarity by using the weighting factor to obtain comprehensive similarity; and matching the most similar K neighbors for the user by utilizing the comprehensive similarity to obtain a recommendation scheme of the insurance product. According to the invention, based on the face matching user information, the risk weighting factor of the user is analyzed to perform weighted recommendation, so that the accuracy of insurance recommendation is improved.
Description
Technical Field
The invention belongs to the technical field of data processing, and particularly relates to an insurance recommendation method, system, equipment and medium based on user information.
Background
The rapid development of the internet brings certain impact to the traditional insurance industry, and the traditional customer-obtaining mode is more and more difficult for insurance agents, but the mobile internet also provides a new customer-obtaining channel and mode, and provides more new possibilities for the customer-obtaining mode.
The traditional customer-obtaining mode is a face-to-face mode generally, the insurance agent can obtain some basic information of the user in the process of communicating with the user, such as approximate age, cultural level, health condition, clothing level and the like, and the insurance agent can rapidly provide insurance products meeting the requirements of the user according to the basic information in the process of communicating with the user. However, after the user obtains the customer by using a telephone or a network, the insurance agent cannot quickly grasp the basic information of the user, and cannot provide the user with an insurance product meeting the user requirement.
Therefore, there is a need in the art for a method for accurately recommending insurance products to a user based on the user's condition.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, an object of the present invention is to provide a method, a system, a device and a medium for recommending insurance based on user information, which are used to solve the problem that in the prior art, when insurance is recommended based on user information, insurance products cannot be accurately recommended according to user conditions.
To achieve the above and other related objects, a first aspect of the present invention provides an insurance recommendation method based on user information, including:
acquiring a face image of a user, and identifying the face image to obtain identity information of the user;
acquiring personal information of a user according to the identity information, and constructing a user portrait by using the personal information;
evaluating the user representation to generate a weighting factor for the user with respect to risk;
generating an insurance product-attribute matrix and a user-insurance product scoring matrix by utilizing a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
carrying out weighting calculation on the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
and matching the most similar K neighbors for the user by using the comprehensive similarity to obtain a recommendation scheme of the insurance product.
In an embodiment of the first aspect, a face image of a user is obtained; performing feature extraction on the face image based on all sequentially connected residual blocks in a depth residual error network to obtain face feature information, wherein any one residual block comprises an identity mapping and at least two convolution layers, and the identity mapping of any one residual block points to the output end of any one residual block from the input end of any one residual block; and performing face retrieval in a face database based on the face feature information to obtain a face retrieval result, wherein the face database stores the corresponding relation between the face feature information and the identity information, and the face retrieval result at least comprises the identity information matched with the target face feature information.
In an embodiment of the first aspect, the method further includes: acquiring personal information of the user according to the identity information, wherein the personal information comprises basic information, health information and insurance information of the user; and constructing a user portrait from multiple dimensions of basic information, health information and insurance information according to the personal information.
In an embodiment of the first aspect, the method further includes: and evaluating the user portrait by using a risk association model, and calculating a weighting factor of the risk of the user based on the relationship between the health information and the insurance information.
In an embodiment of the first aspect, the method further includes:
integrating the insurance product-attribute table to obtain an insurance product-attribute matrix; calculating a first similarity of the insurance products according to the insurance product-attribute matrix;
quantifying user data to obtain a user-insurance product rating table, and integrating data in the user-insurance product rating table to obtain a user-insurance product rating matrix; and obtaining a rating vector of the insurance product based on the user-insurance product rating matrix, and calculating the user-insurance product rating matrix from the user perspective by adopting cosine similarity to obtain a second similarity of the insurance product.
In an embodiment of the first aspect, the method further includes:
and matching the most similar K neighbors of the target insurance product to the user by using the comprehensive similarity to form a nearest neighbor set, calculating the prediction score of the user on the insurance product according to the score vector of the user and the similarity vector of the insurance product, and obtaining a recommendation scheme of the insurance product from high to low according to the prediction score.
In an embodiment of the first aspect, the method further includes: and acquiring a positive feedback factor and a negative feedback factor of the trust between the users, correcting the original second similarity between the users according to the feedback factor, and performing linear weighting on the optimized trust and the second similarity of the users to obtain the optimized second similarity.
A second aspect of the present invention provides an insurance recommendation system based on user information, including:
the acquisition module is used for acquiring a face image of a user and identifying the face image to obtain identity information of the user;
the portrait construction module is used for acquiring personal information of the user according to the identity information and constructing a portrait of the user by utilizing the personal information;
a risk assessment module for assessing the user profile to generate a weighting factor for the user with respect to risk;
the first calculation module is used for generating an insurance product-attribute matrix and a user-insurance product scoring matrix by utilizing a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
the second calculation module is used for performing weighted calculation on the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
and the insurance recommending module is used for matching the most similar K neighbors for the user by utilizing the comprehensive similarity to obtain a recommending scheme of the insurance product.
A third aspect of the present invention provides an insurance recommendation apparatus based on user information, including:
one or more processing devices;
a memory for storing one or more programs; when the one or more programs are executed by the one or more processing devices, the one or more processing devices are caused to implement the insurance recommendation method based on the user information.
A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is configured to cause the computer to execute the above-mentioned insurance recommendation method based on user information.
As described above, the technical solution of the insurance recommendation method, system, device and medium based on user information according to the present invention has the following beneficial effects:
the collaborative filtering algorithm based on the insurance product and the recommendation algorithm based on the insurance product content are combined for use, and then the weighting factor is combined for weighted recommendation, so that the requirement on the novelty of user recommendation is met, the recommendation accuracy is improved, meanwhile, the weighting factor of the user portrait risk is introduced, the real requirement of the user and the product attribute are associated, and the recommendation accuracy of the insurance product is further improved on the original basis.
Drawings
FIG. 1 is a flow chart of an insurance recommendation method based on user information according to the present invention;
FIG. 2 is a block diagram of an insurance recommendation system based on user information according to the present invention;
fig. 3 is a schematic structural diagram of an insurance recommendation apparatus based on user information according to the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
The invention solves the problems that in the prior art, for example, a user goes to a hospital to perform normal physical examination, except for abnormal indexes of self physical examination information, health risks to be caused by the physical examination are unknown, a reasonable and legal insurance product is recommended to the user according to the user physical examination information, the health risks are not known all the time, and if the user goes to the physical examination after suffering from a certain disease, the suspicion of putting a disease into insurance is always existed, so that when the user carries out the claim settlement, the claim settlement dispute occurs, therefore, on the basis of knowing the body health, the risk of the disease of the user is estimated through probability, and the purpose of improving the accurate recommendation of the insurance product is urgently needed to be solved.
Referring to fig. 1, a flowchart of an insurance recommendation method based on user information according to the present invention includes:
step S1, collecting a face image of a user, and identifying the face image to obtain identity information of the user;
the face acquisition terminal installed in a certain area (physical examination center, hospital, nursing center, etc.) sends the acquisition terminal to the server for identification, and obtains the identity information of the user, which is not described herein again.
Step S2, acquiring the personal information of the user according to the identity information, and constructing the user portrait by using the personal information;
the user portrait is constructed according to the personal information of the user by acquiring the personal information of the user from the associated local database.
Step S3, evaluating the user profile to generate a weighting factor of the user about the risk;
step S4, generating an insurance product-attribute matrix and a user-insurance product scoring matrix by using a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
step S5, weighting and calculating the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
and step S6, matching the most similar K neighbors for the user by using the comprehensive similarity to obtain a recommendation scheme of the insurance product.
In step S1, specifically, the method includes: acquiring a face image of a user; performing feature extraction on the face image based on all sequentially connected residual blocks in a depth residual error network to obtain face feature information, wherein any one residual block comprises an identity mapping and at least two convolution layers, and the identity mapping of any one residual block points to the output end of any one residual block from the input end of any one residual block; and performing face retrieval in a face database based on the face feature information to obtain a face retrieval result, wherein the face database stores the corresponding relation between the face feature information and the identity information, and the face retrieval result at least comprises the identity information matched with the target face feature information.
Specifically, the implementation environment comprises a terminal, a face retrieval system and a face database. The face retrieval system is a server, and the face retrieval system and the face database may be configured on the same server or different servers. Types of terminals include, but are not limited to, smart phones, desktop computers, laptop computers, tablet computers, and the like.
In the embodiment of the invention, the terminal and the face retrieval system are based on a RESTful architecture mode, namely, the terminal and the face retrieval system adopt a client/server mode to carry out internet communication. Because the embodiment of the invention provides a RESTful standard protocol interface based on a RESTful architecture, one server can be configured to be accessed by a plurality of clients, and the method and the system are convenient and quick.
In addition, because the data stored in the face database is changed in real time, the face retrieval service is deployed in a set of distributed system, so that not only can a lot of resources and workload be saved, but also a large number of requests can be processed quickly and concurrently. The problem that when each piece of software is independently deployed with face retrieval, the updating task of the database is heavy is solved. For example, most of the software architectures are embedded applications installed in devices such as smart phones and tablet computers, and for this mode, when a data update occurs, the data update needs to be configured on a large number of devices individually, and the task of updating the database is very huge.
Note that identity mapping: for any set a, if the mapping f: a → a is defined as f (a) ═ a, i.e. it is specified that each element a in a corresponds to itself, then f is called an identity mapping on a.
RESTful architecture: RESTful refers to a software architecture style, design style, rather than a standard, that provides a set of design principles and constraints. The method is mainly used for the interactive software of the client and the server. Software designed based on the style can be simpler, more hierarchical and easier to realize mechanisms such as cache and the like.
It should be further noted that the face retrieval method provided by the embodiment of the present invention is innovatively designed from two aspects of a specific face retrieval mode and a software architecture. On one hand, the ResNet network structure is used as a specific algorithm for face retrieval, the face features are learned by a deeper network layer number, and a more accurate face matching and comparison effect is obtained. On the other hand, the embodiment of the invention adopts a RESTful standard-based software architecture, not only can meet the requirement of static retrieval, but also can conveniently configure a large-scale distributed retrieval system, and has higher practical value in the fields of investigation and control, criminal investigation and case handling, security and protection activities and the like.
The face retrieval system mainly comprises a face retrieval service module and a feature extraction service module.
The face retrieval service module is used for warehousing face characteristic information and retrieving the face characteristic information; and the characteristic extraction service module is used for extracting the characteristics of the large-batch images.
For the face retrieval service module, when face feature information is put in storage, the face retrieval service module can be realized in a form of a feature file. The feature file comprises one-to-one identity identification and face feature information. In the embodiment of the invention, the face feature information can be directly put into a storage by calling the storage interface of the face retrieval service module. When face feature information is retrieved, the operations that the face retrieval service module can perform include but are not limited to: after the image is input by the client, base64 coding of the image is automatically carried out, comparison of the similarity of the human face is carried out in a library, and the human face features with the similarity reaching a threshold value are retrieved and output.
For the feature extraction service module, the embodiment of the invention adopts the ResNet network to extract the features of the face image. And because the ResNet network introduces a residual error network structure, the problem of gradient dispersion caused by too deep network layers is solved, the characteristic learning of the face image can be carried out by using a deeper network structure, and the accuracy of face retrieval is ensured. In the embodiment of the present invention, a face image refers to an image including a face.
In summary, the embodiment of the invention adopts the depth residual error network ResNet to perform face retrieval, and solves the problem that the deeper the network, the more obvious the gradient dispersion phenomenon is, and the worse the network training effect is. Compared with other network models, the ResNet network can make the number of network layers deep, even up to 1000 layers, so that a good learning effect of the face feature information can be obtained. In addition, the embodiment of the invention integrates the algorithm with the platform, provides the face retrieval service in the HTTP service mode, and can provide a Restful standard protocol interface to the outside.
In another embodiment, step S2 further includes: acquiring personal information of the user according to the identity information, wherein the personal information comprises basic information, health information and insurance information of the user; and constructing a user portrait from multiple dimensions of basic information, health information and insurance information according to the personal information.
It should be noted that the basic information includes personal basic information such as gender, age, occupation, marital status, and the like of the user. The health information includes health status, family history, disease history, life style, physical examination information, health care style, living environment, mental state, health general knowledge, safety awareness and other aspects, and the health status includes information on whether the user has physical defects, whether congenital diseases exist and whether myopia exists. The family history comprises a family medical history of the user; the disease history comprises information of previous diseases of the user; the life style comprises life information such as smoking condition, drinking condition, eating habit, exercise habit, sleeping habit and the like of the user. The physical examination information includes physical examination information of the user, for example: heart rate, liver function, blood lipid, urinary function, renal function, tumor markers, etc. The health care mode comprises information such as vaccination condition, physical examination frequency and the like. The living environment comprises information such as drinking water condition of a user, harmful substance exposure condition in work or life and the like. The mental states include life and work stress situations of the user. The health knowledge includes knowledge of the user about common sense information in terms of disease prevention, health management, and the like. The safety awareness includes the safety awareness of the user in work and life, such as whether fatigue driving is likely, whether a seat belt is worn during driving, whether a smoke sensor is installed at home, and the like. The insurance information comprises the types and the quantity of purchased insurance, and the insurance types are classified into regular life insurance, regular survival, life-long life insurance, accidental injury, accidental medical treatment, hospitalization subsidy and universal insurance.
Specifically, according to information of multiple dimensions such as user basic information, health information and insurance information, a plurality of user attribute feature data are extracted from the information by using a preset regular expression; generating a plurality of user characteristic labels according to the user attribute characteristic data; and generating a user portrait according to the user characteristic label to generate the user portrait. In the embodiment, the final user portrait is generated and displayed according to the user attribute feature data, and the user can intuitively learn various features of the user according to the user portrait.
In another embodiment, further comprising: and evaluating the relationship between the health information and the insurance information of the user portrait by using a risk association model, and calculating a weighting factor of the risk of the user.
Specifically, the risk association model is a disease risk association model, and is used for searching for related diseases from the disease risk association model according to the health information of the user. The related diseases are diseases which are determined according to the profile sub-information of the user and are suffered by the user or possibly suffered by the user in the future, and the number of the related diseases is one or more.
For example, the prevalence probability of the associated disease is calculated according to a disease risk correlation model. When the number of the related diseases is multiple, the disease probability of each related disease is calculated respectively.
For another example, calculating the weight value of the risk factor of the user according to the disease risk association model; and calculating the disease probability of the related diseases according to the weight value of the risk factor of the user and the disease risk association model, thereby obtaining the weighting factor of the disease risk of the user, for example, calculating the average value of the probabilities as the weighting factor if there is a variety of diseases.
In other embodiments, further comprising: integrating the insurance product-attribute table to obtain an insurance product-attribute matrix; calculating a first similarity of the insurance products according to the insurance product-attribute matrix;
quantifying user data to obtain a user-insurance product rating table, and integrating data in the user-insurance product rating table to obtain a user-insurance product rating matrix; and obtaining a rating vector of the insurance product based on the user-insurance product rating matrix, and calculating the user-insurance product rating matrix from the user perspective by adopting cosine similarity to obtain a second similarity of the insurance product.
For example, calculating the similarity (denoted as sima) of insurance products based on the collaborative filtering idea of the articles; quantifying the user data to obtain a user-insurance product rating table, and integrating the user-insurance product rating table to obtainUser-insurance product scoring matrix, user representing user number, item representing insurance number, rating representing score, rijRepresenting the rating of an insurance product with insurance product number j by a user with user number i, e.g. a rating vector r11, r12, r13, r14,…, rm1]. And solving the similarity of the scoring vectors of all insurance products to obtain a similarity matrix of the insurance products.
Calculating by adopting the modified cosine similarity, and calculating a second similarity of the insurance product from the perspective of a user;
Ri,cscoring insurance product i for user c, Rj,cFor the user c to score insurance product j,
and
the average score of all users on insurance products i, j is given;
for example, similarity (marked as simb) of insurance products is obtained according to a recommendation algorithm idea based on the contents of the articles, and an insurance product-attribute table is integrated to obtain an insurance product-attribute matrix; item represents an insurance number, Property represents the attribute of an insurance product, 1 represents that insurance has the attribute, and 0 represents that the insurance product does not have the attribute, and the first similarity of the insurance product is calculated according to the insurance product-attribute matrix; calculating a similarity matrix for the insurance product by:
wherein the content of the first and second substances,
、
calculating the similarity of the insurance products from two angles by adopting the following mode for the vector corresponding to the insurance product number i, j on the basis of the embodiment, and then carrying out weighted summation on the two similarities to obtain the comprehensive similarity;
simx (x, j) is the integrated similarity, β is the weighting factor for the risk of user portrayal, sima (x, j) is the second similarity for calculating insurance products from the user perspective, simb (x, j) is the similarity between insurance products.
In this embodiment, the insurance recommendation algorithm may be a collaborative filtering algorithm based on articles, recommendation is implemented by analyzing the insurance data of the user, and excellent and novel products are recommended to the client by integrating user selection because the client insurance conditions are not completely the same. The recommendation algorithm based on the content utilizes the characteristic attributes of the articles, can effectively relieve the problem of the data set, and can generate richer vectors according to the attributes through the attributes of the insurance products even if the user information is less, so that the recommendation algorithm can recommend the customers. The collaborative filtering algorithm based on the insurance products and the recommendation algorithm based on the insurance product content are combined for use, and then the weighting factors are combined for weighted recommendation, so that the requirement on the novelty of user recommendation is met, and the recommendation accuracy is improved.
In another embodiment, further comprising: and matching the most similar K neighbors of the target insurance product to the user by using the comprehensive similarity to form a nearest neighbor set, calculating the prediction score of the user on the insurance product according to the score vector of the user and the similarity vector of the insurance product, and obtaining a recommendation scheme of the insurance product from high to low according to the prediction score. For example, the prediction score is calculated using the following formula:
In the formula (4), Pu,IIn order to predict the score(s),
the mean value of the user I's score for insurance products, sim (I, n) is the similarity of user u and insurance products in the nearest neighbor set, Ru,nThe target user's rating of insurance product n,
the mean value of the user n's scores for the insurance products.
In this embodiment, the nearest neighbor set is formed by matching the most similar K neighbors in the insurance products according to the prediction scores, the nearest neighbor set is selected from the highest nearest neighbor set according to the prediction scores, and the insurance product with the highest score is selected to form the recommendation scheme.
In other embodiments, further comprising: and acquiring a positive feedback factor and a negative feedback factor of the trust of each pair of users, correcting the second similarity between the original users according to the feedback factors, and performing linear weighting on the optimized trust and the second similarity of the users to obtain the optimized second similarity.
Among the formulas (5), (6) and (7), Trust*(i, j) is the confidence after optimization, f is a negative feedback factor, t is a positive feedback factor, Trust (i, j) is the confidence before optimization, alpha is a weighting coefficient, the value range is 0 to 1, beta is a weighting factor of the risk of the user portrait, and sima*(x, j) is the similarity of user i and user j after optimizationAnd the degree, sima (x, j) is the similarity between the user i and the user j, and simx (x, j) is the comprehensive similarity.
For example, the similarity between users is calculated according to a user-insurance product scoring matrix, the insurance product scoring matrix calculates the size relationship between the common product score and the actual threshold value between each pair of users, so as to obtain the quantity of positive feedback factors and negative feedback factors of each pair of user trust, the original user trust is corrected by using feedback, and finally the optimized trust and the user similarity are subjected to linear weighting, wherein what needs to be described is that the user similarity is the second similarity calculated from the user perspective; and finally, a recommendation scheme of the insurance product is obtained according to the collaborative filtering algorithm, so that the accurate recommendation capability of the collaborative filtering algorithm is effectively improved.
Specifically, the optimized trust level is shown in formula (6), wherein the user trust level is positively correlated with the number of positive feedback factors, and the number of positive feedback factors represents the number of times that the score difference of each pair of users is small; the user trust degree and the number of negative feedback factors are in negative correlation, and the negative feedback factor number represents the times of large scoring difference of each pair of users; and the initial Trust level Trust (i, j) is the Trust level of the user i to the user j, and the optimized user Trust level is obtained through the method.
On the basis of the implementation, the pearson correlation coefficient and the confidence level are linearly combined to obtain the similarity between the optimized user i and the user j, which is specifically shown in formula (5), so that the accuracy of the similarity between the users is further improved.
And as shown in formula (7), on the basis of the user similarity, the first similarity of the insurance product is calculated by combining an insurance product-attribute matrix, the similarity is calculated from the perspective of the insurance product and the perspective of the user, and the recommendation algorithm based on the article content and the collaborative filtering recommendation algorithm based on the user are integrally combined, so that the problem of poor accuracy caused by serious data sparsity and inaccurate calculation of the user similarity of the traditional collaborative filtering algorithm is solved.
Referring to fig. 2, a structural block diagram of an insurance recommendation system based on user information according to the present invention is shown, in which the insurance recommendation system based on user information is detailed as follows:
the system comprises an acquisition module 1, a display module and a display module, wherein the acquisition module 1 is used for acquiring a face image of a user and identifying the face image to obtain identity information of the user;
the portrait construction module 2 is used for acquiring personal information of the user according to the identity information and constructing a portrait of the user by utilizing the personal information;
a risk evaluation module 3 for evaluating the user profile to generate a weighting factor of the user with respect to risk;
the first calculation module 4 is used for generating an insurance product-attribute matrix and a user-insurance product scoring matrix by utilizing a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of the insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
the second calculation module 5 is used for performing weighted calculation on the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
the insurance recommending module 6 is used for matching the most similar K neighbors for the user by utilizing the comprehensive similarity to obtain a recommending scheme of insurance products
It should be further noted that the insurance recommendation method based on the user information and the insurance recommendation system based on the user information are in a one-to-one correspondence relationship, and here, technical details and technical effects related to the insurance recommendation system based on the user information are the same as those of the identification method, and are not repeated here, please refer to the insurance recommendation method based on the user information.
Referring now to FIG. 3, a schematic diagram of a user information based insurance recommendation device, such as an electronic device or server, suitable for use in implementing embodiments of the present disclosure is shown. The electronic device in the embodiments of the present disclosure may include, but is not limited to, a holder such as a mobile phone, a tablet computer, a laptop computer, a desktop computer, a all-in-one computer, a server, a workstation, a television, a set-top box, smart glasses, a smart watch, a digital camera, an MP4 player, an MP5 player, a learning machine, a point-and-read machine, an electronic paper book, an electronic dictionary, a vehicle-mounted terminal, a Virtual Reality (VR) player, an Augmented Reality (AR) player, or the like. The electronic device shown in fig. 3 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 3, the electronic device 7 may include a processing means (e.g., a central processing unit, a graphics processor, etc.) 701 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)702 or a program loaded from a storage means 708 into a Random Access Memory (RAM) 703. In the RAM703, various programs and data necessary for the operation of the electronic apparatus 700 are also stored. The processing device 701, the ROM702, and the RAM703 are connected to each other by a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
Generally, the following devices may be connected to the I/O interface 705: input devices 706 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; an output device 707 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 708 including, for example, magnetic tape, hard disk, etc.; and a communication device 709. The communication means 709 may allow the electronic device 7 to communicate wirelessly or by wire with other devices to exchange data. While fig. 3 shows the electronic device 7 having various means, it is to be understood that not all of the means shown are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via the communication means 709, or may be installed from the storage means 708, or may be installed from the ROM 702. The computer program, when executed by the processing device 701, performs the above-described functions defined in the methods of the embodiments of the present disclosure.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such embodiments, the computer program may be downloaded and installed from a network via the communication means 709, or may be installed from the storage means 708, or may be installed from the ROM 702. The computer program, when executed by the processing device 701, performs the above-described functions defined in the methods of the embodiments of the present disclosure.
The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.
The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: the method of the above-described steps S1 to S6 is performed.
Computer program code for carrying out operations for aspects of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
In conclusion, the collaborative filtering algorithm based on the insurance product and the recommendation algorithm based on the insurance product content are combined for use, and then the weighting factor is combined for weighted recommendation, so that the recommendation novelty requirement of the user is met, the recommendation accuracy is improved, meanwhile, the weighting factor of the user portrait risk is introduced, the real requirements of the user and the product attributes are associated, the recommendation accuracy of the insurance product is further improved on the original basis, various defects in the prior art are effectively overcome, and the method has high industrial utilization value.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (10)
1. An insurance recommendation method based on user information is characterized by comprising the following steps:
acquiring a face image of a user, and identifying the face image to obtain identity information of the user;
acquiring personal information of a user according to the identity information, and constructing a user portrait by using the personal information;
evaluating the user representation to generate a weighting factor for the user with respect to risk;
generating an insurance product-attribute matrix and a user-insurance product scoring matrix by utilizing a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
carrying out weighting calculation on the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
and matching the most similar K neighbors for the user by using the comprehensive similarity to obtain a recommendation scheme of the insurance product.
2. The insurance recommendation method based on user information according to claim 1, further comprising:
acquiring a face image of a user; performing feature extraction on the face image based on all sequentially connected residual blocks in a depth residual error network to obtain face feature information, wherein any one residual block comprises an identity mapping and at least two convolution layers, and the identity mapping of any one residual block points to the output end of any one residual block from the input end of any one residual block; and performing face retrieval in a face database based on the face feature information to obtain a face retrieval result, wherein the face database stores the corresponding relation between the face feature information and the identity information, and the face retrieval result at least comprises the identity information matched with the target face feature information.
3. The insurance recommendation method based on user information according to claim 1 or 2, further comprising: acquiring personal information of the user according to the identity information, wherein the personal information comprises basic information, health information and insurance information of the user; and constructing a user portrait from multiple dimensions of basic information, health information and insurance information according to the personal information.
4. The user information-based insurance recommendation method according to claim 3, further comprising: and evaluating the user portrait by using a risk association model, and calculating a weighting factor of the risk of the user based on the relationship between the health information and the insurance information.
5. The insurance recommendation method based on user information according to claim 1, further comprising:
integrating the insurance product-attribute table to obtain an insurance product-attribute matrix; calculating a first similarity of the insurance products according to the insurance product-attribute matrix;
quantifying user data to obtain a user-insurance product rating table, and integrating data in the user-insurance product rating table to obtain a user-insurance product rating matrix; and obtaining a rating vector of the insurance product based on the user-insurance product rating matrix, and calculating the user-insurance product rating matrix from the user perspective by adopting cosine similarity to obtain a second similarity of the insurance product.
6. The insurance recommendation method based on user information according to claim 1, further comprising:
and matching the most similar K neighbors of the target insurance product to the user by using the comprehensive similarity to form a nearest neighbor set, calculating the prediction score of the user on the insurance product according to the score vector of the user and the similarity vector of the insurance product, and obtaining a recommendation scheme of the insurance product from high to low according to the prediction score.
7. The insurance recommendation method based on user information according to claim 1, further comprising: and acquiring a positive feedback factor and a negative feedback factor of the trust between the users, correcting the original second similarity between the users according to the feedback factor, and performing linear weighting on the optimized trust and the second similarity of the users to obtain the optimized second similarity.
8. An insurance recommendation system based on user information, comprising:
the acquisition module is used for acquiring a face image of a user and identifying the face image to obtain identity information of the user;
the portrait construction module is used for acquiring personal information of the user according to the identity information and constructing a portrait of the user by utilizing the personal information;
a risk assessment module for assessing the user profile to generate a weighting factor for the user with respect to risk;
the first calculation module is used for generating an insurance product-attribute matrix and a user-insurance product scoring matrix by utilizing a collaborative filtering algorithm, calculating the insurance product-attribute matrix to obtain a first similarity of insurance products, and calculating the user-insurance product scoring matrix to obtain a second similarity between users;
the second calculation module is used for performing weighted calculation on the first similarity and the second similarity by using a weighting factor to obtain a comprehensive similarity;
and the insurance recommending module is used for matching the most similar K neighbors for the user by utilizing the comprehensive similarity to obtain a recommending scheme of the insurance product.
9. An insurance recommendation apparatus based on user information, comprising:
one or more processing devices;
a memory for storing one or more programs; when executed by the one or more processing devices, cause the one or more processing devices to implement the user information-based insurance recommendation method of any of claims 1 to 7.
10. A computer-readable storage medium having stored thereon a computer program for causing a computer to execute the user information-based insurance recommendation method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110740664.1A CN113191911A (en) | 2021-07-01 | 2021-07-01 | Insurance recommendation method, system, equipment and medium based on user information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110740664.1A CN113191911A (en) | 2021-07-01 | 2021-07-01 | Insurance recommendation method, system, equipment and medium based on user information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113191911A true CN113191911A (en) | 2021-07-30 |
Family
ID=76976841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110740664.1A Pending CN113191911A (en) | 2021-07-01 | 2021-07-01 | Insurance recommendation method, system, equipment and medium based on user information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113191911A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643103A (en) * | 2021-08-31 | 2021-11-12 | 平安医疗健康管理股份有限公司 | Product recommendation method, device, equipment and storage medium based on user similarity |
| CN113888203A (en) * | 2021-09-04 | 2022-01-04 | 北京优全智汇信息技术有限公司 | Online insurance product sale system and sale method |
| CN114997956A (en) * | 2022-06-14 | 2022-09-02 | 杭州洋驼网络科技有限公司 | Mother and infant product intelligent recommendation system based on big data |
| CN118014738A (en) * | 2024-01-10 | 2024-05-10 | 保腾网络科技有限公司 | Life insurance recommending method and system based on artificial intelligence |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463630A (en) * | 2014-12-11 | 2015-03-25 | 新一站保险代理有限公司 | Product recommendation method and system based on characteristics of online shopping insurance products |
| CN108108499A (en) * | 2018-02-07 | 2018-06-01 | 腾讯科技（深圳）有限公司 | Face retrieval method, apparatus, storage medium and equipment |
| CN112992317A (en) * | 2021-05-10 | 2021-06-18 | 明品云(北京)数据科技有限公司 | Medical data processing method, system, equipment and medium |
| CN112989210A (en) * | 2021-05-17 | 2021-06-18 | 明品云(北京)数据科技有限公司 | Insurance recommendation method, system, equipment and medium based on health portrait |
-
2021
- 2021-07-01 CN CN202110740664.1A patent/CN113191911A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104463630A (en) * | 2014-12-11 | 2015-03-25 | 新一站保险代理有限公司 | Product recommendation method and system based on characteristics of online shopping insurance products |
| CN108108499A (en) * | 2018-02-07 | 2018-06-01 | 腾讯科技（深圳）有限公司 | Face retrieval method, apparatus, storage medium and equipment |
| CN112992317A (en) * | 2021-05-10 | 2021-06-18 | 明品云(北京)数据科技有限公司 | Medical data processing method, system, equipment and medium |
| CN112989210A (en) * | 2021-05-17 | 2021-06-18 | 明品云(北京)数据科技有限公司 | Insurance recommendation method, system, equipment and medium based on health portrait |
Non-Patent Citations (2)
| Title |
|---|
| 管逸飞: "保险推荐***算法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 黄伟建: "结合隐语义模型与改进用户信任的协同过滤推荐算法", 《现代电子技术》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643103A (en) * | 2021-08-31 | 2021-11-12 | 平安医疗健康管理股份有限公司 | Product recommendation method, device, equipment and storage medium based on user similarity |
| CN113643103B (en) * | 2021-08-31 | 2024-06-25 | 深圳平安医疗健康科技服务有限公司 | Product recommendation method, device, equipment and storage medium based on user similarity |
| CN113888203A (en) * | 2021-09-04 | 2022-01-04 | 北京优全智汇信息技术有限公司 | Online insurance product sale system and sale method |
| CN113888203B (en) * | 2021-09-04 | 2022-09-13 | 北京优全智汇信息技术有限公司 | Online insurance product sale system and sale method |
| CN114997956A (en) * | 2022-06-14 | 2022-09-02 | 杭州洋驼网络科技有限公司 | Mother and infant product intelligent recommendation system based on big data |
| CN114997956B (en) * | 2022-06-14 | 2023-01-10 | 杭州洋驼网络科技有限公司 | Mother and infant product intelligent recommendation system based on big data |
| CN118014738A (en) * | 2024-01-10 | 2024-05-10 | 保腾网络科技有限公司 | Life insurance recommending method and system based on artificial intelligence |
| CN118014738B (en) * | 2024-01-10 | 2024-07-12 | 保腾网络科技有限公司 | Life insurance recommending method and system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113191911A (en) | Insurance recommendation method, system, equipment and medium based on user information | |
| US11670415B2 (en) | Data driven analysis, modeling, and semi-supervised machine learning for qualitative and quantitative determinations | |
| CN111784455B (en) | Article recommendation method and recommendation equipment | |
| Gelman | The connection between varying treatment effects and the crisis of unreplicable research: A Bayesian perspective | |
| US20140114998A1 (en) | Determining demographics based on user interaction | |
| US11875277B2 (en) | Learning and applying contextual similiarities between entities | |
| US20140195255A1 (en) | System And Method For Assessment Of Patient Health Using Patient Generated Data | |
| WO2020238502A1 (en) | Article recommendation method and apparatus, electronic device and storage medium | |
| CN112000894A (en) | Health management information pushing method, device, equipment and storage medium | |
| KR20170067796A (en) | One way and two way data flow systems and methods | |
| CN113192639B (en) | Training method, device, equipment and storage medium of information prediction model | |
| US11650981B1 (en) | Method and system for reducing scan-time for single user query | |
| Goldhaber-Fiebert et al. | Some health states are better than others: using health state rank order to improve probabilistic analyses | |
| Lee et al. | Differences in the perception of blood transfusion risk between laypeople and physicians | |
| CN113094476A (en) | Risk early warning method, system, equipment and medium based on natural language processing | |
| Alzahrani et al. | [Retracted] Towards Understanding the Usability Attributes of AI‐Enabled eHealth Mobile Applications | |
| Baktha et al. | Social network analysis in healthcare | |
| US20220368986A1 (en) | Methods and systems for counseling a user with respect to supervised content | |
| Lin et al. | Design and implementation of a set-top box–based homecare system using hybrid cloud | |
| CN115858876A (en) | Disease knowledge graph-based follow-up content intelligent pushing method and system | |
| US20230075408A1 (en) | Methods and systems for interactively counseling a user with respect to supervised content | |
| Antolín-Prieto et al. | KPIs for mobile apps and digital data management in healthcare | |
| TWM633789U (en) | Matching system | |
| CN113688314B (en) | Physical therapy store recommending method and device | |
| CN110334283A (en) | Information recommendation method, device, server and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210730 |