CN112561709A

CN112561709A - Product information method, device, equipment and medium

Info

Publication number: CN112561709A
Application number: CN202011372435.0A
Authority: CN
Inventors: 王超
Original assignee: Taikang Health Industry Investment Holdings Co ltd; Taikang Insurance Group Co Ltd
Current assignee: Taikang Health Industry Investment Holdings Co ltd; Taikang Insurance Group Co Ltd
Priority date: 2020-11-30
Filing date: 2020-11-30
Publication date: 2021-03-26
Anticipated expiration: 2040-11-30
Also published as: CN112561709B

Abstract

The invention discloses a product information method, a product information device, product information equipment and a product information medium, which are used for solving the problem that products recommended to an account cannot be conveniently and automatically determined in the prior art. In the process of recommending the product information, first service experience information of the target account can be acquired, wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, preference information of the target account, experience time information of the target account and experience evaluation information of the target account. After the first service experience information of the target account is acquired, the first service experience information is fully utilized through a pre-trained prediction model, a first target product corresponding to the first service experience information is determined, then the product information of the first target product is recommended to the target account, and the product is recommended to the target account more conveniently and quickly.

Description

Product information method, device, equipment and medium

Technical Field

The invention relates to the technical field of insurance recommendation, in particular to a product information method, device, equipment and medium.

Background

With the development of society and the change of environment, the health industry is also constantly changing the related policies and development directions. The strategy of 'healthy China 2020' and the planning of 'healthy China 2030' make specific arrangements in the future from different angles and directions. Enhancing the operability and guidance of the national health plan, promoting the deep integration of biomedicine with the national health, the national fitness with the national health, the health maintenance travel with the national health and the like are important subjects needing to be continuously and deeply researched in the practical process of building the 'healthy China'.

The deep integration of health maintenance travel and the health of all people needs deep participation of enterprises in the industry of endowment, medical treatment and insurance, and as an enterprise covering the business of endowment, medical treatment and insurance, on the basis of the national major health strategy, the product service which is provided for common people in the long-life era is provided, and the service is provided for accounts in an online mode. However, in the online system, introduction and guidance of a salesperson to a product are lacked, so that a user to which the account belongs does not know what kind of product is selected, and which kind of product is more matched with the user's own condition or the user's relatives. In such a case, it is difficult for the user to obtain a good use experience. Therefore, how to conveniently and automatically determine to recommend products to an account is a problem to be solved in order to better guarantee the user satisfaction and recognition degree of the service quality by always taking humanized service as a starting point.

Disclosure of Invention

The embodiment of the invention provides a product information method, a product information device, product information equipment and a product information medium, which are used for solving the problem that products recommended to an account cannot be conveniently and automatically determined in the prior art.

The embodiment of the invention provides a product information method, which comprises the following steps:

acquiring first service experience information of a target account;

acquiring a first target product corresponding to the first service experience information through a pre-trained prediction model; wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, hobby information of the target account, experience time information of the target account and experience evaluation information of the target account; the prediction model is obtained by training an original prediction model based on second service experience information, and the second service experience information is obtained by preprocessing sample service experience information in a sample set;

determining product information of the first target product, and recommending the product information of the first target product to the target account;

training an original prediction model based on the second service experience information, comprising:

acquiring second service experience information of any sample account and a corresponding label thereof, wherein the label is used for identifying products purchased by the sample account;

acquiring a second target product corresponding to the second service experience information through the original prediction model;

and training the original prediction model according to the label and the second target product.

Further, the preprocessing the sample service experience information in the sample set includes:

if the sample service experience information comprises age information of the sample account, determining a pre-processed age value of the sample account based on the age information of the sample account, an average value of the age information of each account purchasing the product, and a mean square error of the age information of each account purchasing the product;

and if the sample service experience information comprises the preference information of the sample account, determining the preprocessed preference value of the sample account for the preference based on the survey score of the sample account for the preference and the preset association degree of the preference and the product aiming at each preference.

Further, the preprocessing the sample service experience information in the sample set further includes:

if the sample service experience information comprises the experience time information of the sample account, determining the reserved times of each hour and the reserved hours of each time of the sample account in the service time period according to a preset service time period and each experience time information of the sample account; determining a pre-processed reservation time value of the sample account according to the reservation times and the hours;

and if the sample service experience information comprises the experience evaluation information of the sample account, determining a target experience evaluation information value corresponding to the experience evaluation information of the sample account according to a preset corresponding relationship between the experience evaluation information and the experience evaluation information value.

Further, the experience evaluation information of the target account further includes: and evaluating information corresponding to the behavior of the client in the actual experience process, such as the interest degree of the experience item, the purchase intention and the like.

Further, if the sample service experience information includes age information of the sample account, determining a pre-processed age value of the sample account based on the age information of the sample account, an average of the age information of each account purchasing the product, and a mean square error of the age information of each account purchasing the product includes:

wherein A is_iAn ith product purchased for the sample account; c₁(A_i) Is a pre-processed age value of the sample account; age represents age information of the sample account; c. C_ageAn average value representing age information of each account from which the ith product was purchased; sigma_ageA mean square error representing age information for each account purchasing the ith product;

if the sample service experience information includes preference information of the sample account, determining a preprocessed preference value of the sample account for the preference based on the survey score of the sample account for the preference and a pre-configured degree of association between the preference and the product, including:

C_3j(A_i)＝λ_jG_j，

wherein A is_iAn ith product purchased for the sample account; c_3j(A_i) The preprocessed preference value of the sample account for the j-th preference is obtained; lambda [ alpha ]_jA survey score representing a class j taste for the sample account; g_jRepresenting the relevance of the preset jth hobby and ith product;

if the sample service experience information includes the experience time information of the sample account, determining the pre-processed appointment time value of the sample account according to the appointment times and the hours, including:

wherein A is_iAn ith product purchased for the sample account; c₄(A_i) Is the pre-processed appointment time value for the sample account; m is the reserved times; h is the number of hours.

Further, before the obtaining of the first target product corresponding to the first service experience information through the pre-trained prediction model, the method further includes:

determining a target type for each of the first service experience information;

if the first service experience information is determined to lack the information of the necessary information type according to the pre-configured necessary information type and the target type, determining the historical information of the lacking necessary information type from the pre-stored historical information of each necessary information type of the target account, and adding the historical information of the lacking necessary information type to the first service experience information.

determining a target type for each of the first service experience information;

and filtering the information of which the target type is not a preset necessary information type in the first service experience information.

Further, the recommending the product information of the first target product to the target account includes:

and recommending the product information of the first target product to the target account if the confirmation information of the first target product is received.

An embodiment of the present invention provides a product information apparatus, where the apparatus includes:

the acquisition unit is used for acquiring first service experience information of the target account;

the processing unit is used for acquiring a first target product corresponding to the first service experience information through a pre-trained prediction model; wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, hobby information of the target account, experience time information of the target account and experience evaluation information of the target account; the prediction model is obtained by training an original prediction model based on second service experience information, and the second service experience information is obtained by preprocessing sample service experience information in a sample set;

the recommending unit is used for determining the product information of the first target product and recommending the product information of the first target product to the target account;

wherein the apparatus further comprises: a training unit;

the training unit is used for acquiring second service experience information of any sample account and a corresponding label thereof, wherein the label is used for identifying a product purchased by the sample account; acquiring a second target product corresponding to the second service experience information through the original prediction model; and training the original prediction model according to the label and the second target product.

Further, the apparatus further comprises: a pre-processing unit;

the preprocessing unit is configured to determine a preprocessed age value of the sample account based on the age information of the sample account, an average value of the age information of each account purchasing the product, and a mean square error of the age information of each account purchasing the product if the sample service experience information includes the age information of the sample account; and if the sample service experience information comprises the preference information of the sample account, determining the preprocessed preference value of the sample account for the preference based on the survey score of the sample account for the preference and the preset association degree of the preference and the product aiming at each preference.

Further, the preprocessing unit is further configured to determine, if the sample service experience information includes experience time information of the sample account, a number of reservations per hour and a number of hours of each reservation of the sample account within the service time period according to a preconfigured service time period and each experience time information of the sample account; determining a pre-processed reservation time value of the sample account according to the reservation times and the hours; and if the sample service experience information comprises the experience evaluation information of the sample account, determining a target experience evaluation information value corresponding to the experience evaluation information of the sample account according to a preset corresponding relationship between the experience evaluation information and the experience evaluation information value.

Further, the preprocessing unit is specifically configured to base the sample service experience information on age information of the sample account if the sample service experience information includes the age information of the sample account

Determining a pre-processed age value for the sample account; wherein A is_iAn ith product purchased for the sample account; c₁(A_i) Is a pre-processed age value of the sample account; age represents age information of the sample account; c. C_ageTo representAn average of age information for each account for purchasing the ith product; sigma_ageA mean square error representing age information for each account purchasing the ith product;

if the sample service experience information comprises preference information of the sample account, based on C_3j(A_i)＝λ_jG_jDetermining the preprocessed preference value of the sample account for the love; wherein A is_iAn ith product purchased for the sample account; c_3j(A_i) The preprocessed preference value of the sample account for the j-th preference is obtained; lambda [ alpha ]_jA survey score representing a class j taste for the sample account; g_jRepresenting the relevance of the preset jth hobby and ith product;

if the sample service experience information comprises the experience time information of the sample account, based on

Further, the apparatus further comprises:

the determining unit is used for determining a target type of each piece of information in the first service experience information before a first target product corresponding to the first service experience information is obtained through a pre-trained prediction model;

the processing unit is further configured to determine, if it is determined that the first service experience information lacks information of a necessary information type according to a pre-configured necessary information type and the target type, history information of the missing necessary information type from history information of each necessary information type of the target account stored in advance and add the history information to the first service experience information.

Further, the apparatus further comprises:

the processing unit is further configured to filter information, in the first service experience information, of which a target type is not a pre-configured necessary information type.

Further, the recommending unit is specifically configured to recommend the product information of the first target product to the target account if the confirmation information of the first target product is received.

An embodiment of the present invention provides an electronic device, where the electronic device at least includes a processor and a memory, and the processor is configured to implement the steps of the product information method according to any one of the above descriptions when executing a computer program stored in the memory.

An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program, and the computer program, when executed by a processor, implements the steps of any of the product information methods described above.

In the process of recommending the product information, first service experience information of the target account can be acquired, wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, preference information of the target account, experience time information of the target account and experience evaluation information of the target account. After the first service experience information of the target account is obtained, the first service experience information is fully utilized through a pre-trained prediction model, a first target product corresponding to the first service experience information is determined, then the product information of the first target product is recommended to the target account, products can be recommended to the target account more conveniently and rapidly, the recommended products are determined based on the service experience information of the target account, the method is more suitable for users to a certain extent, and the matching efficiency of products and user requirements and the user experience are improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic diagram of a process for determining product information according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a training process of a prediction model according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a specific process for determining product information according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of an apparatus for determining product information according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the attached drawings, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In order to conveniently and automatically determine products recommended to a target account, the embodiment of the invention provides a product information method, a product information device, product information equipment and a product information medium.

Aiming at the vigorous development of the current medical, endowment and insurance industries, products are recommended to an account through some blind and inappropriate operation means, and the trust degree and the user experience of a user to which the account belongs to an insurance service organization can be reduced. Therefore, in order to support the comprehensive reservation service of the super-experience type service product, the super-experience type product service is enriched, the flexible service provision is facilitated, the user can make a reservation in time, the management capability and the user experience of the super-experience type service are comprehensively improved, and a product recommendation platform for the super-experience type reservation is built, so that the user can reserve various services in the product recommendation system, for example, at least one service such as visiting a scenic spot, visiting an old home, visiting an oral hospital, reserving ophthalmology, reserving physical examination, serious disease registration, community visiting, commemorative garden experience and the like is provided for the user, all-round and multi-level reservation experience is provided for the user, and the enterprise product which is true to mind and is intended for the user can be identified in the reservation experience.

In the specific application process, the product recommendation system is divided into an A end, a B end and a C end, wherein the A end is mainly used by an enterprise operation management department and is responsible for product management, channel management, appointment rule setting, appointment auditing, management report forms and other services, the B end is mainly used by a first-line worker providing services and is responsible for field resource maintenance, resource scheduling management, appointment order confirmation, appointment visit reception and other services, and the C end is mainly used by a user, so that the user can make an appointment at the C end, enter a garden and verify, experience evaluation information, product purchase and other operations. In the product recommendation system, the order of the reserved service is taken as a core to realize the super-experience service productization, for example, a user can reserve the service on a product recommendation platform like purchasing a product, the order is generated and visualized according to the reservation information filled by the user when the user reserves the service, for example, the user and related staff can see the order generated when each user reserves the service, and the related staff can realize the management of the super-experience service at multiple entrances through the modes of resource management, rule management, reservation management, payment management, information management and the like of the order.

Example 1:

fig. 1 is a schematic diagram of a process for determining product information according to an embodiment of the present invention, where the process includes:

s101: first service experience information of a target account is obtained.

S102: acquiring a first target product corresponding to the first service experience information through a pre-trained prediction model; wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, hobby information of the target account, experience time information of the target account and experience evaluation information of the target account; the prediction model is obtained by training an original prediction model based on second service experience information, and the second service experience information is obtained by preprocessing sample service experience information in a sample set.

The product information method provided by the embodiment of the invention is applied to electronic equipment, and the electronic equipment can be intelligent equipment such as an intelligent terminal, an intelligent computer, a tablet and the like, and can also be a server.

In a practical application scenario, when a user wants to purchase a certain product but hesitates to purchase a specific product that needs to be purchased, the reservation service function in the product recommendation system can be used through an account registered in the product recommendation system, i.e., a target account, so as to select a service that the user wants to reserve currently and input reservation information, such as the type of the reserved service, gender information, age information, hobby information, and the like. And determining first service experience information of the target account based on the acquired reservation information of the target account and experience information generated by the target account after the target account experiences the reserved service, such as experience evaluation of the target account on the reserved service. After the electronic device for determining the product information acquires the first service experience information of the target account, corresponding processing is performed based on the reservation request, so that the product information of the first target product recommended to the target account is determined.

The reservation information may be input by inputting the reservation information in a corresponding text input box displayed on the display interface, by sliding a sliding window displayed on the display interface to select corresponding information as the input reservation information, or by inputting voice information. The specific ways of inputting the reservation information are many, and the reservation information can be flexibly set according to the requirements in the specific implementation process, which is not specifically limited herein.

In a specific implementation process, in order to conveniently determine the products recommended to the target account, a prediction model for predicting the recommended products is trained in advance. After the electronic device for product recommendation obtains first service experience information of a target account, a first target product corresponding to the first service experience information is obtained through a pre-trained prediction model based on the first service experience information.

In order to obtain the trained prediction model, a sample set used for training the prediction model is collected in advance, the sample set contains sample service experience information of an account purchased with a product, and the account purchased with the product is determined as a sample account. The sample service information of any sample account is determined according to reservation information input when the sample account reserves services on the product recommendation system and information generated when the reserved services are experienced. And for each sample service experience information in the sample set, preprocessing the sample service experience information, and determining the preprocessed sample service experience information as second service experience information. And training the original prediction model based on each piece of second service experience information, namely adjusting the parameter value of each parameter in the original prediction model to obtain the trained prediction model. And subsequently determining products recommended to any target account based on the trained prediction model.

For example, the target accounts of different ages are relatively old and may be relatively inclined to purchase products of endowment type, the target accounts of different ages are children and may be relatively inclined to purchase products of education fund type, and the target accounts of different genders are also different in inclined purchase products, for example, men are relatively inclined to purchase products of transportation insurance type, women are relatively inclined to purchase products of family financing type, and the preference of the target account has an influence on the products which the target accounts are inclined to purchase, for example, the target accounts which are inclined to travel are inclined to purchase products of travel insurance type, and the target accounts which are inclined to financing are inclined to purchase products of annual insurance type. Therefore, in the embodiment of the present invention, the first service experience information carries at least one of age information of the target account, gender information of the target account, and preference information of the target account.

Wherein the experience evaluation information of the target account further comprises: and evaluating information corresponding to the behavior of the client in the actual experience process, such as the interest degree of the experience item, the purchase intention and the like.

Further, the subscribed services may also have an effect on the products that the target account is inclined to purchase, such as the target account experiencing a certain type of service multiple times, or the target account evaluating a certain subscribed service experience as being very satisfactory, etc. Therefore, on the basis of the above embodiment, the first service experience information may also be at least one of experience time information of the target account and experience evaluation information of the target account, that is, the first service experience information includes at least one of age information of the target account, gender information of the target account, taste information of the target account, experience time information of the target account, and experience evaluation information of the target account.

S102: and determining the product information of the first target product, and recommending the product information of the first target product to the target account.

Based on the method in the embodiment, the obtained first service experience information is processed through a pre-trained prediction model, the probability value recommended to the target account by each product can be obtained, and the first target product corresponding to the service experience information is determined based on each probability value.

In one possible implementation, for each product, it may be determined whether a probability value corresponding to the product is greater than a set probability threshold, and if so, the product is determined as a candidate product. The method includes that a set number of candidate products are randomly selected from each candidate product to serve as a first target product, or a set number of candidate products with a high probability value are selected from each candidate product to serve as the first target product. Of course, the set number of products with larger probability value may be directly used as the first target product, for example, when a certain product corresponds to a probability value

Indicating that the product can be recommended to the target account, determining the product as a first target product; otherwise, it indicates that the product cannot be recommended. Specifically, the mode of determining the first target product may be flexibly set according to actual requirements, and is not specifically limited herein.

In the actual application scene, the general enterprise operation business covers the industries of endowment, medical treatment and insurance, and at present, the enterprise provides a major health strategy of the life-prolonging era, and provides the services of appointment visit physical examination of endowment communities, memorial gardens, oral institutions, physical examination institutions, psychological consultation and green expert medical treatment for users. Therefore, aiming at the appointment visit of the endowment community, the enterprise provides the entity endowment community for the user, the user can make an appointment for the endowment check-in service, so that the real check-in experience is obtained, and the humanistic care in the endowment community is experienced, so that the product corresponding to the appointment service is selected according to the check-in experience, and the product corresponding to the appointment service can have the check-in right of the endowment community and the endowment product; aiming at the appointment visit of the memorial park, the user can make an appointment for the travel service of the memorial park, so that the user can feel the humanistic care provided by the memorial park in a travel mode, enjoy the service mode provided by the memorial park and draw the scenery in the memorial park; aiming at the reserved visit of the oral cavity mechanism, a user can determine whether to purchase products corresponding to the reserved service, such as oral cavity packages and oral cavity products, according to the visit experience by visiting the oral cavity service; aiming at the appointment visit of the physical examination mechanism, a user can personally experience the actual physical examination process, and according to the experience of comparing the appointment physical examination with the conventional physical examination, whether to purchase products corresponding to the appointment service, such as physical examination packages of different age groups and examination types and life insurance products for self health prevention in advance are determined; aiming at the appointed visit of psychological consultation, a user can improve the attention degree of the user to the psychological health by visiting a psychological consultation mechanism, and prevent psychological diseases; aiming at appointment visiting of the green channel medical service, the user can have the service customized by a special person together with the same service through the green channel medical service, the convenient service of the expert resource is reserved in advance, then the medical experience of the green channel medical service is compared with the common medical experience, and whether a product corresponding to the appointment service is purchased or not is determined according to a comparison result. Therefore, in the embodiment of the present invention, the first target product may be a kind of insurance, and may also be a specific insurance product.

If the first target product is of the insurance category, the first target product includes at least one of accident insurance, life insurance, annuity insurance, cancer prevention insurance, commercial medical insurance, travel insurance, property insurance, health insurance, vehicle insurance, traffic insurance, and serious insurance. In the specific implementation process, after a first target product is determined, the product information of each specific product of the first target product configured in advance is recommended to a target account.

If the first target product is a specific insurance product, the first target product includes at least one of an accident insurance type product, a life insurance type product, an annuity insurance type product, a cancer prevention insurance type product, a commercial medical insurance type product, a travel insurance type product, a property insurance type product, a health insurance type product, a vehicle insurance type product, a traffic insurance type product, and a heavy insurance type product. In a specific implementation process, after a first target product is determined, the pre-configured product information of the first target product is directly recommended to a target account.

The product information of the first target product can be recommended in a pop-up window mode in the product recommendation system, the product information of the first target product can also be pushed in a display interface of the product recommendation system, and the product information of the first target product can also be recommended in a short message notification mode. There are many specific recommendation methods, and the method is not limited to the above recommendation method, and is not limited to the specific recommendation method.

S104: training an original prediction model based on the second service experience information, comprising:

In order to conveniently and automatically determine the products recommended to the target account, the original prediction model needs to be trained according to the second service experience information of each sample account. And the second service experience information of any sample account corresponds to a label, and the label is used for identifying the products purchased by the sample account.

It should be noted that the electronic device used for model training may be the same as or different from the electronic device used for determining product information. If the electronic equipment used for model training is different from the electronic equipment used for determining the product information, the electronic equipment used for model training trains the original prediction model in advance according to the second service experience information and the corresponding label thereof, after the trained prediction model is obtained, the trained prediction model is stored in the electronic equipment used for determining the product information, the electronic equipment used for subsequently determining the product information can determine a first target product recommended to a target account through the trained prediction model, and then the product information of the first target product is recommended to the target account.

In the embodiment of the invention, the second service experience information of any sample account is input into the original prediction model, so that a second target product corresponding to the input second service experience information can be obtained, and the original prediction model is trained according to the label corresponding to the second service experience information and the predicted second target product, so as to adjust the parameter values of all parameters in the original prediction model.

Wherein the predictive model may be a Radial Basis (RBF) neural network.

And a plurality of pieces of second service experience information are used for the training of the prediction model, the operation is carried out on each piece of second service experience information, and the training of the prediction model is completed when a preset convergence condition is met.

The condition that the preset convergence condition is satisfied may be that a loss value determined based on a second target product for which each second service experience information is predicted by the trained prediction model and a label corresponding to each second service experience information is smaller than a preset loss threshold, or that the number of iterations for training the original prediction model reaches a set maximum number of iterations, and the like. The specific implementation can be flexibly set, and is not particularly limited herein.

As a possible implementation manner, when the original prediction model is trained, the second service experience information may be divided into a training sample and a testing sample, the original prediction model is trained based on the training sample, and then the reliability of the trained prediction model is verified based on the testing sample.

Example 2:

fig. 2 is a schematic diagram of a training process of a specific prediction model provided in an embodiment of the present invention, and a method for training the prediction model provided in the embodiment of the present invention is described below with reference to fig. 2, with the prediction model being an RBF model:

firstly, sample service experience information used for prediction model training is collected in advance, and a label corresponding to the sample service experience information is determined, wherein the label corresponding to any sample service experience information is used for identifying a product purchased by a sample account corresponding to the sample service experience information. According to the label corresponding to each sample service experience information, determining a probability value, such as 1, corresponding to the label corresponding to each sample service experience information.

In order to facilitate the original prediction model to process the sample service experience information, each sample service experience information can be cleaned to filter out useless or invalid information, then the original prediction model is trained conveniently to obtain a trained prediction model, the sample service experience information is preprocessed, the preprocessed sample service experience information is determined as second service experience information, and then the original prediction model is trained based on the second service experience information to obtain the trained prediction model.

Designing a network structure of an original prediction model, for example, determining a connection mode of the RBF model as p-K-1, that is, p input layer neurons, K hidden layer neurons and 1 output layer neuron of the RBF model, and randomly assigning values to each parameter value in the original prediction model. Wherein the neurons in the hidden layer are small natural numbers.

Updating the current iteration count, namely, the current saved iteration count +1, before the following steps are performed on the second service experience information of each sample account.

Training the original predictive model based on the second service experience information for each sample account. Specifically, for the second service experience information of each sample account, the following steps are executed:

inputting the second service experience information into the original prediction model, for example, the second service experience information of a certain sample account input at the t-th time is x₁(t)，x₂(t)，…，x_p(t), the label corresponding to the second service experience information is y (t) ═ a_i，A_iFor the product purchased by the sample account, through the original prediction model, a second target product corresponding to the currently input second service experience information and a probability value corresponding to the second target product and recommended to the target account can be obtained, and the second target product can be represented as

Specifically, the second target product corresponding to the currently input second service experience information is obtained through the original prediction model, and the second target product corresponding to the currently input second service experience information can be obtained through the following method:

a probability value, ω, corresponding to the second target product corresponding to the second service experience information x (t) input at the t-th time_k(t) represents the connection weight of the kth neuron in the hidden layer and the output layer at the tth moment, wherein K is 1, 2, … and K; theta_k(x (t)) is the output of the k-th neuron of the hidden layer after the second service experience information x (t) is input into the original prediction model.

After the second service experience information x (t) is input into the original prediction model, the output theta of the kth neuron of the hidden layer_kThe formula for the calculation of (x (t)) is:

μ_k(t) denotes the k-th neuron center value, σ, of the hidden layer at time t_kRepresenting the central width of the k-th neuron of the hidden layer at time t.

After a second target product is obtained, determining a loss value according to a probability value corresponding to the second target product and second service experience information x (t) input at the t-th moment, specifically determining by the following method:

wherein E (t) represents the second service experience information x (t) corresponding to the input t timeA loss value y (t) represents a probability value corresponding to the tag corresponding to the second service experience information x (t) input at the time t, and the value is 1,

and the probability value corresponding to the second target product corresponding to the second service experience information x (t) input at the tth moment.

Based on the determined loss value, adjusting each parameter value in the original prediction model, specifically by the following formula:

wherein, W^T(t)＝[w₁(t)，w₂(t)，…，w_K″(t)]^TRepresenting the connection weight of K neurons in the hidden layer and the output layer at the t moment respectively, and representing the t moment; eta ∈ (0, 0.1)]Representing a neural network learning rate; theta_k(x (t)) is the output of the kth neuron of the hidden layer after the second service experience information x (t) is input into the original prediction model;

the probability value corresponding to the second target product corresponding to the second service experience information x (t) input at the t-1 moment, and y (t-1) represents the probability value corresponding to the label corresponding to the second service experience information x (t) input at the t-1 moment.

Through the steps, the original prediction model is trained, whether the currently trained prediction model meets the pre-configured convergence condition is judged, and whether the trained prediction model is obtained is determined. Specifically, if the sum of the loss values corresponding to each piece of second service experience information acquired at the current iteration is smaller than a set loss value threshold, determining to acquire a trained prediction model; otherwise, judging whether the current iteration number 1 reaches the maximum iteration number L, if so, stopping continuous training, and if not, continuously training the prediction model.

When the maximum iteration number is set, the maximum iteration number is generally set to be larger, for example, L ∈ (100, 300), so that the sum of the loss values corresponding to each piece of second service experience information acquired by the current iteration is smaller than the set loss value threshold before the current iteration number reaches the maximum iteration number.

By the method, after the trained prediction model is obtained, the accuracy of the trained prediction model can be tested, and when the trained prediction model is determined to pass the test, the probability value recommended to the target account by each product is determined through the trained prediction model based on the obtained first service experience information. For each product, if the probability value of the product meets the pre-configured recommendation requirement, determining the product as a first target product and recommending the first target product to a target account; otherwise, the product is not recommended to the target account.

Example 3:

for the convenience of the model to process the second service experience information, on the basis of the foregoing embodiments, in an embodiment of the present invention, the preprocessing the sample service experience information in the sample set includes:

In the embodiment of the present invention, since some information that is not convenient for the model to process may exist in the sample service experience information, such as non-numerical experience evaluation information, account gender, hobbies, and the like, the sample service experience information may need to be pre-processed, e.g., regularizing the sample service experience information, deleting punctuation marks in the sample service experience information, unifying capital and small cases in the sample service experience information, converting non-numerical information in the sample service experience information into corresponding numerical values and the like, updating the original sample service experience information by the preprocessed sample service experience information, and based on the updated sample service experience information, and training the original prediction model, thereby facilitating the original prediction model to further process the sample service experience information.

Specifically, for each service experience information that may exist in the sample service experience information, the preprocessing may be performed as follows:

in a first manner, because in an actual application scenario, the sample service experience information may include age information of the sample account, so that the prediction model may learn a potential relationship between the product and the age information of the sample account, in the embodiment of the present invention, if the sample service experience information includes the age information of the sample account, an average value and a mean square deviation of ages of each account purchasing the product in the sample set may be determined for each product. Then, for each sample service experience information in the sample set, according to the age information of the sample account, the average value and the mean square error of the age information of each account of the product in the label corresponding to the purchase sample service experience information, corresponding processing is carried out, the preprocessed age value of the sample account is determined, and second service experience information is determined based on the preprocessed age value of the sample account.

In one possible embodiment, if the sample service experience information includes age information of the sample account, the determining the pre-processed age value of the sample account based on the age information of the sample account, the average value of the age information of each account purchasing the product, and the mean square error of the age information of each account purchasing the product includes:

wherein A is_iAn ith product purchased for the sample account; c₁(A_i) Is a pre-processed age value of the sample account; age represents age information of the sample account; c. C_ageAn average value representing age information of each account from which the ith product was purchased; sigma_ageA mean square error of age information representing each account for purchasing the ith product.

In a second way, in an actual application scenario, the sample service experience information may include gender information of the sample account, and the gender is non-numerical information, which is not favorable for the processing of the original prediction model. In order to enable the prediction model to learn the potential relationship between the product and the gender information of the account, in the embodiment of the present invention, if the sample service experience information includes the gender information of the sample account, a first numerical value corresponding to the gender information of a male and a second numerical value corresponding to the gender information of a female may be preset. For each sample service experience information in the sample set, if the gender information of the target account included in the sample service experience information is male, determining a preset first numerical value as a preprocessed gender value of the sample account; and if the gender information of the sample account included in the sample service experience information is female, determining a preset second numerical value as the preprocessed gender value of the sample account.

For example,

wherein A is_iIth product representing sample account purchase, C₂(A_i) Is the preprocessed sex value corresponding to the sample account purchased the ith product, if the sex information of the sample account is male, the preprocessed sex value of the sample account is determined to be-1, if the sex information of the sample account is female, the sample account is determined to be femaleThe pre-processed gender value of the account is 1.

In a third mode, in an actual application scenario, the sample service experience information may include preference information of the sample account, and the preference information is also non-numerical information, which is not beneficial to processing by the original prediction model. In order to enable the prediction model to learn the potential relationship existing between the product and the preference information of the account, in the embodiment of the invention, the survey score of each sample account for each type of preference is counted in advance to determine the preference degree of the sample account for the type of preference through the survey score, for example, the survey score is lambda_j0, indicating that the sample account has no preference for the j-like preference; lambda [ alpha ]_j0.5, indicating that the sample account has certain preference for the j-type preference; and λ_j1, the preference of the sample account for the j types of hobbies is represented, and the association degree of each type of hobbies and each product is configured in advance, so that the influence of the hobbies on the product purchased by the sample account is determined through the association degree, such as the association degree G of a certain hobbies and a certain product_j1, the association relationship between the hobbies and the product is strong, and the hobbies have a large influence on the purchase of the product in the sample account; if the association degree G of the hobby and the product_j0.75, the association relationship between the hobbies and the product is strong, and the influence of the hobbies on purchasing the product by the sample account is large; if the association degree G of the hobby and the product_j0.5, indicating that the hobby has certain correlation with the product, and the hobby has certain influence on the account to buy the product; if the association degree G of the hobby and the product_j0.25, the association relationship between the hobbies and the product is weak, and the influence of the hobbies on purchasing the product by the sample account is small; if the association degree G of the hobby and the product_j0 indicates that the hobbies are unrelated to the product, and the hobbies having no influence on the purchase of the product in the sample account.

Specifically, sample service experience information of any sample account in a sample set is obtained, and for each type of preference, a preprocessed preference value of the sample account for the type of preference is determined based on survey scores of the sample account for the type of preference and a preset association degree of the type of preference and products purchased by the sample account.

Wherein the association degree of each kind of hobbies with each product is set according to human experience.

In a possible manner, if the sample service experience information includes preference information of the sample account, the determining a pre-processed preference value of the sample account for the type of preference based on the survey score of the sample account for the type of preference and a pre-configured degree of association between the type of preference and the product includes:

C_3j(A_i)＝λ_jG_j，

wherein A is_iAn ith product purchased for the sample account; c_3j(A_i) The preprocessed preference value of the sample account for the j-th preference is obtained; lambda [ alpha ]_jA survey score representing a class j taste for the sample account; g_jIndicating the relevance of the pre-configured jth category preference to the ith product.

Fourth, since in an actual application scenario, the sample service experience information may include the experience time information of the sample account, in order to enable the prediction model to learn a potential relationship existing between the product and the experience time information of the account, in an embodiment of the present invention, if the sample service experience information includes the experience time information of the sample account, for the sample service experience information of each sample account, the number of times that the sample account makes a reservation in each hour within the service time period may be determined according to a preconfigured service time period, for example, 8:00-17:00 per day, and the reservation time of each reservation of the sample account, for example, if a certain sample account makes a reservation twice in 14:30 and once in 14:40, then the sample account is determined to make a reservation 3 times within the service time period of 14:00, and determining the hours of each reservation of the sample account, and determining the pre-processed reservation time value of the sample account according to the reservation times and the hours of the sample account.

In a possible implementation manner, if the sample service experience information includes experience time information of the sample account, the determining the pre-processed appointment time value of the sample account according to the appointment number and the number of hours includes:

In a fifth mode, in an actual application scenario, the sample service experience information may include experience evaluation information of the sample account, and the experience evaluation information is also non-numerical information, which is not beneficial to processing by the original prediction model. In order to enable the prediction model to learn the potential relationship between the product and the experience evaluation information of the account, in the embodiment of the invention, if the sample service experience information includes the experience evaluation information of the sample account, the corresponding relationship between the experience evaluation information and the experience evaluation information value can be preset. In a specific implementation process, for each sample service experience information, according to a preset corresponding relationship between experience evaluation information and experience evaluation information values, a target experience evaluation information value corresponding to the experience evaluation information of the sample account is determined. For example, C₅(A_i) 0, 0.5 and 1, C₅(A_i) And for the target experience evaluation information value corresponding to the experience evaluation information of the sample account, the target experience evaluation information value is 0, which indicates that the sample account experiences the service unsatisfied, the target experience evaluation information value is 0.5, which indicates that the sample account experiences the service satisfactorily, and the target experience evaluation information value is 1, which indicates that the sample account experiences the service very satisfactorily.

It can be seen from the trained prediction model that since the information of the user such as age, gender, hobbies and the like is different, each service that can be reserved has different attractions for target accounts of different age groups, different sexes and different hobbies, and further, the association degree between the product finally purchased by the target account and the number of times the target account experiences the service is different, for example, some target accounts experience a certain service for many times, a product corresponding to the service is definitely purchased, some target accounts experience a certain service once, the product corresponding to the service is purchased, even some target accounts reserve a certain service, but finally purchase a product corresponding to another service. Therefore, through the prediction model, based on a large amount of second service experience information, the nonlinear input-output relationship can be learned to a certain extent, and the influence on products purchased by the account, such as 30% of users who have reserved visiting a memorial park, can purchase different-dimension physical examination packages.

Example 4:

in order to accurately determine the first target product, on the basis of the foregoing embodiments, in an embodiment of the present invention, before the obtaining, by the pre-trained prediction model, the first target product corresponding to the first service experience information, the method further includes:

determining a target type for each of the first service experience information;

In an actual application scenario, a user may not want to fill some information, such as name, gender, hobby, and the like, which has been filled in a previous service reservation every time a reservation is made, so that there may be a case that information is missing in the acquired reservation information, and further there may be a case that information is missing in the first service experience information determined based on the reservation information and the experience information of the target account. When it is determined that the first service experience information has a condition of information missing, it is indicated that the content in the first service experience information is incomplete, and the first target product cannot be accurately recommended to the target user according to the information contained in the first service experience information. Therefore, in order to ensure that the first target product recommended to the target account is accurately determined, necessary information types are configured in advance, and for each necessary information type, historical information of the necessary information type recorded by the target account in the historical experience service process is saved. In a specific implementation process, before a first target product corresponding to first service experience information is acquired through a pre-trained prediction model, a target type of each piece of acquired first service experience information is determined, then the target type is matched with a pre-configured necessary information type, and whether the first service experience information lacks any information of the necessary information type is determined. If it is determined that the first service experience information lacks information of the necessary information type, the current missing history information of the necessary information type may be determined from the pre-stored history information of each necessary information type of the target account and added to the first service experience information, so as to perform the subsequent steps according to the added first service experience information.

If it is determined that the first service experience information does not lack the information of the necessary information type, the subsequent steps can be directly performed according to the currently acquired first service experience information.

It should be noted that, if it is determined that the history information of a certain necessary information type of the target account is not currently stored, in order to ensure that the first target product recommended to the target account is subsequently and accurately determined, in the embodiment of the present invention, a prompt message for supplementing the necessary information type lacking the history information, for example, "please input age", may be output, so that the target account further improves the input first service experience information.

The prompt information for outputting the necessary information type for supplementing the missing history information may be a prompt information in a voice broadcast audio format, for example, a prompt information "please input age" for supplementing the necessary information type for missing history information is a prompt information in a voice broadcast, or a prompt information corresponding to a text form may be displayed on the display interface, for example, a prompt information "please supplement sex" for supplementing the necessary information type for missing history information is displayed on the display interface. The two modes of outputting the prompt information can also be combined at the same time, namely the prompt information in the audio format is broadcasted and the prompt information in the text format is displayed on the display interface.

Specifically, which mode is selected to output the prompt information may be preset according to the preference of the user, or may be selected according to the capability of the electronic device, for example, some electronic devices do not have a display interface capable of displaying the prompt information, and for these electronic devices, when the prompt information is output, the prompt information in the audio format may be broadcasted.

After the prompt information for supplementing the necessary information type lacking the historical information is output, the information supplemented by the user can be received, the supplemented information and the information type corresponding to the information are correspondingly stored, and according to the supplemented information and the previously acquired first service experience information, a first target product recommended to the target account is determined through a pre-trained prediction model.

Further, when the user subscribes to the service, some privacy information, such as an identification number, a mobile phone number, and the like, may be input, and if the privacy information is leaked, the property safety and the like of the user may be caused. Therefore, after the first service experience information is acquired, the privacy information in the first service experience information needs to be extracted, and desensitization processing is performed on the extracted privacy information. Specifically, the method for desensitizing the private information belongs to the prior art, and is not limited herein.

In order to further facilitate determining a product that can be recommended to a target account, in an embodiment of the present invention, before the obtaining, by using a pre-trained prediction model, a first target product corresponding to the first service experience information, the method further includes:

determining a target type for each of the first service experience information;

In the actual application process, the first service experience information input by the user may include many other useless contents besides the information of the necessary information type, and the useless contents are generally large and redundant, which not only consumes storage resources, but also wastes a large amount of resources for data processing. Therefore, in the embodiment of the present invention, in order to further facilitate determining the first target product that can be recommended to the target account, before the first target product corresponding to the first service experience information is acquired through the pre-trained prediction model, the content included in the acquired first service experience information is filtered.

In a specific implementation process, a target type of each piece of information in the first service experience information is determined. And then matching each target type with a preset necessary information type, and filtering information corresponding to the target type which is not matched with any preset necessary information type, namely filtering information of which the target type in the first service experience information is not the preset necessary information type, so as to realize filtering of useless information in the first service experience information.

Example 5:

fig. 3 is a schematic view of a specific process for determining product information according to an embodiment of the present invention, and a detailed description is now made of the product information method according to the embodiment of the present invention with reference to fig. 3:

the user makes a service reservation on the product recommendation system.

Specifically, the user can perform reservation service of old-age community visit, memorial park visit, oral cavity mechanism visit, physical examination mechanism visit, psychological consultation experience and green medical experience on the C-end APP of the product recommendation system. All the services can lead the user to enjoy experience in advance, thereby improving the user experience and providing different products for the user in a targeted manner according to the real feeling and evaluation of the user.

Sample service experience information is obtained.

In the specific implementation process, the product recommendation system mainly comprises a product center, a resource center and an order center, wherein the product center is responsible for managing products and service items sold on a management line and standardizing the products; the resource center is used for scheduling personnel, equipment, places and the like for providing services, managing resources online, facilitating online reservation of users and improving the utilization rate of the resources; the order center stores service experience information determined based on the reservation information and the experience information input by the account.

The persistent layer in the order center synchronizes the service experience information stored in the Mysql database and the mongoDB database to a distributed Search (es) data center in real time, and finally stores the service experience information in the es data center. The scheme for synchronizing the service experience information stored in the Mysql database to the es data center is as follows:

the canal real-time synchronizer is deployed as a slave to the Mysql database, reads the log data binlog, and sends updates to the table data in the Mysql database to the kafka message center in a structured json format.

Because the service experience information stored in the Mysql database is stored in the es data center in real time, the standardization requirement on the data is high, the kafka message is directly stored in the es, the data processing difficulty is high, and the maintenance is not easy. Therefore, in the embodiment of the invention, the message data in the kafka message center is not directly stored in the es data center, but the message of the kafka message center is subscribed through a micro-service consuming the kafka message, so that the message instantaneity is ensured, and the service experience information can be flexibly processed. When consuming the kafka message, the kafka message can be personalized according to different service types.

In the embodiment of the invention, the cluster construction of the es data center adopts a master-slave cluster mode, for example, a master cluster is 3 es data center master nodes and 5 es data center data nodes. The standby cluster is also 3 es data center main nodes and 5 es data center data nodes, and when the main cluster provides different services, the standby cluster is automatically started.

After the service experience information stored in the Mysql database is stored in the es data center, a rest api interface enabling product is provided as an input parameter for prediction model training based on the ability of es full-text fast indexing.

The scheme for synchronizing the service experience information stored in the mongoDB database to the es data center is as follows:

at present, there are various ways to synchronize data stored in the mongoDB database to the es data center, and services can be written to the es data center and the mongoDB database through web server synchronization in a double-writing mode. Buffered inputs in logstack can also be exploited by adding a mongo input and ES output plug-in, but this approach is very dependent on Java database connectivity (JDBC). Therefore, in the embodiment of the invention, a mongo-connector mode can be adopted, the tool synchronizes data between the mongo DB and the es data center, tracks the operation log data oplog of the mongo DB, and can perform real-time synchronization operation on the data stored in the es data center when the operation on the data in the mongo DB is kept.

In the specific implementation process, firstly, a copy set mode of the mongoDB is started, a copy set name is set, and initialization and state verification are carried out on the copy set.

And configuring an es data center, and setting an es data center connected by the mongoDB database through the mongo-connector, specifically determining the ip and the port of the mongoDB database and information of an es data center cluster.

Verifying that the service experience information is updated in the mongoDB database, and inquiring whether the corresponding data is changed by the es data center.

The service experience information stored in the es data center is determined as sample service experience information, each sample service experience information is preprocessed, such as cleaning processing, desensitization processing and the like, the preprocessed sample service experience information is determined as second service experience information, an original prediction model is trained subsequently based on each second service experience information, whether the currently trained prediction model meets a preconfigured convergence condition is judged, and therefore whether the trained prediction model is obtained is determined.

If the currently trained prediction model is determined to meet the pre-configured convergence condition, the trained prediction model is determined to be obtained, the trained prediction model is output, and the first service experience information is obtained based on the method in the embodiment. And determining a first target product corresponding to the first service experience information based on the first service experience information through the trained prediction model, so that the product information of the first target product is recommended to a target account.

The method comprises the steps of carrying out analysis training on products finally purchased by a target account based on age information of the target account, gender information of the target account, preference information of the target account, experience time information of the target account and experience evaluation information of the target account, establishing a neural network prediction model, and carrying out intelligent recommendation on the products based on the model, namely, the first target products more suitable for the target account are provided for the target account in a targeted manner through the trained prediction model and based on first service experience information of the target account.

In order to further improve the accuracy of the first target products recommended to the target account, before the product information of the first target products is recommended to the target account, each determined first target product may be sent to the relevant staff, so that the accuracy of each first target product may be screened through the experience of the relevant staff. When the relevant staff member determines that a first target product is suitable to be recommended to the target account, the determination information of the first target product can be input. After the electronic equipment for determining the product information receives the determination information of the first target product, the product information of the first target product is recommended to the target account. When the related staff determines that a certain first target product is not suitable for being recommended to the target account, the determination information of the first target product is not input, or the error information of the first target product is input, so that the product information of the first target product is not recommended to the target account subsequently.

Example 6:

an embodiment of the present invention provides a product information apparatus, and fig. 4 is a schematic structural diagram of an apparatus for determining product information according to an embodiment of the present invention, where the apparatus includes:

an obtaining unit 41, configured to obtain first service experience information of a target account;

the processing unit 42 is configured to obtain, through a pre-trained prediction model, a first target product corresponding to the first service experience information; wherein the first service experience information comprises at least one of age information of the target account, gender information of the target account, hobby information of the target account, experience time information of the target account and experience evaluation information of the target account; the prediction model is obtained by training an original prediction model based on second service experience information, and the second service experience information is obtained by preprocessing sample service experience information in a sample set;

a recommending unit 43, configured to determine product information of the first target product, and recommend the product information of the first target product to the target account;

wherein the apparatus further comprises: a training unit 44;

the training unit 44 is configured to obtain second service experience information of any sample account and a corresponding label thereof, where the label is used to identify a product purchased by the sample account; acquiring a second target product corresponding to the second service experience information through the original prediction model; and training the original prediction model according to the label and the second target product.

The device further comprises: a pre-processing unit;

Further, the preprocessing unit is specifically configured to, if the sample service experience information includes age information of the sample account, determine the preprocessed age value of the sample account based on the age information of the sample account, an average value of the age information of each account for purchasing the product, and a mean square error of the age information of each account for purchasing the product, and includes:

C_3j(A_i)＝λ_jG_j，

Further, the apparatus further comprises:

the processing unit 42 is further configured to, if it is determined that the first service experience information lacks information of a necessary information type according to a pre-configured necessary information type and the target type, determine history information of the missing necessary information type from history information of each necessary information type of the target account stored in advance, and add the history information to the first service experience information.

Further, the apparatus further comprises:

the processing unit 42 is further configured to filter information in the first service experience information, where the target type is not a pre-configured necessary information type.

Further, the recommending unit 43 is specifically configured to recommend the product information of the first target product to the target account if the confirmation information of the first target product is received.

Example 7:

as shown in fig. 5, which is a schematic structural diagram of an electronic device according to an embodiment of the present invention, on the basis of the foregoing embodiments, the electronic device includes: the system comprises a processor 51, a communication interface 52, a memory 53 and a communication bus 54, wherein the processor 51, the communication interface 52 and the memory 53 are communicated with each other through the communication bus 54; the memory 53 has stored therein a computer program which, when executed by the processor 51, causes the processor 51 to perform the steps of:

acquiring first service experience information of a target account;

Because the principle of the electronic device for solving the problems is similar to the product information method, the implementation of the electronic device can refer to the implementation of the method, and repeated details are not repeated.

The communication bus mentioned in the above embodiments may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The communication interface 52 is used for communication between the above-described electronic apparatus and other apparatuses. The Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Alternatively, the memory may be at least one memory device located remotely from the processor. The Processor may be a general-purpose Processor, including a central processing unit, a Network Processor (NP), and the like; but may also be a Digital instruction processor (DSP), an application specific integrated circuit, a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or the like.

Example 8:

on the basis of the foregoing embodiments, the present invention further provides a computer-readable storage medium, in which a computer program executable by a processor is stored, and when the program is run on the processor, the processor is caused to execute the following steps:

acquiring first service experience information of a target account;

Since the principle of the computer-readable storage medium to solve the problem is similar to the product information method in the above-described embodiment, specific implementation may be referred to implementation of the product information method.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method of product information, the method comprising:

acquiring first service experience information of a target account;

2. The method of claim 1, wherein preprocessing the sample service experience information in the sample set comprises:

3. The method of claim 2, wherein preprocessing the sample service experience information in the sample set further comprises:

if the sample service experience information comprises the experience time information of the sample account, determining the reserved times of each hour and the reserved hours of each time of the sample account in the service time period according to a preset service time period and each experience time information of the sample account; determining a preprocessed experience time value of the sample account according to the reserved times and the hours;

4. The method of claim 3, wherein if the sample service experience information includes age information for the sample account, the determining the pre-processed age value for the sample account based on the age information for the sample account, the average of the age information for each account that purchased the product, and the mean square error of the age information for each account that purchased the product comprises:

C_3j(A_i)＝λ_jG_j，

5. The method of claim 1, wherein before the pre-trained predictive model is used to obtain the first target product corresponding to the first service experience information, the method further comprises:

determining a target type for each of the first service experience information;

6. The method of claim 1, wherein before the pre-trained predictive model is used to obtain the first target product corresponding to the first service experience information, the method further comprises:

determining a target type for each of the first service experience information;

7. The method of claim 1, wherein recommending product information for the first target product to the target account comprises:

8. A product information device, the device comprising:

wherein the apparatus further comprises: a training unit;

9. An electronic device, characterized in that the electronic device comprises at least a processor and a memory, the processor being adapted to carry out the steps of the product information method according to any of claims 1-7 when executing a computer program stored in the memory.

10. A computer-readable storage medium, characterized in that it stores a computer program which, when being executed by a processor, carries out the steps of the product information method according to any one of claims 1 to 7.