CN113706317A

CN113706317A - Data processing method and device, electronic equipment and storage medium

Info

Publication number: CN113706317A
Application number: CN202110994800.XA
Authority: CN
Inventors: 周朋飞; 张捷
Original assignee: Shanghai Sensetime Intelligent Technology Co Ltd
Current assignee: Shanghai Sensetime Intelligent Technology Co Ltd
Priority date: 2021-08-27
Filing date: 2021-08-27
Publication date: 2021-11-26

Abstract

The application discloses a data processing method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring current underwriting data of an underwriting bill; acquiring at least one historical data of the warrant to be checked; fusing the current underwriting data and the at least one historical data to obtain first fused data; and obtaining a processing result of the to-be-checked insurance policy based on the first fusion data. Based on the technical scheme provided by the application, the processing accuracy can be improved by processing the insurance policy to be checked.

Description

Data processing method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of machine learning technologies, and in particular, to a data processing method and apparatus, an electronic device, and a storage medium.

Background

Insurance underwriting refers to the determination of underwriting by the insurance company through risk assessment of the policy. In the process of underwriting, if the risk control is too loose, the insurance company can be maliciously cheated by some customers, and if the risk control is too tight, part of the customers can be refused to be protected, and the performance of the insurance company is further declined. Therefore, how to accurately verify the insurance has very important significance for the insurance company.

Disclosure of Invention

The application provides a data processing method and device, electronic equipment and a storage medium.

In a first aspect, a data processing method is provided, the method including:

acquiring current underwriting data of an underwriting bill;

acquiring at least one historical data of the warrant to be checked;

fusing the current underwriting data and the at least one historical data to obtain first fused data;

and obtaining a processing result of the to-be-checked insurance policy based on the first fusion data.

With reference to any embodiment of the present application, the at least one historical data includes a first target historical data, and the fusing the current underwriting data and the at least one historical data to obtain a first fused data includes:

determining a first correlation of the first target historical data with the current underwriting data;

obtaining a first weight of the first target historical data based on the first correlation, wherein the first weight is positively correlated with the first correlation;

obtaining first updated first target historical data based on the product of the first weight and the first target historical data; and

and performing fusion processing on the current underwriting data and the first updated first target historical data to obtain the first fusion data.

With reference to any one of the embodiments of the present application, before the determining the first correlation between the first target historical data and the current underwriting data, the method further includes:

constructing a time sequence of the current underwriting data and the first target historical data to obtain a characteristic vector sequence;

carrying out nonlinear transformation on the feature vector sequence to obtain a second weight of the first target historical data, wherein the second weight is positively correlated with the information content carried by the first target historical data;

obtaining second updated first target historical data based on the product of the first target historical data and the second weight; and

the determining a first correlation of the first target historical data and the current underwriting data comprises:

and determining a second correlation between the second updated first target historical data and the current underwriting data as the first correlation.

With reference to any embodiment of the present application, the obtaining first updated target history data based on a product of the first weight and the first target history data includes:

obtaining the first updated first target history data based on a product of the first weight and the second updated first target history data.

With reference to any embodiment of the present application, before the obtaining the first updated first target historical data based on a product of the first weight and the second updated first target historical data, the at least one historical data further includes a second target historical data, and the method further includes:

determining a third correlation between the second target historical data and the current underwriting data;

obtaining a third weight of the second target historical data based on the third correlation, wherein the third weight is positively correlated with the third correlation; and

obtaining the first updated first target history data based on a product of the first weight and the second updated first target history data, including:

and performing fusion processing on the second updated first target historical data and the second target historical data based on the first weight and the third weight to obtain the first updated first target historical data.

With reference to any embodiment of the present application, the fusing the second updated first target historical data and the second target historical data based on the first weight and the third weight to obtain the first updated first target historical data includes:

and performing weighted average on the second updated first target historical data and the second target historical data based on the first weight and the third weight to obtain the first updated first target historical data.

With reference to any embodiment of the present application, before obtaining the processing result of the policy to be checked based on the first fusion data, the method further includes:

acquiring user data carrying user information, wherein the user information comprises at least one of the following: information of the applicant of the to-be-verified insurance policy and information of the to-be-verified object of the to-be-verified insurance policy; and

the obtaining of the processing result of the to-be-certified policy based on the first fusion data includes:

fusing the user data and the first fused data to obtain second fused data;

and obtaining the processing result based on the second fusion data.

With reference to any embodiment of the present application, the obtaining the processing result based on the second fusion data includes:

performing fusion processing on the first dimensional data and the second dimensional data to obtain third fusion data;

and obtaining the processing result based on the third fusion data.

With reference to any embodiment of the present application, the fusing the user data and the first fused data to obtain second fused data includes:

splicing the first fusion data and the user data to obtain spliced data; and

and carrying out nonlinear transformation on the spliced data to obtain the second fusion data.

With reference to any embodiment of the present application, the acquiring user data carrying user information includes:

acquiring the user information; and

and coding the user information to obtain the user data.

In a second aspect, there is provided a data processing apparatus comprising:

the acquisition unit is used for acquiring the current underwriting data of the to-be-underwriting bill;

the acquisition unit is further used for acquiring at least one historical data of the to-be-certified insurance policy;

the first processing unit is used for carrying out fusion processing on the current underwriting data and the at least one historical data to obtain first fusion data;

and the second processing unit is used for obtaining a processing result of the to-be-checked insurance policy based on the first fusion data.

With reference to any one of the embodiments of the present application, the first processing unit is configured to:

With reference to any one of the embodiments of the present application, the data processing apparatus further includes: a third processing unit, configured to construct a time sequence of the current underwriting data and the first target historical data to obtain a feature vector sequence before the first correlation between the first target historical data and the current underwriting data is determined;

the first processing unit is configured to:

With reference to any one of the embodiments of the present application, the at least one historical data further includes a second target historical data, and the data processing apparatus further includes:

a fourth processing unit, configured to determine a third correlation between the second target history data and the current underwriting data before the first updated first target history data is obtained based on a product of the first weight and the second updated first target history data;

the first processing unit is configured to:

With reference to any embodiment of the present application, the obtaining unit is further configured to obtain user data carrying user information before obtaining a processing result of the policy to be certified based on the first fusion data, where the user information includes at least one of: information of the applicant of the to-be-verified insurance policy and information of the to-be-verified object of the to-be-verified insurance policy; and

the second processing unit is configured to:

fusing the user data and the first fused data to obtain second fused data;

and obtaining the processing result based on the second fusion data.

With reference to any embodiment of the present application, the second processing unit is configured to:

and obtaining the processing result based on the third fusion data.

splicing the first fusion data and the user data to obtain spliced data; and

With reference to any embodiment of the present application, the obtaining unit is configured to:

acquiring the user information; and

and coding the user information to obtain the user data.

In a third aspect, an electronic device is provided, which includes: a processor and a memory for storing computer program code comprising computer instructions, the electronic device performing the method of the first aspect and any one of its possible implementations as described above, if the processor executes the computer instructions.

In a fourth aspect, another electronic device is provided, including: a processor, transmitting means, input means, output means, and a memory for storing computer program code comprising computer instructions, which, when executed by the processor, cause the electronic device to perform the method of the first aspect and any one of its possible implementations.

In a fifth aspect, there is provided a computer-readable storage medium having stored therein a computer program comprising program instructions which, if executed by a processor, cause the processor to perform the method of the first aspect and any one of its possible implementations.

A sixth aspect provides a computer program product comprising a computer program or instructions which, when run on a computer, causes the computer to perform the method of the first aspect and any of its possible implementations.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments or the background art of the present application, the drawings required to be used in the embodiments or the background art of the present application will be described below.

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application.

Fig. 1 is a schematic flowchart of a data processing method according to an embodiment of the present application;

fig. 2 is a schematic diagram of a splicing process provided in an embodiment of the present application;

fig. 3 is a schematic structural diagram of an underwriting model according to an embodiment of the present disclosure;

FIG. 4 is a schematic structural diagram of another underwriting model provided in the embodiments of the present application;

fig. 5 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present application;

fig. 6 is a schematic hardware structure diagram of a data processing apparatus according to an embodiment of the present application.

Detailed Description

In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, 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 application.

The terms "first," "second," and the like in the description and claims of the present application and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.

It should be understood that in the present application, "at least one" means one or more, "a plurality" means two or more, "at least two" means two or three and three or more, "and/or" for describing an association relationship of associated objects, meaning that three relationships may exist, for example, "a and/or B" may mean: only A, only B and both A and B are present, wherein A and B may be singular or plural.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.

The traditional underwriting work is completed manually, the method needs to consume large labor cost, and the accuracy of manual underwriting is low. With the development of science and technology, intelligent underwriting is applied.

At present, an intelligent data processing method adopts an underwriting model to underwritten a to-be-underwritten policy, but because information carried by the to-be-underwritten policy is limited, a processing result of the to-be-underwritten policy is determined only based on the information carried by the to-be-underwritten policy, so that the accuracy of the processing result is low. Based on this, the embodiment of the application provides a data processing method to improve the accuracy of a processing result.

The execution subject of the embodiment of the application is a data processing device. Optionally, the data processing device may be one of the following: cell-phone, computer, server, panel computer. The embodiments of the present application will be described below with reference to the drawings. Referring to fig. 1, fig. 1 is a schematic flowchart illustrating a data processing method according to an embodiment of the present disclosure.

101. And acquiring current underwriting data of the to-be-underwriting bill.

102. At least one historical data of the to-be-checked insurance policy is obtained.

In the embodiment of the application, the policy to be checked is the current policy, and the current checking data carries information carried by the policy to be checked. The historical data carries information of historical policies, wherein the historical policies correspond to the policies to be checked, and the historical policies corresponding to any two pieces of historical data are different, that is, the timestamps of different pieces of historical data are different when the number of the pieces of historical data exceeds 1. For example, the historical data includes historical data a and historical data b, wherein the timestamp of the policy corresponding to the historical data a is 3, 20 and 2018, and the timestamp of the policy corresponding to the historical data b is 5, 1 and 2020.

In the embodiment of the application, the applicant of the policy to be checked is the same as the applicant of the historical policy corresponding to the historical data, and/or the object to be guaranteed of the policy to be checked is the same as the object to be guaranteed of the historical policy corresponding to the historical data. For example, the applicant of the policy to be checked and the applicant of the historical policy corresponding to the historical data are Zhang III; the insured person of the policy to be checked and the insured person of the historical policy corresponding to the historical data are all LiIV; the guaranteed vehicle of the warranty to be checked and the guaranteed vehicle of the historical warranty corresponding to the historical data are both vehicles a; the applicant of the policy to be underwritten and the applicant of the historical policy corresponding to the historical data are Zhang three, and the insured person of the policy to be underwritten and the insured person of the historical policy corresponding to the historical data are Li four.

In the embodiment of the application, the information carried by the policy to be checked and the information of the historical policy can be the information with the same dimension or the information with different dimensions.

For example, the to-be-certified policy includes the following information: the age of Zhang III and the work of Zhang III. The information of the historical policy includes: the physical examination report of Zhang III and the work of Zhang III.

For another example, the to-be-certified policy includes the following information: age of Zhang III. The historical policy includes the following information: zhang San physical examination report.

For another example, the to-be-certified policy includes the following information: the age of Zhang III and the work of Zhang III. The information of the historical policy includes: physical examination report of Zhang III, age of Zhang III.

Optionally, when the number of the historical data exceeds 1, the information carried by the different historical data may be information of the same dimension, or information of different dimensions. For example, the at least one historical data includes historical data a and historical data B, where the historical data a carries information of historical policy a and the historical data B carries information of historical policy B. The information of the historical policy A comprises the age of Zhang III and the work of Zhang III, and the information of the historical policy B comprises: age of Zhang III. At this time, the history data a carries information of two dimensions, and the history data b carries information of one dimension.

Optionally, the information carried by the policy (including the policy to be checked and the historical policy) in the embodiment of the present application includes at least one of the following: insurance period, insurance amount, insurance type.

Optionally, the information carried by the historical policy includes the claim settlement result of the historical policy. For example, the applicant for historical policy a is Zhang III, and it is confirmed that historical policy a was cheated by Zhang III. Thus, the claim settlement results of the historical policy a include fraud. For another example, the applicant of historical policy a is Zhang III, which hides his own disease history from historical policy a. Thus, the claim settlement results for historical policy a include a true disease history of Zhang III. As another example, the claim settlement results for historical policy a include the claim settlement cost for policy a.

Optionally, the electronic device obtains current underwriting data by encoding the to-be-underwriting policy, and obtains historical data of the to-be-underwriting policy by encoding the historical underwriting policy.

In one implementation of obtaining the current underwriting data, the data processing device receives the current underwriting data of the warranty to be underwritten, which is input by a user through the input component. The above-mentioned input assembly includes: keyboard, mouse, touch screen, touch pad, audio input device, etc.

In another implementation manner of acquiring the current underwriting data, the data processing device receives the current underwriting data of the warranty to be underwrited, which is sent by the terminal. Optionally, the terminal may be any one of the following: cell-phone, computer, panel computer, server, wearable equipment.

In one implementation of obtaining at least one historical data, a data processing apparatus receives at least one historical data input by a user via an input component.

In another implementation of obtaining at least one history data, the data processing apparatus receives at least one history data transmitted by the terminal.

103. And performing fusion processing on the current underwriting data and at least one historical data to obtain first fusion data.

The data processing device can fuse the information carried by the current underwriting data and the information carried by the historical data by fusing the current underwriting data and the historical data, namely, the information carried by the current underwriting data and the information carried by the historical data are fused.

In the case where the number of the history data exceeds 1, the data processing apparatus may fuse the first feature vector and the partial history data to obtain first fused data.

For example, the at least one historical data includes: history data a and history data b. The data processing device can fuse the current underwriting data and the historical data a to obtain first fused data. The data processing device can also fuse the current underwriting data and the historical data b to obtain first fused data.

As another example, the at least one historical data includes: history data a, history data b, and history data c. The data processing device can fuse the current underwriting data and the historical data a to obtain first fused data. The data processing device can also fuse the current underwriting data, the historical data a and the historical data b to obtain first fused data.

Under the condition that the quantity of the historical data exceeds 1, the data processing device can also fuse the current underwriting data with all the historical data to obtain first fused data;

for example, the at least one historical data includes: history data a and history data b. The data processing device can fuse the current underwriting data, the historical data a and the historical data b to obtain first fused data.

In one possible implementation manner, the data processing device obtains the first fusion data by performing weighted average on the current underwriting data and the at least one historical data.

In another possible implementation manner, the data processing device obtains the first fusion data by summing the current underwriting data and the at least one historical data.

In yet another possible implementation manner, the data processing apparatus calculates an intelligent product (element-wise) between the current underwriting data and the at least one historical data as the first fused data.

104. And obtaining a processing result of the insurance policy to be checked based on the first fusion data.

Because the applicant of the to-be-certified insurance policy is the same as the applicant of the historical insurance policy, and/or the protected object of the to-be-certified insurance policy is the same as the protected object of the historical insurance policy, the information of the historical insurance policy carried by the historical data has referential property for auditing the to-be-certified insurance policy. And the data processing device fuses the information carried by the policy to be checked and the information of the historical policy to the first fused data by executing step 103.

In this way, the data processing device obtains the processing result of the to-be-certified security policy according to the first fusion data, and the processing result can be obtained by using the information carried by the to-be-certified security policy and the information carried by the historical security policy. Optionally, the processing result includes an underwriting result of the warranty to be underwritten.

In the embodiment of the application, the processing result of the policy to be checked includes at least one of the following: insurance acceptance, insurance rejection, charging amount and prolonging period. For example, the processing result of the warranty to be underwritten may be an underwriting; the insurance policy to be checked can be charging insurance, wherein the charging amount is 5 ten-thousand yuan; the insurance policy to be checked can also be a delay underwriting, wherein the extension period is half a year.

In one possible implementation, the data processing apparatus obtains a Support Vector Machine (SVM) before executing step 104. And processing the first fusion data by using the SVM to obtain a processing result of the to-be-verified insurance policy.

In another possible implementation manner, the data processing apparatus acquires a multilayer perceptron (MLP) before executing step 104. And processing the first fusion data by using MLP to obtain a processing result of the to-be-checked insurance policy.

In the embodiment of the application, the data processing device performs fusion processing on the current underwriting data and the historical data, fuses information carried by the to-be-underwriting policy and information carried by the historical policy, and obtains first fusion data. Therefore, when the processing result of the warranty to be checked is obtained according to the first fusion data, the processing result can be obtained by utilizing the information carried by the warranty to be checked and the information carried by the historical warranty, and the accuracy of the processing result can be improved.

As an optional implementation, the at least one history data includes the first target history data, that is, the first target history data is any one of the at least one history data. The data processing apparatus performs the following steps in executing step 103:

1. a first correlation between the first target historical data and the current underwriting data is determined.

In the embodiment of the application, the greater the correlation between the historical data and the current underwriting data is, the higher the accuracy of the processing result of the to-be-underwriting policy corresponding to the current underwriting data is obtained by the electronic device according to the information carried by one of the historical data.

For example, the current underwriting data of the policy to be underwritten is current underwriting data a, the historical policy corresponding to the historical data 1 is a feature vector b, and the feature vector of the historical policy 2 is a feature vector c. The correlation between the current underwriting data a and the historical data b is larger than that between the current underwriting data a and the historical data c, and the processing result of the to-be-underwriting bill obtained by the electronic equipment according to the information carried by the historical data b is represented more accurately than the processing result of the to-be-underwriting bill obtained by the electronic equipment according to the information carried by the historical data c.

In a possible implementation manner, the data processing apparatus acquires a deep learning model with an attention mechanism (attention mechanism) before performing step 1, and processes the current underwriting data and the first target historical data using the deep learning model to obtain a correlation between the current underwriting data and the first target historical data, that is, a first correlation.

In another possible implementation manner, the data processing apparatus calculates a correlation coefficient between the current underwriting data and the first target history data as a correlation between the current underwriting data and the first target history data, that is, a first correlation.

Optionally, the correlation coefficient is one of the following: spearman (Spearman) rank correlation coefficient, Kendall (Kendall) rank correlation coefficient.

2. And obtaining a first weight of the first target historical data based on the first correlation.

In the embodiment of the application, the first weight is positively correlated with the first correlation, that is, the more accurate the processing result of the warranty to be checked, which is obtained by the electronic device according to the first target historical data, the larger the first weight is.

Assume that the first weight is w₁The first correlation is c₁。

In one possible implementation, w₁And c₁Satisfies the following formula:

w₁＝k₁×c₁… formula (1)

Wherein k is₁Is a positive number. Optionally, k₁＝1。

In another possible implementation, w₁And c₁Satisfies the following formula:

w₁＝k₁×c₁+d₁… formula (2)

Wherein k is₁Is a positive number, d₁Are real numbers. Optionally, k₁＝1，d₁＝0。

In yet another possible implementation, w₁And c₁Satisfies the following formula:

3. And obtaining first updated first target historical data based on the product of the first weight and the first target historical data.

Because the first weight is positively correlated with the first correlation, the data processing device obtains first updated target historical data according to the product of the first target historical data and the first weight, namely, the first updated target historical data is obtained by correcting the first target historical data according to the correlation between the first target historical data and the current underwriting data. In this way, in the subsequent processing, the accuracy of the processing result obtained by the data processing apparatus based on the first updated first target history data is higher than the accuracy of the processing result obtained based on the first target history data.

Assuming that the product of the first weight and the first target history data is e₁The first updated first target history data is p₁。

In one possible implementation, e₁And p₁Satisfies the following formula:

p₁＝k₂×e₁… formula (4)

Wherein k is₂Is a positive number. Optionally, k₂＝1。

In another possible implementation, e₁And p₁Satisfies the following formula:

p₁＝k₂×e₁+d₂… formula (5)

Wherein k is₂Is a positive number, d₂Are real numbers. Optionally, k₂＝1，d₂＝0。

In yet another possible implementation, e₁And p₁Satisfies the following formula:

It should be understood that, in practical applications, in the case where the number of history data exceeds 1, the data processing apparatus may determine the correlation between each history data and the current underwriting data, respectively. And respectively determining the weight of each historical data, and respectively correcting the historical data, namely obtaining the first corrected historical data according to the product of the weight of the historical data and the historical data.

For example, the at least one historical data includes a first target historical data and a first intermediate target historical data. The data processing device corrects the first target history data by executing step 2 to step 3 to obtain first updated first target history data. The data processing apparatus may also determine a first intermediate correlation between the first intermediate target historical data and the current underwriting data. A first intermediate weight of the first intermediate target historical data is obtained according to the first intermediate correlation. And obtaining updated second target historical data according to the product of the first intermediate weight and the first intermediate target historical data, wherein the updated second target historical data is the first corrected historical data obtained by correcting the first intermediate target historical data.

4. And performing fusion processing on the current underwriting data and the first updated first target historical data to obtain the first fusion data.

Since the first updated first target history data is obtained by correcting the first target history data, the accuracy of the processing result obtained by the data processing apparatus based on the first updated first target history data is higher than the accuracy of the processing result obtained based on the first target history data. Therefore, the data processing device obtains the first fusion data by fusing the current underwriting data and the first updated first target historical data, and the accuracy of the first fusion data can be improved.

In a possible implementation manner, the data processing apparatus obtains the first fusion data by performing weighted average on the current underwriting data and the first updated first target historical data.

In another possible implementation manner, the data processing apparatus obtains the first fusion data by summing the current underwriting data and the first updated first target historical data.

In yet another possible implementation manner, the data processing apparatus calculates an intelligent product between the current underwriting data and the first updated first target historical data as the first fused data.

As an alternative embodiment, before performing step 1, the data processing apparatus further performs the following steps:

5. and constructing a time sequence of the current underwriting data and the first target historical data to obtain a characteristic vector sequence.

In the embodiment of the application, the time sequence refers to a sequence obtained by sequencing policy data by electronic equipment according to the sequence of the timestamps from large to small or from small to large. For example, the data a is current underwriting data of the policy to be underwritten, and the data b is historical data corresponding to the historical policy a. Suppose that the effective date of the policy to be checked is 9/24/2020, and the effective time of the historical policy A is 1/3/2020. Then the timestamp for data a is 9/24/2020 and the timestamp for data b is 1/3/2020.

The electronic equipment sequences the data a and the data b according to the sequence of the timestamps from large to small, and the obtained time sequence is as follows: 1. data a; 2. and (c) data b. The electronic equipment sequences the data a and the data b according to the sequence of the timestamps from small to large, and the obtained time sequence is as follows: 1. data b; 2. data a.

And the data processing device constructs a time sequence of the current underwriting data and the first target historical data according to the time stamp of the current underwriting data and the time stamp of the first target historical data to obtain a characteristic vector sequence.

6. And carrying out nonlinear transformation on the characteristic vector sequence to obtain a second weight of the first target historical data.

In the implementation of the application, the second weight is positively correlated with the information amount carried by the first target historical data, that is, the greater the second weight is, the higher the accuracy of the processing result of the to-be-certified insurance policy obtained according to the first target historical data is.

Due to the correlation between the future development trend of the event and the historical development trend of the event, the importance degree of different historical data of the event for obtaining the future development trend of the event can be determined according to the historical development trend of the event.

For example, assume that the event is a predicted pedestrian motion trajectory. According to the motion track of the pedestrian in the past period of time, the motion track of the pedestrian in the future period of time can be predicted. However, it is obvious that, in the motion trajectory in the past period, the motion trajectory data with the larger time stamp is more important for predicting the future motion trajectory of the pedestrian.

For another example, assume an event is a prediction of rainfall in a certain place. According to the rainfall of the place in the past period, the rainfall of the place in the future period can be predicted. Since the rainfall in one place has a periodic tendency, the more the time interval between the time stamp of the rainfall data and the time stamp to be predicted is close to a positive number times the period in the rainfall in the past period, the more important the rainfall data is to predict the future rainfall.

In the embodiment of the application, the event is an underwriting, and the development trend of the future event is the processing result of the warranty to be underwrited. Because the characteristic vector sequence carries the historical development trend information of the underwriting, the data processing device can determine the improvement degree of the processing result of the underwriting policy to be processed according to the characteristic vector of the historical policy according to the historical development trend information by carrying out nonlinear transformation on the characteristic vector sequence.

For example, the feature vector sequence includes: the data processing method comprises data a, data b and data c, wherein the timestamp of the data a is minimum, the timestamp of the data c is maximum, the claim settlement result of the data a is cheated by a policyholder, and the claim settlement results of the data b and the data c are both not cheated by the policyholder.

Although the claim settlement result of the data a is cheated by the policyholder, the time stamp of the data a is minimum, and the claim settlement result after the data a is not cheated by the policyholder. Thus, the electronic device may determine that data a has a lower impact on the accuracy of the processing results. And the data b and the data c have higher improvement on the accuracy of the processing result, and the electronic device can determine that the improvement on the accuracy of the processing result by the data c is higher than that by the data b because the timestamp of the data c is greater than that of the data b.

In the embodiment of the application, the greater the improvement of the accuracy of the processing result of the to-be-checked insurance policy by the information carried by certain data in the feature vector sequence is, the greater the weight of the data is. For example, the feature vector sequence includes: the weight of the data a is greater than that of the data b when the accuracy of the processing result of the to-be-certified bill is improved by the data a than that of the data b.

In this step, the data processing apparatus performs nonlinear transformation on the feature vector to obtain a weight of a fourth feature vector, that is, a second weight, where the second weight and the fourth feature vector have a positive correlation with respect to a degree of improvement of accuracy of the processing result.

7. And obtaining second updated first target historical data based on the product of the first target historical data and the second weight.

Because the second weight is in positive correlation with the first target historical data to improve the accuracy of the processing result of the to-be-certified bill, the data processing device obtains the second updated first target historical data according to the product of the first target historical data and the second weight, which is equivalent to improve the accuracy of the processing result of the to-be-certified bill according to the first target historical data, and corrects the first target historical data to obtain the second updated first target historical data.

Assuming that the product of the second weight and the first target history data is e₂The second updated first target history data is p₂。

In one possible implementation, e₂And p₂Satisfies the following formula:

p₂＝k₃×e₂… formula (7)

Wherein k is₃Is a positive number. Optionally, k₃＝1。

In another possible implementation, p₂And p₃Satisfies the following formula:

p₂＝k₃×e₂+d₃… formula (8)

Wherein k is₃Is a positive number, d₃Are real numbers. Optionally, k₃＝1，d₃＝0。

In yet another possible implementation, p₂And p₃Satisfies the following formula:

It is to be understood that, in practical applications, in the case where the number of history data exceeds 1, the data processing apparatus may construct a feature vector sequence (hereinafter referred to as a first intermediate feature vector sequence) of all the history data and the current underwriting data. The data processing device can obtain the weight of corresponding historical data according to the improvement of the accuracy of the obtained processing result of the historical data by carrying out nonlinear transformation on the first intermediate characteristic vector sequence. And then, correcting the corresponding historical data according to the weight of the historical data, namely obtaining second corrected historical data according to the product of the weight of the historical data and the historical data.

For example, the at least one historical data includes a first target historical data and a third target historical data. The data processing device corrects the first target history data by executing step 5 to step 7, and obtains second updated first target history data. The data processing device may further obtain a second intermediate weight of the third target history data according to the improvement of the accuracy of the third target history data to the obtained processing result. And obtaining updated third target historical data according to the product of the second intermediate weight and the third target historical data, wherein the updated third target historical data is second modified historical data obtained by modifying the third target historical data.

After obtaining the second updated first target history data, the data processing apparatus performs the following steps in the process of performing step 1:

8. and determining a second correlation between the second updated first target historical data and the current underwriting data as the first correlation.

The implementation manner of this step can be seen in the implementation manner of step 1, specifically, the second updated first target history data in this step corresponds to the first target history data in step 1, and the second correlation in this step corresponds to the first correlation in step 1.

After executing step 8, the data processing apparatus executes the following steps in executing step 3:

9. and obtaining the first updated first target history data based on a product of the first weight and the second updated first target history data.

Because the first weight is positively correlated with the second weight, the data processing device obtains the first updated target historical data according to the product of the first updated target historical data and the first weight, and equivalently, corrects the first updated target historical data according to the correlation between the first updated target historical data and the current underwriting data to obtain the first updated target historical data. In this way, the first updated first target history data improves the accuracy of obtaining the processing result more than the second updated first target history data.

Assuming that the product of the first weight and the second updated first target history data is e₃The first updated first target history data is p₁₃。

In one possible implementation, e₃And p₁Satisfies the following formula:

p₁＝k₄×e₃… formula (10)

Wherein k is₄Is a positive number. Optionally, k₄＝1。

In another possible implementationIn the formula, e₃And p₁Satisfies the following formula:

p₁＝k₄×e₃+d₄… formula (11)

Wherein k is₄Is a positive number, d₄Are real numbers. Optionally, k₄＝1，d₄＝0。

In yet another possible implementation, e₃And p₁Satisfies the following formula:

As an alternative embodiment, the at least one history data includes not only the first target history data but also the second target history data. The data processing apparatus further performs the following steps before performing step 9:

10. and determining a third correlation between the second target historical data and the current underwriting data.

For an implementation manner of determining the correlation between the two data in this step, see the implementation manner of determining the correlation between the two data in step 1, specifically, the second target history data in this step corresponds to the fourth feature vector in step 1, and the third correlation in this step corresponds to the first correlation in step 1.

11. And obtaining a third weight of the second target historical data based on the third correlation.

In the embodiment of the present application, the third weight is positively correlated with the third correlation. In this step, the implementation manner of obtaining the weight of the data according to the correlation of the data may be referred to as the implementation manner of obtaining the weight of the data according to the correlation of the data in step 2, specifically, the third correlation in this step corresponds to the first correlation in step 2, the second target history data in this step corresponds to the first target history data in step 2, and the third weight in this step corresponds to the first weight in step 2.

After obtaining the third weight, the data processing apparatus performs the following steps in performing step 9:

12. and performing fusion processing on the second updated first target history data and the second target history data based on the first weight and the third weight to obtain the first updated first target history data.

Because the information carried by different historical data is different, and the accuracy of the processing result of the underwriting policy is improved by the information carried by the historical data, the data processing device obtains the first updated target historical data by fusing the different historical data, and the information for improving the accuracy of the processing result in the first updated target historical data can be enriched.

In a possible implementation manner, the data processing apparatus performs weighted average on the second updated first target history data and the second target history data according to the first weight and the third weight to obtain the first updated first target history data.

For example, assume that the first weight is w₁The third weight is w₂The first updated first target history data is p₁The sixth feature vector is p₂The seventh feature vector is p₃. Then

In another possible implementation manner, the data processing apparatus performs weighted summation on the second updated first target historical data and the second target historical data according to the first weight and the third weight to obtain the first updated first target historical data.

In yet another possible implementation manner, the data processing apparatus calculates a product of the first weight and the second updated first target history data to obtain fourth target history data, and calculates a product of the second weight and the second target history data to obtain fifth target history data. An intelligent product (element-wise) between the fourth target history data and the fifth target history data is calculated as the first updated first target history data.

As an alternative embodiment, the data processing apparatus further performs the following steps before performing step 104:

13. and acquiring user data carrying user information.

In the embodiment of the present application, the user information includes at least one of: insurance applicant information of the to-be-verified insurance policy and protected object information of the to-be-verified insurance policy. The user data is data carrying user information.

The applicant information includes at least one of: name of the applicant, age of the applicant, sex of the applicant, disease history of the applicant, job of the applicant, and annual income of the applicant. In the case where the secured object is a person, the secured object information includes at least one of: name of the insured life, age of the insured life, sex of the insured life, disease history of the insured life, job of the insured life. In the case where the secured object is a vehicle, the secured object information includes at least one of: the model of the vehicle to be protected, the brand of the vehicle to be protected, the age of the vehicle to be protected, the mileage of the vehicle to be protected, and the maintenance history of the vehicle to be protected.

In one implementation of obtaining user data, a data processing apparatus receives user data input by a user through an input component.

In another implementation of acquiring user data, the data processing apparatus receives user data sent by a terminal.

After acquiring the user data, the data processing apparatus executes the following steps in executing step 103:

14. and performing fusion processing on the user data and the first fusion data to obtain second fusion data.

The data processing device can fuse the information carried by the user data and the information carried by the first fused data by fusing the user data and the first fused data to obtain second fused data, wherein the second fused data carries the policy information (including the information of the policy to be checked and the information of the historical policy) and the user information.

15. And obtaining the processing result based on the second fusion data.

The user data is favorable for improving the accuracy of the processing result, the second fusion data carries the policy information and the user information, and the data processing device obtains the processing result according to the second fusion data, so that the accuracy of the processing result can be improved.

The implementation manner of this step can be seen in the implementation manner of step 104, specifically, the second fused data in this step corresponds to the first fused data in step 104.

As an alternative embodiment, the data processing device performs the following steps in the course of performing step 14:

16. and splicing the first fusion data and the user data to obtain spliced data.

In the embodiment of the present application, the splicing process (splice) may be referred to in fig. 2. As shown in fig. 2, the electronic device may obtain data c by performing a splicing process on data a and data b.

17. And carrying out nonlinear transformation on the spliced data to obtain the second fusion data.

As can be seen from fig. 2, the dimension of the data obtained by the stitching process is larger than the dimension of the data before the stitching process. The nonlinear transformation in this step can convert the high-dimensional vector into the low-dimensional vector, and fuse the information carried by the data of different dimensions into the data of the same dimension. For example, the data a includes: first dimension data, second dimension data, and third dimension data. The electronic equipment performs nonlinear transformation on the data A, and can fuse the first dimensional data and the second dimensional data to obtain fourth dimensional data. And taking data containing the third dimension data and the fourth dimension data as data obtained by carrying out nonlinear transformation on the data A.

In this step, the data processing apparatus performs nonlinear transformation on the spliced data, fuses data of different dimensions in the spliced data, and obtains second fused data.

As an alternative embodiment, the second fused data includes data of at least two dimensions, and the data processing device performs the following steps in the course of performing step 15:

18. and performing fusion processing on the data of all dimensions in the second fusion data to obtain third fusion data.

Because the information carried by the data with different dimensions is different, the information carried by the data can be used for obtaining the processing result, the data processing device obtains third fusion data by fusing the data with all dimensions in the second fusion data, and obtains the processing result according to the third fusion data in the subsequent processing, so that the accuracy of the processing result can be improved.

In one possible implementation manner, the data processing device obtains the third fused data by performing weighted average on the data of all the dimensions in the second fused data.

In another possible implementation manner, the data processing device obtains the third fused data by summing data of all dimensions in the second fused data.

In yet another possible implementation, the data processing apparatus calculates an intelligent product between data of all dimensions in the second fused data as the third fused data.

Optionally, the second fused data includes data of two dimensions, and the data of the two dimensions are the first dimension data and the second dimension data, respectively. And the data processing device performs fusion processing on the first dimensional data and the second dimensional data to obtain third fusion data.

19. And obtaining the processing result based on the third fusion data.

The implementation manner of this step can be seen in the implementation manner of step 104, and specifically, the third fused data in this step corresponds to the first fused data in step 104.

In the embodiment of the application, the data processing device performs fusion processing on the data of all dimensions in the second fusion data to obtain third fusion data with richer carried information. Therefore, the data processing device obtains the processing result according to the third fusion data, and the accuracy of the processing result can be improved.

As an alternative embodiment, the data processing device executes the following steps in the process of executing step 13:

20. and acquiring the user information.

In one implementation of obtaining user information, a data processing apparatus receives user information input by a user through an input component.

In another implementation of obtaining the user information, the data processing apparatus receives the user information sent by the terminal.

21. And coding the user information to obtain the user data.

In an embodiment of the present application, the encoding process is used to convert data into feature vectors. Optionally, the encoding process is one of the following: an embedding encoding process (embedding encoder) and a one-hot encoding process (one-hot encoder).

In this step, the data processing apparatus performs encoding processing on the user information to obtain a feature vector of the user information as user data.

Optionally, when the user information includes both continuous data and discrete data, the data processing apparatus constructs a first intermediate vector according to the continuous data, and performs encoding processing on the discrete data to obtain a second intermediate vector. And integrating the first intermediate vector and the second intermediate vector to obtain a characteristic vector of the user information, and taking the characteristic vector of the user information as user data.

As an optional implementation manner, based on the technical solutions disclosed above, an embodiment of the present application further provides an underwriting model, which is used for obtaining a processing result of an underwriting form.

Referring to fig. 3, fig. 3 is a schematic structural diagram of a first underwriting model 300. As shown in FIG. 3, the first underwriting model 300 includes: an embedded encoding module 301, a weight generation module 302, a weighted average module 303, a stitching module 304, a deep neural network module 305, a logistic regression module 306, and an activation module 307.

The embedded encoding module 301 is configured to perform encoding processing on the discrete data to obtain a feature vector. Specifically, the embedded encoding module 301 performs encoding processing on the policy to be certified to obtain a feature vector of the policy to be certified, that is, current policy data, the embedded encoding module 301 performs encoding processing on at least one historical policy to obtain at least one historical data, the embedded encoding module 301 performs encoding processing on discrete data in the user information to obtain an encoded feature vector, and integrates the encoded feature vector with continuous data in the user information to obtain the user data.

The weight generation module 302 is configured to determine a first correlation between the first target historical data and the current underwriting data, and obtain a first weight of the first target historical data according to the first correlation.

The weighted average module 303 is configured to perform weighted average on the second updated first target historical data and the second target historical data according to the first weight and the third weight to obtain first updated first target historical data.

The splicing module 304 is configured to splice the first fusion data and the user data to obtain spliced data.

The deep neural network module 305 comprises a deep neural network, and the module 305 is configured to perform nonlinear transformation on the spliced data to obtain second fused data.

The logistic regression module 306 is configured to perform fusion processing on the first dimensional data and the second dimensional data to obtain third fusion data.

The activation module 307 is configured to obtain a processing result of the to-be-checked insurance policy according to the first fusion data. Optionally, the activation module 307 comprises an activation function. And the data processing device inputs the first fusion data into the activation function, so that the probability of the passing of the to-be-verified insurance policy can be obtained. Under the condition that the passing probability of the to-be-certified bill exceeds the threshold value, the data processing device determines that the processing result of the to-be-certified bill is passing; and in the case that the probability of the passing of the warranty to be verified does not exceed the threshold value, the data processing device determines that the processing result of the warranty to be verified is not passed. The activation function may be one of the following: sigmoid, Tanh, ReLU.

Optionally, the training device may be used to train the first underwriting model 300 before the data processing device uses the first underwriting model 300 to process the current underwriting data and the at least one historical policy. The training device and the data processing device may be the same or different, and this is not limited in this embodiment of the present application. Optionally, the training device may be any one of: computer, panel computer, server, treater.

The training process comprises the following steps: the training device obtains training data and a first underwriting model before training, wherein the training data comprises a first training warrant to be underwritten and at least one first training historical warrant; the training device uses a first underwriting model before training to encode a first training to-be-underwriting policy to obtain first current training underwriting data, and the training device uses the first underwriting model before training to encode at least one first training historical policy to obtain at least one first training historical data. The training device uses a first underwriting model before training to process the training historical data and at least one training historical data to obtain a first training prediction result. And the training device obtains the loss of the first underwriting model before training according to the difference between the first training prediction result and the label of the first training underwriting sheet. The training device updates the lost parameters of the first underwriting model before training according to the loss of the first underwriting model before training until the parameters of the first underwriting model before training are converged, completes the training of the first underwriting model before training, and obtains the first underwriting model.

As an optional implementation manner, based on the technical solutions disclosed above, the embodiment of the present application further provides another underwriting model, which is used for obtaining a processing result of the to-be-underwrited policy.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a second underwriting model 400. As shown in FIG. 4, the second underwriting model 400 includes: an embedded coding module 401, a Gate Recurrence Unit (GRU) module 402, a weight generation module 403, an attention gating loop module 404, a stitching module 405, a deep neural network module 406, a logistic regression module 407, and an activation module 408.

The embedded encoding module 401 is configured to perform encoding processing on the discrete data to obtain a feature vector. Specifically, the embedded encoding module 401 performs encoding processing on the policy to be certified to obtain a feature vector of the policy to be certified, that is, current policy data, the embedded encoding module 401 performs encoding processing on at least one historical policy to obtain at least one historical data, the embedded encoding module performs encoding processing on discrete data in the user information to obtain an encoded feature vector, and the encoded feature vector and continuous data in the user information are integrated to obtain the user data.

The GRU module 402 is configured to perform nonlinear transformation on the feature vector sequence to obtain a second weight of the first target history data, and obtain second updated first target history data according to a product of the first target history data and the second weight.

The weight generating module 403 is configured to determine a first correlation between the first target historical data and the current underwriting data, obtain a first weight of the first target historical data according to the first correlation, and obtain a third weight of the second target historical data according to the second correlation.

The attention GRU module 404 is configured to perform weighted averaging on the first target history data and the second target history data according to the first weight and the third weight to obtain first updated target history data.

The splicing module 405 is configured to splice the first fusion data and the user data to obtain spliced data.

The deep neural network module 406 includes a deep neural network, and the module 406 is configured to perform a nonlinear transformation on the spliced data to obtain second fused data.

The logistic regression module 407 is configured to perform fusion processing on the first dimensional data and the second dimensional data to obtain third fusion data.

The activation module 408 is configured to obtain a processing result of the to-be-verified insurance policy according to the first fusion data. Optionally, the activation module 408 contains an activation function. And the data processing device inputs the first fusion data into the activation function, so that the probability of the passing of the to-be-verified insurance policy can be obtained. Under the condition that the passing probability of the to-be-certified bill exceeds the threshold value, the data processing device determines that the processing result of the to-be-certified bill is passing; and in the case that the probability of the passing of the warranty to be verified does not exceed the threshold value, the data processing device determines that the processing result of the warranty to be verified is not passed. The activation function may be one of the following: sigmoid, Tanh, ReLU.

Optionally, the training device may be used to train the second underwriting model 400 before the data processing device uses the second underwriting model 400 to process the current underwriting data and the at least one historical policy. The training device and the data processing device may be the same or different, and this is not limited in this embodiment of the present application. Optionally, the training device may be any one of: computer, panel computer, server, treater.

The training process comprises the following steps: the training device obtains training data and a second underwriting model before training, wherein the training data comprises a second training warranty to be underwritten and at least one second training historical warranty; the training device uses a second underwriting model before training to encode a second training to-be-underwriting policy to obtain second current training underwriting data, and the training device uses the second underwriting model before training to encode at least one second training historical policy to obtain at least one second training historical data. The training device uses a second underwriting model before training to process second current training underwriting data and at least one second training historical data to obtain a second training prediction result. And the training device obtains the loss of the second underwriting model before training according to the difference between the second training prediction result and the label of the second training warranty to be underwritten. The training device updates the parameters of the loss of the second underwriting model before training according to the loss of the second underwriting model before training until the parameters of the second underwriting model before training are converged, completes the training of the second underwriting model before training, and obtains the second underwriting model.

It will be understood by those skilled in the art that in the method of the present invention, the order of writing the steps does not imply a strict order of execution and any limitations on the implementation, and the specific order of execution of the steps should be determined by their function and possible inherent logic.

The method of the embodiments of the present application is set forth above in detail and the apparatus of the embodiments of the present application is provided below.

Referring to fig. 5, fig. 5 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present disclosure, in which the data processing apparatus 1 includes an obtaining unit 11, a first processing unit 12, and a second processing unit 13. Optionally, the data processing apparatus 1 further comprises a third processing unit 14 and a fourth processing unit 15, wherein:

the acquiring unit 11 is used for acquiring current underwriting data of the to-be-underwriting bill;

the obtaining unit 11 is further configured to obtain at least one historical data of the to-be-certified policy;

the first processing unit 12 is configured to perform fusion processing on the current underwriting data and the at least one historical data to obtain first fusion data;

and the second processing unit 13 is configured to obtain a processing result of the to-be-certified insurance policy based on the first fusion data.

In combination with any embodiment of the present application, the first processing unit 12 is configured to:

With reference to any one of the embodiments of the present application, the data processing apparatus further includes: a third processing unit 14, configured to construct a time sequence of the current underwriting data and the first target historical data before the determining of the first correlation between the first target historical data and the current underwriting data, so as to obtain a feature vector sequence;

the first processing unit 12 is configured to:

a fourth processing unit 15, configured to determine a third correlation between the second target history data and the current underwriting data before the first updated first target history data is obtained based on a product of the first weight and the second updated first target history data;

the first processing unit 12 is configured to:

With reference to any embodiment of the present application, the obtaining unit 11 is further configured to obtain user data carrying user information before obtaining a processing result of the policy to be certified based on the first fusion data, where the user information includes at least one of: information of the applicant of the to-be-verified insurance policy and information of the to-be-verified object of the to-be-verified insurance policy; and

the second processing unit 13 is configured to:

fusing the user data and the first fused data to obtain second fused data;

and obtaining the processing result based on the second fusion data.

With reference to any embodiment of the present application, the second processing unit 13 is configured to:

and obtaining the processing result based on the third fusion data.

splicing the first fusion data and the user data to obtain spliced data; and

With reference to any embodiment of the present application, the obtaining unit 11 is configured to:

acquiring the user information; and

and coding the user information to obtain the user data.

In some embodiments, functions of or modules included in the apparatus provided in the embodiments of the present application may be used to execute the method described in the above method embodiments, and specific implementation thereof may refer to the description of the above method embodiments, and for brevity, will not be described again here.

Fig. 6 is a schematic hardware structure diagram of a data processing apparatus according to an embodiment of the present application. The data processing device 2 comprises a processor 21, a memory 22, an input device 23, an output device 24. The processor 21, the memory 22, the input device 23 and the output device 24 are coupled by a connector, which includes various interfaces, transmission lines or buses, etc., and the embodiment of the present application is not limited thereto. It should be appreciated that in various embodiments of the present application, coupled refers to being interconnected in a particular manner, including being directly connected or indirectly connected through other devices, such as through various interfaces, transmission lines, buses, and the like.

The processor 21 may include one or more processors, for example, one or more Central Processing Units (CPUs), and in the case that the processor 21 is one CPU, the CPU may be a single-core CPU or a multi-core CPU.

The processor 21 is used for calling the program codes and data in the memory and executing the steps in the above method embodiments. Specifically, reference may be made to the description of the method embodiment, which is not repeated herein.

The memory 22 is used to store program codes and data for the network devices.

The memory 22 includes, but is not limited to, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), or a portable read-only memory (CD-ROM), which is used for related instructions and data.

The input means 23 are for inputting data and/or signals and the output means 24 are for outputting data and/or signals. The output device 24 and the input device 23 may be separate devices or may be an integral device.

It is understood that, in the embodiment of the present application, the memory 22 may be used to store not only the relevant instructions, but also relevant data, for example, the memory 22 may be used to store the current underwriting data of the to-be-underwriting policy acquired through the input device 23, or the memory 22 may also be used to store the processing result of the to-be-underwriting policy obtained by the processor 21, and the like, and the embodiment of the present application is not limited to the data specifically stored in the memory.

It will be appreciated that fig. 6 only shows a simplified design of a data processing device. In practical applications, the data processing apparatus may further include other necessary components, including but not limited to any number of input/output devices, processors, memories, etc., and all data processing apparatuses that can implement the embodiments of the present application are within the protection scope of the present application.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. It is also clear to those skilled in the art that the descriptions of the various embodiments of the present application have different emphasis, and for convenience and brevity of description, the same or similar parts may not be repeated in different embodiments, so that the parts that are not described or not described in detail in a certain embodiment may refer to the descriptions of other embodiments.

In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the above-described division of units is only one type of division of logical functions, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The processes or functions described above in accordance with the embodiments of the present application occur wholly or in part upon loading and execution of the above-described computer program instructions on a computer. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on or transmitted over a computer-readable storage medium. The computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a Digital Versatile Disk (DVD)), a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.

One of ordinary skill in the art will appreciate that all or part of the processes in the methods of the above embodiments may be implemented by hardware related to instructions of a computer program, which may be stored in a computer-readable storage medium, and when executed, may include the processes of the above method embodiments. And the aforementioned storage medium includes: various media that can store program codes, such as a read-only memory (ROM) or a Random Access Memory (RAM), a magnetic disk, or an optical disk.

Claims

1. A method of data processing, the method comprising:

acquiring current underwriting data of an underwriting bill;

acquiring at least one historical data of the warrant to be checked;

2. The method of claim 1, wherein the at least one historical data includes a first target historical data, and the fusing the current underwriting data and the at least one historical data to obtain a first fused data includes:

3. The method of claim 2, wherein prior to the determining the first correlation of the first target historical data to the current underwriting data, the method further comprises:

4. The method of claim 3, wherein obtaining the first updated first target history data based on a product of the first weight and the first target history data comprises:

5. The method of claim 4, wherein the at least one historical data further comprises a second target historical data, and wherein the method further comprises, prior to the obtaining the first updated first target historical data based on a product of the first weight and the second updated first target historical data:

6. The method according to claim 5, wherein the fusing the second updated first target history data and the second target history data based on the first weight and the third weight to obtain the first updated first target history data comprises:

7. The method according to any one of claims 1 to 6, wherein before the obtaining of the processing result of the policy to be certified based on the first fusion data, the method further comprises:

fusing the user data and the first fused data to obtain second fused data;

and obtaining the processing result based on the second fusion data.

8. The method of claim 7, wherein the second fused data comprises first dimension data and second dimension data, and wherein obtaining the processing result based on the second fused data comprises:

and obtaining the processing result based on the third fusion data.

9. The method according to claim 7 or 8, wherein the fusing the user data and the first fused data to obtain second fused data comprises:

splicing the first fusion data and the user data to obtain spliced data; and

10. The method according to any one of claims 7 to 9, wherein the obtaining user data carrying user information comprises:

acquiring the user information; and

and coding the user information to obtain the user data.

11. A data processing apparatus, characterized in that the data processing apparatus comprises:

12. An electronic device, comprising: a processor and a memory for storing computer program code comprising computer instructions which, if executed by the processor, the electronic device performs the method of any of claims 1 to 10.

13. A computer-readable storage medium, in which a computer program is stored, which computer program comprises program instructions which, if executed by a processor, cause the processor to carry out the method of any one of claims 1 to 10.