CN113177660A

CN113177660A - Driving intention prediction and processing method, device, equipment and storage medium

Info

Publication number: CN113177660A
Application number: CN202110464205.5A
Authority: CN
Inventors: 张霖; 朱磊; 付盼春
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2021-04-28
Filing date: 2021-04-28
Publication date: 2021-07-27
Anticipated expiration: 2041-04-28
Also published as: CN113177660B

Abstract

The invention relates to the technical field of artificial intelligence, and discloses a driving intention prediction and processing method, device, equipment and storage medium, which are used for improving the accuracy of predicting the driving intention of a user. The driving intention prediction and processing method comprises the following steps: acquiring position service track information, action hotspot track information and credit information according to the driving intention prediction request; determining a target characteristic factor data set based on the position service track information, the action hotspot track information and the credit information; acquiring a driving intention prediction result for the target characteristic factor data set through a preset risk user prediction model, and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not; if so, acquiring and pushing vehicle compensation service order information; and if not, generating and pushing a driving intention risk assessment report and early warning information of the target object. In addition, the invention also relates to a blockchain technology, and the driving intention risk assessment report can be stored in a blockchain node.

Description

Driving intention prediction and processing method, device, equipment and storage medium

Technical Field

The invention relates to the technical field of artificial intelligence neural networks, in particular to a driving intention prediction and processing method, device, equipment and storage medium.

Background

At present, vehicle damage caused by traffic accidents has an indemnity standard, and the damaged vehicle is economically indemnified according to the indemnity standard, so that the economic pressure of vehicle owners can be reduced. However, vehicle damage caused by traffic accidents has the condition that drivers intentionally collide with the rear or wipe vehicles to obtain vehicle compensation, and the healthy development of vehicle compensation enterprises is hindered.

Traditional risk identification schemes use models based on historical fraud data and empirically developed rules for determining fraud by insurance companies. Data that can be used to detect fraud is often limited to claims data collected by vehicle reimbursement enterprises and small amounts of external data. In the vehicle loss compensation business, drivers can cooperate with each other to obtain vehicle compensation by forging call records or transaction records, and the traditional risk identification scheme has the problems of low accuracy in predicting the driving intention of the users and low reliability of the vehicle loss compensation.

Disclosure of Invention

The invention provides a driving intention prediction and processing method, a driving intention prediction and processing device, a driving intention prediction device and a driving intention processing device and a storage medium, which are used for predicting driving intentions of a target object through a preset risk user prediction model, improving the accuracy and recognition efficiency of predicting the driving intentions of users and improving the reliability of vehicle loss compensation.

In order to achieve the above object, a first aspect of the present invention provides a driving intention prediction and processing method, including: receiving a driving intention prediction request of a target object, and acquiring position service track information of the target object, action hotspot track information of the target object and credit information of the target object according to the driving intention prediction request; performing feature analysis and screening on the position service track information, the action hotspot track information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating multi-dimensional feature factors related to target object fraud risk behaviors; predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not; if the driving intention prediction result is smaller than a preset risk threshold value, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information, and pushing the vehicle compensation service order information to a target terminal; and if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of a target object according to the driving intention prediction result and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal.

Optionally, in a first implementation manner of the first aspect of the present invention, the receiving a driving intention prediction request of a target object, and acquiring location service trajectory information of the target object, action hotspot trajectory information of the target object, and credit information of the target object according to the driving intention prediction request includes: receiving a driving intention prediction request of a target object, distributing the driving intention prediction request and starting a message processing thread; packaging the driving intention prediction request into a target business message through the message processing thread, and caching the target business message into a business message queue; calling a service processing thread corresponding to the service message queue, and performing message analysis on the target service message to obtain a unique identifier and a case identifier of a target object; and inquiring a preset map database according to the unique identifier of the target object and the case identification to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

Optionally, in a second implementation manner of the first aspect of the present invention, the performing feature analysis and screening on the location service trajectory information, the action hotspot trajectory information, and the credit information to obtain a target feature factor dataset, where the target feature factor dataset is used to indicate a multidimensional feature factor related to target object fraud risk behavior, includes: reading multidimensional candidate factors from a preset data table, and respectively performing characteristic analysis on the position service track information, the action hotspot track information and the credit information according to the multidimensional candidate factors to obtain a plurality of position service characteristic factors, a plurality of action hotspot characteristic factors and a plurality of credit characteristic factors; and clustering the plurality of location service characteristic factors, the plurality of action hotspot characteristic factors and the plurality of credit characteristic factors respectively through Euclidean distances to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

Optionally, in a third implementation manner of the first aspect of the present invention, the predicting driving intention of the target feature factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and determining whether the driving intention prediction result is smaller than a preset risk threshold includes: inputting the target characteristic factor data set into a preset risk user prediction model, and performing characteristic extraction on the target characteristic factor data set through an input layer in the preset risk user prediction model to obtain a plurality of sparse characteristics, wherein the preset risk user prediction model is a trained neural network factorization model; obtaining the dimensionality of each sparse feature, embedding each sparse feature into a corresponding dimensionality vector according to the dimensionality of each sparse feature through a neural network layer in the preset risk user prediction model to obtain a plurality of feature embedding vectors, wherein the plurality of feature embedding vectors comprise a plurality of low-order feature vectors and a plurality of high-order feature vectors; performing feature cross processing on the plurality of low-order feature vectors through a factor decomposition machine layer in the preset risk user prediction model to obtain a low-order feature combination set; carrying out nonlinear transformation on the plurality of high-order feature vectors through a hidden layer in the preset risk user prediction model to obtain a high-order feature combination set; performing feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in the preset risk user prediction model, and outputting a driving intention prediction result; performing difference operation on the driving intention prediction result and a preset risk threshold to obtain a target difference value, and judging whether the target difference value is smaller than 0; if the target difference value is smaller than 0, determining that the driving intention prediction result is smaller than a preset risk threshold; and if the target difference value is greater than or equal to 0, determining that the driving intention prediction result is greater than or equal to a preset risk threshold value.

Optionally, in a fourth implementation manner of the first aspect of the present invention, if the driving intention prediction result is smaller than a preset risk threshold, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle reimbursement service order information, and pushing the vehicle reimbursement service order information to a target terminal, where the method includes: if the driving intention prediction result is smaller than a preset risk threshold value, extracting the unique identification and the case identification of the target object from the driving intention prediction request, and acquiring basic information and case compensation information of the target object according to the unique identification and the case identification of the target object; reading a target business processing interface from a preset interface information configuration table according to the unique identifier of the target object and the case identifier, calling and executing the target business processing interface according to the basic information of the target object and the case compensation information, and obtaining vehicle compensation business order information; and pushing the vehicle compensation business order information to a target terminal according to a preset pushing mode, so that the target terminal carries out approval processing on the vehicle compensation business order information.

Optionally, in a fifth implementation manner of the first aspect of the present invention, if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and warning information of a target object according to the driving intention prediction result and the target feature factor data set, and sending the driving intention risk assessment report and the warning information to the target terminal, where the method includes: if the driving intention prediction result is larger than or equal to a preset risk threshold value, carrying out numerical conversion on the driving intention prediction result according to a preset risk conversion rule to obtain a driving intention risk value, and matching a preset risk configuration data table according to the driving intention risk value to obtain a driving risk grade; setting the driving risk level as a target key, and inquiring a preset memory database according to the target key to obtain a target value; when the target value is not a null value, inquiring a preset file directory according to the target value to obtain a target report template and an early warning template; and performing data packaging and conversion processing on the driving intention prediction result and the target characteristic factor data set based on the target report template and the early warning template to obtain a driving intention risk assessment report and early warning information of a target object, and sending the driving intention risk assessment report and the early warning information to the target terminal.

Optionally, in a sixth implementation manner of the first aspect of the present invention, before the receiving a driving intention prediction request of a target object, and acquiring location service trajectory information of the target object, action hotspot trajectory information of the target object, and credit information of the target object according to the driving intention prediction request, the driving intention prediction and processing method further includes: acquiring position service information and action hotspot information which are reported by a mobile terminal corresponding to the target object at regular time through a preset action track acquisition task; respectively carrying out data cleaning and data classification processing on the position service information and the action hotspot information to obtain position service track information of the target object and action hotspot track information of the target object; and acquiring a unique identifier of a target object, and updating the position service track information of the target object and the action hotspot track information of the target object into a preset graph database according to the unique identifier of the target object so as to construct a user portrait of the target object.

A second aspect of the present invention provides a driving intention prediction and processing apparatus, including: the acquisition module is used for receiving a driving intention prediction request of a target object and acquiring position service track information of the target object, action hotspot track information of the target object and credit information of the target object according to the driving intention prediction request; the screening module is used for performing feature analysis screening on the position service track information, the action hotspot track information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating multi-dimensional feature factors related to target object fraud risk behaviors; the prediction module is used for predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not; the processing module is used for calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information and pushing the vehicle compensation service order information to a target terminal if the driving intention prediction result is smaller than a preset risk threshold value; and the generating module is used for generating a driving intention risk assessment report and early warning information of a target object according to the driving intention prediction result and the target characteristic factor data set if the driving intention prediction result is greater than or equal to a preset risk threshold value, and sending the driving intention risk assessment report and the early warning information to the target terminal.

Optionally, in a first implementation manner of the second aspect of the present invention, the obtaining module is specifically configured to: receiving a driving intention prediction request of a target object, distributing the driving intention prediction request and starting a message processing thread; packaging the driving intention prediction request into a target business message through the message processing thread, and caching the target business message into a business message queue; calling a service processing thread corresponding to the service message queue, and performing message analysis on the target service message to obtain a unique identifier and a case identifier of a target object; and inquiring a preset map database according to the unique identifier of the target object and the case identification to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

Optionally, in a second implementation manner of the second aspect of the present invention, the screening module is specifically configured to: reading multidimensional candidate factors from a preset data table, and respectively performing characteristic analysis on the position service track information, the action hotspot track information and the credit information according to the multidimensional candidate factors to obtain a plurality of position service characteristic factors, a plurality of action hotspot characteristic factors and a plurality of credit characteristic factors; and clustering the plurality of location service characteristic factors, the plurality of action hotspot characteristic factors and the plurality of credit characteristic factors respectively through Euclidean distances to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

Optionally, in a third implementation manner of the second aspect of the present invention, the prediction module is specifically configured to: inputting the target characteristic factor data set into a preset risk user prediction model, and performing characteristic extraction on the target characteristic factor data set through an input layer in the preset risk user prediction model to obtain a plurality of sparse characteristics, wherein the preset risk user prediction model is a trained neural network factorization model; obtaining the dimensionality of each sparse feature, embedding each sparse feature into a corresponding dimensionality vector according to the dimensionality of each sparse feature through a neural network layer in the preset risk user prediction model to obtain a plurality of feature embedding vectors, wherein the plurality of feature embedding vectors comprise a plurality of low-order feature vectors and a plurality of high-order feature vectors; performing feature cross processing on the plurality of low-order feature vectors through a factor decomposition machine layer in the preset risk user prediction model to obtain a low-order feature combination set; carrying out nonlinear transformation on the plurality of high-order feature vectors through a hidden layer in the preset risk user prediction model to obtain a high-order feature combination set; performing feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in the preset risk user prediction model, and outputting a driving intention prediction result; performing difference operation on the driving intention prediction result and a preset risk threshold to obtain a target difference value, and judging whether the target difference value is smaller than 0; if the target difference value is smaller than 0, determining that the driving intention prediction result is smaller than a preset risk threshold; and if the target difference value is greater than or equal to 0, determining that the driving intention prediction result is greater than or equal to a preset risk threshold value.

Optionally, in a fourth implementation manner of the second aspect of the present invention, the processing module is specifically configured to: if the driving intention prediction result is smaller than a preset risk threshold value, extracting the unique identification and the case identification of the target object from the driving intention prediction request, and acquiring basic information and case compensation information of the target object according to the unique identification and the case identification of the target object; reading a target business processing interface from a preset interface information configuration table according to the unique identifier of the target object and the case identifier, calling and executing the target business processing interface according to the basic information of the target object and the case compensation information, and obtaining vehicle compensation business order information; and pushing the vehicle compensation business order information to a target terminal according to a preset pushing mode, so that the target terminal carries out approval processing on the vehicle compensation business order information.

Optionally, in a fifth implementation manner of the second aspect of the present invention, the generating module is specifically configured to: if the driving intention prediction result is larger than or equal to a preset risk threshold value, carrying out numerical conversion on the driving intention prediction result according to a preset risk conversion rule to obtain a driving intention risk value, and matching a preset risk configuration data table according to the driving intention risk value to obtain a driving risk grade; setting the driving risk level as a target key, and inquiring a preset memory database according to the target key to obtain a target value; when the target value is not a null value, inquiring a preset file directory according to the target value to obtain a target report template and an early warning template; and performing data packaging and conversion processing on the driving intention prediction result and the target characteristic factor data set based on the target report template and the early warning template to obtain a driving intention risk assessment report and early warning information of a target object, and sending the driving intention risk assessment report and the early warning information to the target terminal.

Optionally, in a sixth implementation manner of the second aspect of the present invention, the driving intention predicting and processing apparatus further includes: the acquisition module is used for acquiring the position service information and the action hotspot information which are reported by the mobile terminal corresponding to the target object at regular time through a preset action track acquisition task; the cleaning module is used for respectively carrying out data cleaning and data classification processing on the position service information and the action hotspot information to obtain position service track information of the target object and action hotspot track information of the target object; and the updating module is used for acquiring the unique identifier of the target object and updating the position service track information of the target object and the action hotspot track information of the target object into a preset map database according to the unique identifier of the target object so as to construct a user portrait of the target object.

A third aspect of the present invention provides a driving intention predicting and processing apparatus including: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the driving intent prediction and processing device to perform the driving intent prediction and processing method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which, when run on a computer, cause the computer to perform the driving intention prediction and processing method described above.

According to the technical scheme provided by the invention, a driving intention prediction request of a target object is received, and position service track information of the target object, action hotspot track information of the target object and credit information of the target object are obtained according to the driving intention prediction request; performing feature analysis and screening on the position service track information, the action hotspot track information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating multi-dimensional feature factors related to target object fraud risk behaviors; predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not; if the driving intention prediction result is smaller than a preset risk threshold value, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information, and pushing the vehicle compensation service order information to a target terminal; and if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of a target object according to the driving intention prediction result and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal. In the embodiment of the invention, a target characteristic factor data set is determined based on position service track information, action hotspot track information and credit information; predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result; and if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information. The accuracy and the recognition efficiency of the driving intention of the user are improved, and the reliability of vehicle loss compensation is improved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a driving intent prediction and processing method according to an embodiment of the invention;

FIG. 2 is a schematic diagram of another embodiment of a driving intent prediction and processing method according to an embodiment of the invention;

FIG. 3 is a schematic diagram of an embodiment of a driving intention prediction and processing device according to an embodiment of the invention;

FIG. 4 is a schematic diagram of another embodiment of the driving intention prediction and processing device according to the embodiment of the invention;

fig. 5 is a schematic diagram of an embodiment of a driving intention prediction and processing device in the embodiment of the invention.

Detailed Description

The embodiment of the invention provides a driving intention prediction and processing method, a driving intention prediction and processing device, a driving intention prediction device and a driving intention processing device, and a storage medium, which are used for predicting a driving intention of a target object through a preset risk user prediction model, improving the accuracy and recognition efficiency of predicting the driving intention of a user, and improving the reliability of vehicle loss compensation.

The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

For convenience of understanding, a detailed flow of an embodiment of the present invention is described below, and referring to fig. 1, an embodiment of a driving intention prediction and processing method according to an embodiment of the present invention includes:

101. and receiving a driving intention prediction request of the target object, and acquiring position service track information of the target object, action hotspot track information of the target object and credit information of the target object according to the driving intention prediction request.

Wherein the driving intention prediction request of the target object comprises a unique identifier of the target object and case identification. Specifically, the server receives a driving intention prediction request of a target object, and the server analyzes the driving intention prediction request of the target object to obtain a unique identifier and a case identifier of the target object; the server judges whether the target object is a vehicle loss compensation object or not according to the unique identifier; if the target object is not the vehicle loss compensation object, the server refuses the driving intention prediction request of the target object; if the target object is a vehicle loss compensation object, the server queries target relation map data from a preset map database according to the unique identifier and case identification of the target object, wherein the target relation map data comprises position service track information of the target object, action hotspot track information of the target object and credit information of the target object.

It is to be understood that the executing subject of the present invention may be a driving intention predicting and processing device, and may also be a terminal or a server, which is not limited herein. The embodiment of the present invention is described by taking a server as an execution subject.

102. And performing characteristic analysis and screening on the position service track information, the action hotspot track information and the credit information to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

It should be noted that the driving intention prediction request of the target object is used for predicting and identifying a purposeful fraudulent object, that is, the server determines whether the driving intention of the target object exists or not, and achieves the purpose of obtaining the vehicle loss compensation by a fraudulent means. The server constructs a characteristic factor rule according to preset historical claim information (including normal claim cases and claim refusal cases), preset position service track information and preset action hotspot track information, wherein the characteristic factor rule comprises a multi-dimensional candidate factor; the server analyzes and screens position service track information of the target object, action hotspot track information of the target object and credit information of the target object respectively according to the multi-dimensional candidate factors to obtain a target characteristic factor data set, and the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors. The multi-dimensional candidate factors may include job-related feature factors, social network blacklist-related feature factors, and user credit-related feature factors.

Further, the user analyzes the characteristics of the purposeful cheating object in advance, determines the characteristic correlation corresponding to the purposeful cheating object, and then the server processes the multi-dimensional candidate factors according to the content of characteristic analysis. It will be appreciated that the feature correlation corresponding to a rogue object may correspond to the following scenario: one is that there is a characteristic correlation with the occupational dimension, and the target objects appearing in a particular location, e.g., a repair shop or 4S shop, which is familiar with car insurance claim settlement methods, result in a high probability of fraud; characteristic relevance to social network dimensions, e.g., the same region or connection of mobile hotspot devices may occur between fraudulent objects of interest; there is a characteristic correlation with the user credit dimension, e.g., users who are overdue, have poor credit motivated to commit fraud. The feature correlation corresponding to the fraudulent object may also include other situations, which are not limited herein.

103. And predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value.

The preset risk user prediction model can be subjected to model training processing in advance according to an actual business scene. The preset risk user prediction model may be a random forest model based on machine learning, a long-short term memory artificial neural network model based on deep learning, or a back propagation neural network model, and an attention mechanism may also be introduced into each model, which is not limited herein. Optionally, the server sets the trained neural network factorization model (i.e., the depefm model) as a preset risk user prediction model; the server performs feature extraction on the target feature factor data set through an input layer in a preset risk user prediction model to obtain a plurality of sparse features; the server embeds each sparse feature into a corresponding dimension vector according to the dimension of each sparse feature through a neural network layer which is based on an attention introducing mechanism and in a preset risk user prediction model to obtain a plurality of feature embedded vectors; the server performs feature cross processing on the feature embedded vectors through a factor decomposition machine layer in a preset risk user prediction model to obtain a low-order feature combination set; the server carries out nonlinear transformation on the multiple feature embedded vectors through a hidden layer in a preset risk user prediction model to obtain a high-order feature combination set; and the server performs feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in the preset risk user prediction model, outputs a driving intention prediction result, and judges whether the driving intention prediction result is smaller than a preset risk threshold value.

104. If the driving intention prediction result is smaller than the preset risk threshold, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information, and pushing the vehicle compensation service order information to a target terminal.

That is, if the driving intention prediction result is smaller than the preset risk threshold, the server determines that the target object is a normal driving behavior, and there is no fraud risk. Specifically, if the driving intention prediction result is smaller than a preset risk threshold, the server performs parameter analysis on the driving intention prediction request to obtain a unique identifier and a case identifier of the target object; the server sets the unique identification and the case identification of the target object as a target index, and queries a preset database according to the target index to obtain basic information and case compensation information of the target object; the server reads a target service processing interface from a preset interface information configuration table according to the unique identifier of the target object and the case identifier; the server sets the basic information and case compensation information of the target object as interface input parameters, calls and executes a target service processing interface based on the interface input parameters, and obtains an interface response result and interface response data; when the interface response result is a preset value, the server acquires the vehicle compensation service order information from the interface response data; the server calls a preset pushing interface according to a preset pushing mode, and pushes the vehicle compensation business order information to the target terminal through the preset pushing interface, so that the target terminal can examine and approve the vehicle compensation business order information.

105. And if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of the target object according to the driving intention prediction result and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal.

That is, if the driving intention prediction result is greater than or equal to the preset risk threshold, the server determines that the target object is an abnormal driving behavior and a fraud risk exists. Specifically, if the driving intention prediction result is greater than or equal to a preset risk threshold, the server determines a driving risk level according to the driving intention prediction result, acquires a target report template and an early warning template according to the driving risk level, statistically determines the driving risk behavior type of the target object based on the target characteristic factor data set, and sends a driving intention risk evaluation report and early warning information of the target object according to the target report template, the early warning template, the driving intention prediction result and the driving risk behavior type to the target terminal. Further, the server stores the driving intention risk assessment report in a blockchain database, which is not limited herein.

In the embodiment of the invention, a target characteristic factor data set is determined based on position service track information, action hotspot track information and credit information; predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result; and if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information. The accuracy and the recognition efficiency of the driving intention of the user are improved, and the reliability of vehicle loss compensation is improved.

Referring to fig. 2, another embodiment of the driving intention prediction and processing method according to the embodiment of the invention includes:

201. and receiving a driving intention prediction request of the target object, distributing the driving intention prediction request and starting a message processing thread.

The driving intention prediction request carries at least one service parameter related to the target object, and each service parameter comprises a parameter name and a parameter value. The driving intention prediction request has a one-to-one correspondence with the message processing thread. Optionally, the server acquires the position service information and the action hotspot information which are reported by the mobile terminal corresponding to the target object at regular time through a preset action track acquisition task; the server respectively carries out data cleaning and data classification processing on the position service information and the action hotspot information to obtain position service track information of the target object and action hotspot track information of the target object; the server acquires the unique identification of the target object, and updates the position service track information of the target object and the action hotspot track information of the target object into a preset graph database according to the unique identification of the target object so as to construct a user portrait of the target object.

It should be noted that the location service information and the action hotspot information are already authorized by the target object and do not relate to the privacy of the target object. The mobile terminal corresponding to the target object is used for acquiring the position information in real time, compressing and encrypting the position information and sending the processed position information to a preset mobile positioning center; the mobile positioning center is used for decompressing, decrypting, matching and data packaging the position information to obtain the position service information and returning the position service information to the mobile terminal. And the mobile terminal reports the position service information and the action hotspot information to the position service information and the action hotspot information at regular time.

202. And packaging the driving intention prediction request into a target business message through a message processing thread, and caching the target business message into a business message queue.

Further, the server acquires a service type corresponding to the driving intention prediction request, encapsulates the driving intention prediction request into a target service message through a message processing thread, caches the target service message into a service message queue corresponding to the service type according to the service type corresponding to the driving intention prediction request, and terminates the message processing thread.

203. And calling a service processing thread corresponding to the service message queue, and performing message analysis on the target service message to obtain the unique identifier and the case identifier of the target object.

Specifically, the server calls a service processing thread corresponding to the service message queue, and performs message analysis on the target service message according to a preset message format to obtain a message analysis result; the server carries out data verification on the message analysis result to obtain a verification result; and when the verification result is successful, the server extracts the unique identification and the case identification of the target object from the verification result. It should be noted that, when the server performs service processing on the target service message, the server may also monitor whether the target service message is abnormal, and if the target service message is abnormal, the server performs abnormal processing on the target service message.

204. And inquiring a preset graph database according to the unique identifier and the case identification of the target object to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

Specifically, the server calculates entity parallel relationship similarity from a preset map database according to the unique identifier and case identification of the target object, and extracts and counts the upper-lower relationship to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

205. And performing characteristic analysis and screening on the position service track information, the action hotspot track information and the credit information to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

It should be noted that the corresponding number of the multidimensional characteristic factors related to the fraud risk behavior of the target object is less than or equal to the corresponding number of the candidate factors. Multidimensional scaling factors associated with target object fraud risk behavior may also include scaling factors associated with professional dimensions, scaling factors associated with social network blacklist dimensions, and scaling factors associated with user credit dimensions. For example, the characteristic factors related to the occupation dimension include places where the target object frequently appears on working days, the number of times the target object appears, and the appearance time. The characteristic factors related to the social network blacklist dimension are used for indicating that a social network blacklist is established based on historical fraud users, analyzing the times and moments of the target object and the preset blacklist users appearing in the same area, and the times, moments and connection types of connecting the same action hotspot. The characteristic factors related to the credit dimension of the user comprise overdue times, overdue types, overdue amount and the like. The multidimensional characteristic factor may also include a vehicle driving track dimensional characteristic of the target object, and may also include other types of factors, which are not limited herein.

Optionally, the server reads the multidimensional candidate factors from a preset data table, and performs feature analysis on the position service track information, the action hotspot track information and the credit information according to the multidimensional candidate factors to obtain a plurality of position service feature factors, a plurality of action hotspot feature factors and a plurality of credit feature factors; the server carries out clustering processing on a plurality of location service characteristic factors (belonging to characteristic factors related to professional dimensions), a plurality of action hotspot characteristic factors (belonging to characteristic factors related to social network blacklist dimensions) and a plurality of credit characteristic factors (namely, characteristic factors related to user credit dimensions) through Euclidean distances to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

206. And predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value.

It can be understood that, when the server is used for model training, the server acquires historical claim case information (including normal claim cases and claim refusal cases) of a plurality of users, location service information of the plurality of users, action hotspot track information of the plurality of users and credit information of the plurality of users, and performs data extraction, data cleaning and data clustering processing on the historical claim case information of the plurality of users, the location service information of the plurality of users, the action hotspot track information of the plurality of users and the credit information of the plurality of users respectively to obtain a training data set; and performing model training based on the training data set to obtain a trained neural network factorization model, and taking the trained neural network factorization model as a preset risk user prediction model. The efficiency and the accuracy of the driving intention prediction are improved, and further a judgment basis can be provided for the reasonable decision and control of the vehicle compensation business.

Optionally, the server inputs the target feature factor data set into a preset risk user prediction model, and performs feature extraction on the target feature factor data set through an input layer in the preset risk user prediction model to obtain a plurality of sparse features, wherein the preset risk user prediction model is a trained neural network factorization machine model; the server acquires the dimensionality of each sparse feature, and embeds each sparse feature into a corresponding dimensionality vector according to the dimensionality of each sparse feature through a neural network layer in a preset risk user prediction model to obtain a plurality of feature embedding vectors, wherein the plurality of feature embedding vectors comprise a plurality of low-order feature vectors and a plurality of high-order feature vectors; the server performs feature cross processing on the low-order feature vectors through a factor decomposition machine layer in a preset risk user prediction model to obtain a low-order feature combination set; the server performs nonlinear transformation on the high-order feature vectors through a hidden layer in a preset risk user prediction model to obtain a high-order feature combination set; the server performs feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in a preset risk user prediction model, and outputs a driving intention prediction result; the server performs difference operation on the driving intention prediction result and a preset risk threshold value to obtain a target difference value, and judges whether the target difference value is smaller than 0; if the target difference value is smaller than 0, the server determines that the driving intention prediction result is smaller than a preset risk threshold value; if the target difference is greater than or equal to 0, the server determines that the driving intention prediction result is greater than or equal to a preset risk threshold. For example, the preset risk threshold is 0.65.

207. If the driving intention prediction result is smaller than the preset risk threshold, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information, and pushing the vehicle compensation service order information to a target terminal.

The vehicle compensation business order information is used for indicating the compensation of vehicle loss on the target object, namely, the target object conforms to the preset vehicle loss compensation regulation. Optionally, if the driving intention prediction result is smaller than a preset risk threshold, the server extracts the unique identifier and the case identifier of the target object from the driving intention prediction request, and obtains the basic information and the case compensation information of the target object according to the unique identifier and the case identifier of the target object; the server reads a target business processing interface from a preset interface information configuration table according to the unique identifier and the case identifier of the target object, calls and executes the target business processing interface according to the basic information and the case compensation information of the target object, and obtains vehicle compensation business order information; the server pushes the vehicle compensation business order information to the target terminal according to a preset pushing mode, so that the target terminal carries out approval processing on the vehicle compensation business order information. It should be noted that the preset push mode may include a message notification mode or an email mode, and is not limited herein. Further, the server receives an approval result and approval information sent by the target terminal; when the approval result is that the approval is passed, the server triggers a preset indemnity service flow according to the vehicle indemnity service order information and sends approval information to the target user; and when the approval result is that the approval is passed, the server predicts and audits the redriving intention of the vehicle compensation business order information according to the approval information.

208. And if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of the target object according to the driving intention prediction result and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal.

The driving intention risk assessment report is a report file with a preset format. Furthermore, the target terminal draws and displays the driving intention risk assessment report and the early warning information to the target personnel, and the target personnel can carry out site investigation according to the driving intention risk assessment report and the early warning information and feed back investigation results. So that the server carries out vehicle loss compensation processing and revises the driving intention risk assessment report on the target object according to the reconnaissance result, and the loss of the vehicle loss compensation is reduced. Optionally, the server updates the driving intention prediction result and the early warning information to a graph database corresponding to the target object according to the unique identifier and the case identifier of the target object, so that the portrait accuracy of the target object is improved. And the server transmits or transmits the driving intention risk assessment report to a preset file system, so that the preset file system performs unified file storage and maintenance on the driving intention risk assessment report.

Optionally, first, if the driving intention prediction result is greater than or equal to the preset risk threshold, the server performs numerical conversion on the driving intention prediction result according to a preset risk conversion rule to obtain a driving intention risk value, and matches a preset risk configuration data table according to the driving intention risk value to obtain a driving risk level, where the driving intention prediction result and the driving intention risk value have a one-to-one correspondence relationship, and the driving intention risk value and the driving risk level have a correspondence relationship. Then, the server sets the driving risk level as a target key, queries a preset memory database according to the target key to obtain a target value, further, judges whether the target value is a null value, and if the target value is the null value, the server generates prompt information and sends the prompt information to the target terminal. Secondly, when the target value is not a null value, the server queries a preset file directory according to the target value to obtain a target report template and an early warning template, namely, when the target value is not a null value, the server determines that the target value is first file path information corresponding to the target report template and second file path information corresponding to the early warning template, and the server reads the target report template and the early warning template from the preset file directory based on the first file path information and the second file path information. The first file path information and the second file path information both have uniqueness. And finally, the server performs data packaging and conversion processing on the driving intention prediction result and the target characteristic factor data set based on the target report template and the early warning template to obtain a driving intention risk assessment report and early warning information of the target object, and sends the driving intention risk assessment report and the early warning information to the target terminal.

With reference to fig. 3, the driving intention prediction and processing method in the embodiment of the present invention is described above, and the driving intention prediction and processing device in the embodiment of the present invention is described below, where one embodiment of the driving intention prediction and processing device in the embodiment of the present invention includes:

the obtaining module 301 is configured to receive a driving intention prediction request of a target object, and obtain location service trajectory information of the target object, action hotspot trajectory information of the target object, and credit information of the target object according to the driving intention prediction request; a screening module 302, configured to perform feature analysis screening on the location service trajectory information, the action hotspot trajectory information, and the credit information to obtain a target feature factor dataset, where the target feature factor dataset is used to indicate a multidimensional feature factor related to a target object fraud risk behavior; the prediction module 303 is configured to predict the driving intention of the target feature factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and determine whether the driving intention prediction result is smaller than a preset risk threshold; the processing module 304 is configured to, if the driving intention prediction result is smaller than the preset risk threshold, call a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle reimbursement service order information, and push the vehicle reimbursement service order information to the target terminal; and a generating module 305, configured to generate a driving intention risk assessment report and early warning information of the target object according to the driving intention prediction result and the target feature factor data set if the driving intention prediction result is greater than or equal to a preset risk threshold, and send the driving intention risk assessment report and the early warning information to the target terminal.

Further, the driving intention risk assessment report is stored in the blockchain database, which is not limited herein.

Referring to fig. 4, another embodiment of the driving intention prediction and processing apparatus according to the embodiment of the present invention includes:

Optionally, the obtaining module 301 may be further specifically configured to:

receiving a driving intention prediction request of a target object, distributing the driving intention prediction request and starting a message processing thread; packaging the driving intention prediction request into a target service message through a message processing thread, and caching the target service message into a service message queue; calling a service processing thread corresponding to the service message queue, and performing message analysis on the target service message to obtain a unique identifier and a case identifier of the target object; and inquiring a preset graph database according to the unique identifier and the case identification of the target object to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

Optionally, the screening module 302 may be further specifically configured to: reading multi-dimensional candidate factors from a preset data table, and respectively carrying out feature analysis on the position service track information, the action hotspot track information and the credit information according to the multi-dimensional candidate factors to obtain a plurality of position service feature factors, a plurality of action hotspot feature factors and a plurality of credit feature factors; and clustering the plurality of position service characteristic factors, the plurality of action hotspot characteristic factors and the plurality of credit characteristic factors respectively through Euclidean distances to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

Optionally, the prediction module 303 may be further specifically configured to: inputting the target characteristic factor data set into a preset risk user prediction model, and performing characteristic extraction on the target characteristic factor data set through an input layer in the preset risk user prediction model to obtain a plurality of sparse characteristics, wherein the preset risk user prediction model is a trained neural network factorization machine model; obtaining the dimensionality of each sparse feature, embedding each sparse feature into a corresponding dimensionality vector according to the dimensionality of each sparse feature through a neural network layer in a preset risk user prediction model to obtain a plurality of feature embedding vectors, wherein the plurality of feature embedding vectors comprise a plurality of low-order feature vectors and a plurality of high-order feature vectors; performing feature cross processing on the plurality of low-order feature vectors through a factor decomposition machine layer in a preset risk user prediction model to obtain a low-order feature combination set; carrying out nonlinear transformation on a plurality of high-order feature vectors through a hidden layer in a preset risk user prediction model to obtain a high-order feature combination set; performing feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in a preset risk user prediction model, and outputting a driving intention prediction result; performing difference operation on the driving intention prediction result and a preset risk threshold to obtain a target difference value, and judging whether the target difference value is smaller than 0; if the target difference value is smaller than 0, determining that the driving intention prediction result is smaller than a preset risk threshold value; and if the target difference value is greater than or equal to 0, determining that the driving intention prediction result is greater than or equal to a preset risk threshold value.

Optionally, the processing module 304 may be further specifically configured to: if the driving intention prediction result is smaller than a preset risk threshold value, extracting the unique identification and the case identification of the target object from the driving intention prediction request, and acquiring basic information and case compensation information of the target object according to the unique identification and the case identification of the target object; reading a target business processing interface from a preset interface information configuration table according to the unique identifier and the case identifier of the target object, calling and executing the target business processing interface according to the basic information and the case compensation information of the target object, and obtaining vehicle compensation business order information; and pushing the vehicle compensation business order information to the target terminal according to a preset pushing mode, so that the target terminal carries out approval processing on the vehicle compensation business order information.

Optionally, the generating module 305 may be further specifically configured to: if the driving intention prediction result is larger than or equal to a preset risk threshold value, carrying out numerical conversion on the driving intention prediction result according to a preset risk conversion rule to obtain a driving intention risk value, and matching a preset risk configuration data table according to the driving intention risk value to obtain a driving risk grade; setting the driving risk level as a target key, and inquiring a preset memory database according to the target key to obtain a target value; when the target value is not a null value, inquiring a preset file directory according to the target value to obtain a target report template and an early warning template; and performing data encapsulation and conversion processing on the driving intention prediction result and the target characteristic factor data set based on the target report template and the early warning template to obtain a driving intention risk assessment report and early warning information of the target object, and sending the driving intention risk assessment report and the early warning information to the target terminal.

Optionally, the driving intention predicting and processing device may further include: the acquisition module 306 is configured to acquire, through a preset behavior trace acquisition task, location service information and action hotspot information periodically reported by a mobile terminal corresponding to a target object; a cleaning module 307, configured to perform data cleaning and data classification processing on the location service information and the action hotspot information respectively to obtain location service track information of the target object and action hotspot track information of the target object; and the updating module 308 is configured to update the location service trajectory information of the target object and the action hotspot trajectory information of the target object to a preset map database according to the unique identifier of the target object, so as to construct a user portrait of the target object.

Fig. 3 and 4 describe the driving intention prediction and processing device in the embodiment of the present invention in detail from the viewpoint of modularization, and the driving intention prediction and processing device in the embodiment of the present invention is described in detail from the viewpoint of hardware processing.

Fig. 5 is a schematic structural diagram of a driving intention prediction and processing device 500 according to an embodiment of the present invention, which may include one or more processors (CPUs) 510 (e.g., one or more processors) and a memory 520, and one or more storage media 530 (e.g., one or more mass storage devices) storing applications 533 or data 532. Memory 520 and storage media 530 may be, among other things, transient or persistent storage. The program stored in the storage medium 530 may include one or more modules (not shown), each of which may include a series of instruction operations in the driving intention prediction and processing apparatus 500. Still further, the processor 510 may be configured to communicate with the storage medium 530, and execute a series of instruction operations in the storage medium 530 on the driving intent prediction and processing device 500.

The driving intent prediction and processing device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input-output interfaces 560, and/or one or more operating systems 531, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the configuration of the driving intention prediction and processing device shown in fig. 5 does not constitute a limitation of the driving intention prediction and processing device, and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components may be used.

The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when run on a computer, cause the computer to perform the steps of the driving intention prediction and processing method.

The invention also provides a driving intention predicting and processing device, which comprises a memory and a processor, wherein the memory stores instructions, and the instructions are executed by the processor, so that the processor executes the steps of the driving intention predicting and processing method in the embodiments.

Further, the computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

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.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A driving intention prediction and processing method is characterized by comprising the following steps:

receiving a driving intention prediction request of a target object, and acquiring position service track information of the target object, action hotspot track information of the target object and credit information of the target object according to the driving intention prediction request;

performing feature analysis and screening on the position service track information, the action hotspot track information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating multi-dimensional feature factors related to target object fraud risk behaviors;

predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not;

if the driving intention prediction result is smaller than a preset risk threshold value, calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information, and pushing the vehicle compensation service order information to a target terminal;

and if the driving intention prediction result is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of a target object according to the driving intention prediction result and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal.

2. The method for predicting and processing the driving intention according to claim 1, wherein the receiving of the request for predicting the driving intention of the target object and the obtaining of the location service trajectory information of the target object, the action hotspot trajectory information of the target object and the credit information of the target object according to the request for predicting the driving intention comprises:

receiving a driving intention prediction request of a target object, distributing the driving intention prediction request and starting a message processing thread;

packaging the driving intention prediction request into a target business message through the message processing thread, and caching the target business message into a business message queue;

calling a service processing thread corresponding to the service message queue, and performing message analysis on the target service message to obtain a unique identifier and a case identifier of a target object;

and inquiring a preset map database according to the unique identifier of the target object and the case identification to obtain the position service track information of the target object, the action hotspot track information of the target object and the credit information of the target object.

3. The method for predicting and processing driving intention according to claim 1, wherein the step of performing feature analysis and screening on the location service trajectory information, the action hotspot trajectory information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating a multi-dimensional feature factor related to target object fraud risk behaviors, and comprises the following steps:

reading multidimensional candidate factors from a preset data table, and respectively performing characteristic analysis on the position service track information, the action hotspot track information and the credit information according to the multidimensional candidate factors to obtain a plurality of position service characteristic factors, a plurality of action hotspot characteristic factors and a plurality of credit characteristic factors;

and clustering the plurality of location service characteristic factors, the plurality of action hotspot characteristic factors and the plurality of credit characteristic factors respectively through Euclidean distances to obtain a target characteristic factor data set, wherein the target characteristic factor data set is used for indicating multi-dimensional characteristic factors related to target object fraud risk behaviors.

4. The method for predicting and processing the driving intention according to claim 1, wherein the predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result, and judging whether the driving intention prediction result is smaller than a preset risk threshold value comprises:

inputting the target characteristic factor data set into a preset risk user prediction model, and performing characteristic extraction on the target characteristic factor data set through an input layer in the preset risk user prediction model to obtain a plurality of sparse characteristics, wherein the preset risk user prediction model is a trained neural network factorization model;

obtaining the dimensionality of each sparse feature, embedding each sparse feature into a corresponding dimensionality vector according to the dimensionality of each sparse feature through a neural network layer in the preset risk user prediction model to obtain a plurality of feature embedding vectors, wherein the plurality of feature embedding vectors comprise a plurality of low-order feature vectors and a plurality of high-order feature vectors;

performing feature cross processing on the plurality of low-order feature vectors through a factor decomposition machine layer in the preset risk user prediction model to obtain a low-order feature combination set;

carrying out nonlinear transformation on the plurality of high-order feature vectors through a hidden layer in the preset risk user prediction model to obtain a high-order feature combination set;

performing feature fusion processing on the low-order feature combination set and the high-order feature combination set through an output layer in the preset risk user prediction model, and outputting a driving intention prediction result;

performing difference operation on the driving intention prediction result and a preset risk threshold to obtain a target difference value, and judging whether the target difference value is smaller than 0;

if the target difference value is smaller than 0, determining that the driving intention prediction result is smaller than a preset risk threshold;

and if the target difference value is greater than or equal to 0, determining that the driving intention prediction result is greater than or equal to a preset risk threshold value.

5. The method for predicting and processing the driving intention according to claim 1, wherein if the result of the prediction of the driving intention is smaller than a preset risk threshold, calling a corresponding target service processing interface according to the request for predicting the driving intention to obtain vehicle reimbursement service order information, and pushing the vehicle reimbursement service order information to a target terminal comprises:

if the driving intention prediction result is smaller than a preset risk threshold value, extracting the unique identification and the case identification of the target object from the driving intention prediction request, and acquiring basic information and case compensation information of the target object according to the unique identification and the case identification of the target object;

reading a target business processing interface from a preset interface information configuration table according to the unique identifier of the target object and the case identifier, calling and executing the target business processing interface according to the basic information of the target object and the case compensation information, and obtaining vehicle compensation business order information;

and pushing the vehicle compensation business order information to a target terminal according to a preset pushing mode, so that the target terminal carries out approval processing on the vehicle compensation business order information.

6. The method for predicting and processing the driving intention according to claim 1, wherein if the predicted result of the driving intention is greater than or equal to a preset risk threshold, generating a driving intention risk assessment report and early warning information of a target object according to the predicted result of the driving intention and the target characteristic factor data set, and sending the driving intention risk assessment report and the early warning information to the target terminal, the method comprises:

if the driving intention prediction result is larger than or equal to a preset risk threshold value, carrying out numerical conversion on the driving intention prediction result according to a preset risk conversion rule to obtain a driving intention risk value, and matching a preset risk configuration data table according to the driving intention risk value to obtain a driving risk grade;

setting the driving risk level as a target key, and inquiring a preset memory database according to the target key to obtain a target value;

when the target value is not a null value, inquiring a preset file directory according to the target value to obtain a target report template and an early warning template;

and performing data packaging and conversion processing on the driving intention prediction result and the target characteristic factor data set based on the target report template and the early warning template to obtain a driving intention risk assessment report and early warning information of a target object, and sending the driving intention risk assessment report and the early warning information to the target terminal.

7. The driving intention prediction and processing method according to any one of claims 1 to 6, wherein before the receiving of the driving intention prediction request of the target object, and the obtaining of the location service trajectory information of the target object, the action hotspot trajectory information of the target object, and the credit information of the target object according to the driving intention prediction request, the driving intention prediction and processing method further comprises:

acquiring position service information and action hotspot information which are reported by a mobile terminal corresponding to the target object at regular time through a preset action track acquisition task;

respectively carrying out data cleaning and data classification processing on the position service information and the action hotspot information to obtain position service track information of the target object and action hotspot track information of the target object;

and acquiring a unique identifier of a target object, and updating the position service track information of the target object and the action hotspot track information of the target object into a preset graph database according to the unique identifier of the target object so as to construct a user portrait of the target object.

8. A driving intention prediction and processing device, characterized by comprising:

the acquisition module is used for receiving a driving intention prediction request of a target object and acquiring position service track information of the target object, action hotspot track information of the target object and credit information of the target object according to the driving intention prediction request;

the screening module is used for performing feature analysis screening on the position service track information, the action hotspot track information and the credit information to obtain a target feature factor data set, wherein the target feature factor data set is used for indicating multi-dimensional feature factors related to target object fraud risk behaviors;

the prediction module is used for predicting the driving intention of the target characteristic factor data set through a preset risk user prediction model to obtain a driving intention prediction result and judging whether the driving intention prediction result is smaller than a preset risk threshold value or not;

the processing module is used for calling a corresponding target service processing interface according to the driving intention prediction request to obtain vehicle compensation service order information and pushing the vehicle compensation service order information to a target terminal if the driving intention prediction result is smaller than a preset risk threshold value;

and the generating module is used for generating a driving intention risk assessment report and early warning information of a target object according to the driving intention prediction result and the target characteristic factor data set if the driving intention prediction result is greater than or equal to a preset risk threshold value, and sending the driving intention risk assessment report and the early warning information to the target terminal.

9. A driving intention prediction and processing apparatus characterized by comprising: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the driving intent prediction and processing device to perform the driving intent prediction and processing method of any of claims 1-7.

10. A computer-readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement the driving intent prediction and processing method of any of claims 1-7.