A repair shop lead distribution method, device, equipment and medium

Abstract

The application discloses a repair factory clue distribution method, device, equipment and medium. The repair factory clue distribution method comprises the following steps: based on a target clue to be distributed, obtaining distribution priorities of a plurality of initial repair factories; according to the distribution priorities, selecting a plurality of candidate repair factories from the initial repair factories, and obtaining initial clue amounts and target product values received by the candidate repair factories; according to the initial clue amounts and the target product values, comprehensively evaluating the candidate repair factories to obtain comprehensive adaptation scores of the candidate repair factories; according to the comprehensive adaptation scores, determining a candidate repair factory with the highest comprehensive adaptation score from the candidate repair factories as a target repair factory, and distributing the target clue to the target repair factory. The method improves the accuracy and real-time performance of clue distribution decisions through multi-dimensional evaluation and accurate screening of repair factories, is suitable for a vehicle maintenance and claim settlement scene in the financial insurance industry, and significantly improves the utilization efficiency of repair factory resources.

Description

Technical Field

[0001] This invention relates to the field of financial insurance technology, and in particular to a method, apparatus, equipment and medium for allocating repair shop leads. Background Technology

[0002] In vehicle accident claims, repair shop leads refer to the information about vehicles awaiting repair that insurance companies push to their partner repair shops. The rational allocation of these leads is crucial for improving claims efficiency and optimizing resource allocation. However, the current industry standard employs a crude allocation strategy based on geographical distance or historical frequency of cooperation. This can easily lead to some repair shops exceeding their repair capacity due to lead overload, while other repair shops waste resources due to idle leads, resulting in low accuracy in lead allocation. Summary of the Invention

[0003] This invention provides a repair shop clue allocation method, apparatus, equipment, and medium to solve the technical problem of low accuracy in existing extensive allocation strategies.

[0004] In a first aspect, a method for allocating repair shop leads is provided, comprising the steps of: obtaining the allocation priorities of multiple initial repair shops based on the target leads to be allocated; selecting multiple candidate repair shops from each of the initial repair shops according to the allocation priorities, and obtaining the number of initial leads received and the target output value of each candidate repair shop; comprehensively evaluating each candidate repair shop according to the number of initial leads and the target output value to obtain a comprehensive fit score for each candidate repair shop; determining the candidate repair shop with the highest comprehensive fit score as the target repair shop according to the comprehensive fit score, and allocating the target lead to the target repair shop.

[0005] Secondly, a repair shop lead allocation device is provided, comprising: a priority calculation module, used to obtain the allocation priority of multiple initial repair shops based on the target lead to be allocated; a candidate screening module, used to select multiple candidate repair shops from each of the initial repair shops according to the allocation priority, and obtain the number of initial leads received and the target output value of each candidate repair shop; a comprehensive evaluation module, used to perform a comprehensive evaluation of each candidate repair shop according to each of the initial lead volume and each of the target output values, and obtain a comprehensive fit score for each candidate repair shop; and an allocation decision module, used to determine the candidate repair shop with the highest comprehensive fit score from each of the candidate repair shops as the target repair shop according to each of the comprehensive fit scores, and allocate the target lead to the target repair shop.

[0006] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described repair shop clue allocation method.

[0007] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described repair shop clue allocation method.

[0008] The aforementioned repair shop lead allocation method, apparatus, equipment, and medium implement a solution whereby the repair shop lead allocation method includes the following steps: based on the target lead to be allocated, obtaining the allocation priorities of multiple initial repair shops; selecting multiple candidate repair shops from each initial repair shop according to the allocation priorities, and obtaining the initial lead volume and target output value received by each candidate repair shop; comprehensively evaluating each candidate repair shop based on each initial lead volume and each target output value to obtain a comprehensive fit score for each candidate repair shop; and determining the candidate repair shop with the highest comprehensive fit score as the target repair shop based on each comprehensive fit score, and allocating the target lead to the target repair shop. In this invention, for the repair shop lead allocation problem, the initial repair shops are first prioritized based on the target lead, then the initial repair shops with higher allocation priorities are selected as candidate repair shops, and a comprehensive evaluation is performed based on the initial lead volume and target output value received by each candidate repair shop to finally determine the most suitable target repair shop for lead allocation. This allocation method, by comprehensively evaluating the overall suitability scores of each candidate repair shop, can more accurately determine their repair capabilities and output value achievement capacity, thereby improving the accuracy and real-time nature of allocation decisions. Furthermore, the solution also boasts good scalability and adaptability, capable of handling complex scenarios such as sudden accident peaks or regional fluctuations in repair demand, providing insurance companies with a more efficient and intelligent repair shop lead allocation solution. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of an application environment for a repair shop clue allocation method according to an embodiment of the present invention; Figure 2 This is a flowchart of a repair shop clue allocation method according to an embodiment of the present invention; Figure 3This is a detailed flowchart of step S1 in the repair shop clue allocation method in one embodiment of the present invention; Figure 4 This is a detailed flowchart of step S3 in the repair shop clue allocation method in one embodiment of the present invention; Figure 5 This is a detailed flowchart of step S31 in the repair shop clue allocation method in one embodiment of the present invention; Figure 6 This is a detailed flowchart of step S32 in the repair shop clue allocation method in one embodiment of the present invention; Figure 7 This is a detailed flowchart of step S33 in the repair shop clue allocation method in one embodiment of the present invention; Figure 8 This is a schematic diagram of a repair shop clue distribution device according to an embodiment of the present invention; Figure 9 This is a schematic diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] The repair shop lead allocation method provided in this embodiment of the invention can be applied to, for example... Figure 1 The application environment is shown. Specifically, this repair shop lead allocation method is applied in a repair shop lead allocation system, which includes, for example, the following... Figure 1 The diagram illustrates a client and server that communicate over a network to achieve precise allocation and efficient scheduling of repair shop resources. The client, also known as the user terminal, is the program that provides local services to the customer, corresponding to the server. The client can be installed on, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0013] In one embodiment, such as Figure 2 As shown, a repair shop clue allocation method is provided, which is applied to... Figure 1 Taking the server in the example, the following steps are included: Step S1: Based on the target clues to be assigned, obtain the assignment priority of multiple initial repair shops.

[0014] It should be noted that target leads are accident repair request information generated from accident information uploaded by relevant users such as insurance claims departments, traffic management departments, or vehicle owners through the client. This information includes key attributes such as accident location, vehicle type, extent of damage, and the user's urgency level. Initial repair shops refer to those registered in the system with basic order-taking qualifications. These qualifications cover at least two core business areas, including bodywork and painting, electromechanical repair, and accident damage assessment. Their qualification review is automatically completed by the system and is regularly updated with data from the industry regulatory platform. The qualification review process strictly follows relevant regulations to ensure the authenticity, validity, and traceability of the review results. Allocation priority is a comprehensive weight value dynamically calculated based on multi-dimensional data such as the initial repair shop's geographical location, historical capacity, real-time availability, professional qualifications, and user ratings. This ensures that lead allocation adheres to both the principle of geographical proximity and considers service capacity and quality stability.

[0015] In this embodiment, by obtaining the accident location associated with the target clue, the relative distance between each initial repair shop and the accident location can be obtained. Then, based on the type and relative distance of each initial repair shop, each initial repair shop can be grouped to obtain a set of repair shops with different allocation priorities, thereby obtaining the allocation priority of each initial repair shop.

[0016] Specifically, such as Figure 3 As shown, step S1 includes the following sub-steps: Step S11: Based on the target clues, obtain the relative distance between each initial repair shop and the accident site.

[0017] It should be noted that relative distance refers to the weighted composite value of the straight-line distance from the initial repair shop to the accident site and the actual travel distance of the traffic network. It takes into account both geographical proximity and dynamic factors such as real-time traffic conditions, road grade and congestion index to ensure that the calculation results of relative distance truly reflect the timeliness of response.

[0018] In this embodiment, multi-dimensional data such as the geographical location, repair capabilities, real-time order volume, and historical response time of each initial repair shop are pre-stored in the server database. By calling the Geographic Information System (GIS) interface, the spatial straight-line distance between each initial repair shop and the accident site can be calculated in real time. Combined with the latest traffic flow speed and congestion heat map obtained from the dynamic traffic API, a weighted relative distance is generated, thus providing a quantitative basis for subsequent clue allocation that is both spatially reasonable and temporally feasible. This weighted model is not a static threshold determination, but is dynamically calibrated according to day and night, weather conditions, and regional events, making each clue push a precise response to the complexity of reality—distance is not only a physical scale, but also a time contract for the delivery of trust.

[0019] For example, after a car owner uploads accident information through the client, corresponding accident repair request information, i.e. target clues, is generated. Then, based on the target clues, the spatiotemporal coordinates of the accident are analyzed. By calling the corresponding Geographic Information System (GIS) interface, the spatial straight-line distance between each initial repair shop and the accident location is calculated in real time. Real-time traffic condition weight coefficients are injected simultaneously to generate a dynamically weighted relative distance, thereby accurately anchoring the optimal service touchpoint within the response radius, so that every emergency request can find the most resilient support point in the spatiotemporal tension.

[0020] Step S12: Based on the type of each initial repair shop and the corresponding relative distance, group the initial repair shops according to the corresponding distance division strategy to obtain multiple repair shop sets.

[0021] It should be noted that, due to the different service positioning, technical capabilities, and response radii of different types of initial repair shops, their service radii and response capabilities vary significantly. Therefore, a type-aware distance partitioning strategy needs to be constructed to ensure that the grouping logic meets the realistic constraints of type adaptability and service response. The distance partitioning strategy is closely related to the type of initial repair shop. Different distance thresholds are set according to the type of initial repair shop (e.g., 4S shops, comprehensive repair shops). Different distance thresholds are used for different types of initial repair shops. Based on the comparison between the relative distance and the corresponding distance threshold, different types of initial repair shops are assigned to corresponding repair shop sets. The repair shop set is the result of grouping initial repair shops based on their type and the relative distance between the initial repair shop and the accident location. Essentially, it is a concrete mapping of the tension between service capability and spatiotemporal constraints, providing a structured and computable hierarchical decision-making basis for subsequent clue allocation.

[0022] In this embodiment, for 4S stores, those within a 30-kilometer radius (relative distance ≤ 30 kilometers) are grouped into a core set of repair shops, with the corresponding set index denoted as 0. For 4S stores at longer distances (relative distance > 30 kilometers), they are grouped incrementally in units of 10 kilometers, with the corresponding set indices denoted as 1, 2, 3, and so on. The set index of the repair shop set corresponding to each 4S store can be represented using the first set index formula: ; Similarly, for comprehensive repair shops, using 5 km and 30 km as double boundaries, nearby comprehensive repair shops (relative distance ≤ 5 km) are grouped into the core repair shop set, with a set index of 0; medium-distance comprehensive repair shops (5 km < relative distance ≤ 30 km) are grouped in increments of 5 km, with corresponding set indices of 1, 2, 3... up to 5; long-distance comprehensive repair shops (relative distance > 30 km) are grouped in increments of 10 km, with corresponding set indices starting from 6. The set index of the repair shop set corresponding to a comprehensive repair shop can be represented using the second set index formula: ; Where d represents the relative distance (in kilometers) between the initial repair shop and the accident site. This indicates the rounding up operation; I_4S is the set index of the set of repair shops to which the initial repair shop belongs, used to identify the distance grouping level of the 4S shop; I_ZX is the set index of the set of repair shops to which the initial repair shop belongs, used to identify the distance grouping level of the comprehensive repair shop.

[0023] In this embodiment, by combining the brand characteristics and geographical distance of each initial repair shop, differentiated dynamic grouping rules are formulated. Then, through the core group division and distance segmentation incremental strategy, the dynamic set index of each repair shop is obtained in real time, realizing the fine-grained grouping of different types of repair shops, thereby providing a grouping basis that fits the actual operational needs for subsequent repair shop resource scheduling.

[0024] In other embodiments, the distance threshold and grouping granularity can be dynamically adjusted according to the complexity of the actual business scenario. For example, in high-density urban areas, the distance threshold for 4S stores can be lowered to 20 kilometers, and the distance range for comprehensive repair shops can be compressed to 3-25 kilometers, with real-time correction of the traffic congestion index. In low-density counties, the threshold can be appropriately relaxed and the coverage radius of a single group can be expanded to improve scheduling robustness. At the same time, weather influencing factors and holiday fluctuation coefficients can be introduced to dynamically compensate the grouping strategy in the spatiotemporal dimension, thereby accurately adapting to the spatiotemporal dynamic characteristics of different regions and more closely meeting actual operational needs.

[0025] Step S13: Determine the allocation priority of each initial repair shop based on the set of repair shops.

[0026] In this embodiment, each set of repair shops corresponds to an allocation priority. For each set of repair shops, its set index directly determines the allocation priority of each initial repair shop within that set. The set of repair shops corresponding to set index 0 has the highest allocation priority, and the allocation priority of the corresponding set of repair shops decreases with each increment of the set index. This allocation priority decay mechanism ensures that the initial repair shops in the core set of repair shops that are closest to the accident site and have the strongest response capability are always prioritized; while as the set index increases, although the response time of distant repair shops is slightly lower, they become a powerful supplement when resources are scarce.

[0027] For example, core repair shops with a set index of 0 (such as the Zhongshan Road 4S store in City A and the Chengxi Comprehensive Repair Center in District B) automatically receive the highest priority of 99 points in the system. Secondary core repair shops with a set index of 1 (such as the Jiefang Road 4S store in County C and the Dongjiao Comprehensive Repair Station in Town D) receive a priority of 95 points, while peripheral repair shops with a set index of 2 (such as the various stations on the auto repair street in Township E) receive 90 points. This hierarchical system uses 5-point increments to construct a priority scale that respects both the rigid constraints of geography and accommodates the elastic value of services, providing quantifiable, traceable, and optimizable underlying support for subsequent intelligent scheduling decisions.

[0028] Furthermore, this priority decay mechanism can also be compatible with brand attribute weights. For example, under the same set index, 4S stores have a higher priority than general repair shops, ensuring both timeliness and repair quality. For example, among repair shops under set index 1, the 4S store in City A's High-tech Zone has a priority of 97 points due to its dual qualifications of direct supply of original parts and technical certification, while the nearby general repair shop in the B District auto parts market maintains a baseline of 95 points. This ensures a more reliable and standardized repair service experience for accident car owners while maintaining timely response. This refined priority design not only injects business interpretability into the algorithm scheduling but also achieves a dual alignment of geographical proximity and service quality at the system level, laying a solid data foundation and logical support for subsequent multi-objective collaborative optimization.

[0029] In other embodiments, each initial repair shop corresponds to an allocation priority, and each set of repair shops corresponds to an allocation priority range. The initial repair shops within each set are sorted according to their specific allocation priority values, thereby achieving secondary sorting based on brand level within the same distance group. For example, within the allocation priority range of set index 2, the allocation priority values ​​for all 4S shops are [95, 99], while for general repair shops it is [85, 89], ensuring that resources of different qualities at the same distance can still be scheduled in an orderly manner according to quality gradients.

[0030] In this application, the above-mentioned operation in step S1 performs geographical clustering of each initial repair shop based on the relative distance between the accident location associated with the target clue and each initial repair shop, generating several repair shop sets. Based on the geographical clustering results, a unique set index is assigned to each repair shop set, and the allocation priority value of each initial repair shop is dynamically generated based on the index value and brand attribute. This completes the semantic mapping from spatial relationship to computable scheduling weight, enabling subsequent intelligent scheduling decisions based on the allocation priority value. Thus, the abstract geographical and brand semantics are transformed into scheduling variables that can participate in real-time calculations.

[0031] Step S2: Based on the allocation priority, select multiple candidate repair shops from each initial repair shop, and obtain the initial clue volume and target output value received by each candidate repair shop.

[0032] It should be noted that the initial lead count refers to the total number of leads currently pending or being processed by the corresponding candidate repair shop. This represents the total number of accident assessments, repair requests, and other tasks that the repair shop is currently handling or awaiting allocation, reflecting its immediate workload. The target revenue is the repair service revenue indicator that the corresponding candidate repair shop must achieve within the assessment period. This target is typically set quarterly or annually and dynamically linked to regional market capacity, historical performance capabilities, and strategic development positioning.

[0033] In this embodiment, firstly, based on each set of repair shops, the set of repair shops with the highest allocation priority is selected as the candidate repair shop set; then, each initial repair shop in the corresponding candidate repair shop set is extracted as a candidate repair shop, and the initial thread count and target output value of each candidate repair shop in the corresponding candidate repair shop set are obtained. This not only allows resource scheduling to focus on the core group with high responsiveness and high quality, but also accurately depicts the actual capacity reserve of each candidate repair shop through real-time perception of both thread count and target output value.

[0034] For example, the set of repair shops with the highest allocation priority is first selected as the candidate repair shop set. Specifically, the set of repair shops with index 0 is chosen as the candidate repair shop set because it has the highest overall allocation priority. All initial repair shops within this set are included in the first round of candidate pools, covering six highly qualified institutions, including the 4S flagship store in City A's Economic Development Zone, the original factory authorized center in City B's High-tech Zone, and the intelligent fast repair station in City C's Bonded Zone. These candidate repair shops have an average allocation priority of 96.3 points, far exceeding the average of the entire set (89.7 points), and their historical lead fulfillment rate is consistently above 98.2%, providing a high-quality, low-risk resource base for subsequent precise scheduling. Specifically, the 4S flagship store in City A's Economic Development Zone currently has 23 initial leads with a target output value of 12.8 million yuan; the original factory authorized center in City B's High-tech Zone currently has 19 initial leads with a target output value of 11.2 million yuan; and the intelligent fast repair station in City C's Bonded Zone currently has 31 initial leads with a target output value of 8.6 million yuan.

[0035] First, the core repair shops with the highest allocation priority are selected as the candidate repair shop set. Then, all initial repair shops in the core repair shop set are considered as candidate repair shops. Next, the initial lead volume and target output value received by each candidate repair shop are obtained. Then, based on the initial lead volume and target output value of each candidate repair shop, the specific operations of subsequent steps S3 and S4 can be performed to select the target repair shop from the core repair shop set.

[0036] Furthermore, the server dynamically collects and aggregates the initial lead volume and target output data received by each candidate repair shop by monitoring their work order system and financial interface in real time, ensuring the timeliness and accuracy of load assessment; at the same time, a sliding time window mechanism is introduced to eliminate lead records that have not responded or have been cancelled due to timeout, avoiding misjudgment of capacity.

[0037] In this application, the above operation in step S2 uses the allocation priority determined in step S1 to select the set of repair shops with the highest allocation priority from each set of repair shops as the candidate set of repair shops, so as to obtain the initial clue quantity and target output value of each candidate repair shop in the candidate set of repair shops, and provide solid, dynamic and verifiable data support for subsequent steps S3 and S4.

[0038] It should be noted that the candidate repair shops in the core repair shop set may be operating at full capacity. When performing subsequent steps S3 and S4, it may be impossible to select a target repair shop that can be allocated target leads. In this case, the candidate range needs to be expanded level by level to the repair shop sets corresponding to adjacent set indices, incorporating the repair shop set with the second highest allocation priority. Correspondingly, step S2 needs to be returned to obtain the repair shop set with the second highest allocation priority as the new candidate repair shop set. The initial repair shops in this new candidate repair shop set are then used as the new candidate repair shops, and the initial lead quantity and target output value of each candidate repair shop are simultaneously obtained. Steps S3 and S4 are then re-executed. This dynamic expansion mechanism not only ensures the continuity of lead allocation but also, in the event of sudden peaks or local resource depletion, leverages the collaborative response capabilities of multi-level repair shop sets to achieve smooth load transfer and elastic capacity release.

[0039] For example, when determining the core repair shop set with set index 0 as the candidate repair shop set, if all candidate repair shops in this set are already operating at full capacity, the candidate range needs to be expanded to the repair shop set with set index 1, that is, the repair shop set with the second highest priority is assigned, and step S2 is executed again, using all the initial repair shops in this set as new candidate repair shops, and the initial thread volume and target output value of each new candidate repair shop are obtained simultaneously, and then steps S3 and S4 are executed again.

[0040] Step S3: Based on the initial clue quantity and target output value, conduct a comprehensive evaluation of each candidate repair shop to obtain the comprehensive fit score of each candidate repair shop.

[0041] It should be noted that the comprehensive evaluation is based on the matching degree between the real-time lead load and target output value of each candidate repair shop. It quantifies and scores the degree of fit between each candidate repair shop and the target lead, thus providing a quantifiable, rankable, and traceable decision-making basis for the subsequent accurate selection of target repair shops. The comprehensive fit score is a quantified result of the matching between each candidate repair shop and the target lead. The higher the comprehensive fit score, the stronger the corresponding candidate repair shop's ability to undertake the target lead and the better its response quality under the dual constraints of current load and capacity target.

[0042] In this embodiment, for the set of candidate repair shops selected in step S2, a corresponding clue probability distribution matrix and output probability distribution vector are constructed. By matching the clue probability distribution matrix with the corresponding output probability distribution vector, a comprehensive fit score for each candidate repair shop is generated. This comprehensive fit score can ensure that the evaluation results reflect both the current resource occupancy status and predict future performance stability.

[0043] Specifically, such as Figure 4 As shown, step S3 includes the following sub-steps: Step S31: Based on the initial number of clues, obtain the clue probability distribution matrix corresponding to the candidate repair shop set.

[0044] In this embodiment, the clue distribution matrix corresponding to the candidate repair shop set is first obtained based on the initial clue quantity of each candidate repair shop in the corresponding candidate repair shop set. Then, the original clue distribution matrix is ​​normalized to a probability distribution form using the proportion of each initial clue quantity to the total clue quantity of the corresponding candidate repair shop set as weights, so that the sum of each row is 1. This statistically represents the relative carrying capacity of each candidate repair shop for clue resources. This transformation not only eliminates the difference in absolute numerical dimensions, but also makes the clue carrying capacity of candidate repair shops of different sizes comparable and quantifiable. The normalized clue probability distribution matrix accurately depicts the resource elasticity space of each candidate repair shop under the current load, providing a unified benchmark for subsequent dynamic coupling with output targets.

[0045] Specifically, such as Figure 5 As shown, step S31 includes the following sub-steps: Step S311: Based on the initial number of clues, obtain the clue distribution matrix corresponding to the candidate repair shop set.

[0046] In this embodiment, for the candidate repair shop set, it is set to include a total of One candidate repair shop, and The candidate repair shops are ranked according to their relative distance, numbered from 1 to... Correspondingly, the first The initial number of leads for each candidate repair shop is denoted as: Therefore, it can be constructed based on this. × The clue distribution matrix P is dimensional, and the formula for calculating the clue distribution matrix P is: ; Where I is × An identity matrix of dimension 1, representing "assuming a new target clue is assigned to the 1st dimension". "One candidate repair shop", S is... Vector of pushed clues Extended diagonal matrix (i.e.) off-diagonal elements ), Vector of pushed clues This indicates the distribution of leads received by each candidate repair shop.

[0047] For example, if the candidate repair shop set has 3 candidate repair shops, their initial thread counts are respectively , , The corresponding vector of pushed leads is , Corresponding clue distribution matrix Among them, the clue distribution matrix The first line assumes that after assigning the target clue to the first candidate repair shop, the distribution of the number of clues received by each candidate repair shop is as follows: Become The second line assumes that after assigning the target clue to the second candidate repair shop, the distribution of the number of clues received by each candidate repair shop is as follows: Become The same applies to the third line, so I won't repeat it here.

[0048] Step S312: Perform a clue probability transformation on the clue distribution matrix to obtain the clue probability distribution matrix.

[0049] In this embodiment, the clue distribution matrix P corresponding to the candidate repair shop set is normalized row by row so that the sum of the elements in each row is 1, thereby obtaining the clue probability distribution matrix. Specifically, for each element in the clue distribution matrix P, dividing it by the sum of all elements in that row yields the corresponding clue probability distribution matrix. Corresponding clue probability distribution matrix The Line number Column elements can be represented as: in, ;No. Line number Column elements This indicates that the target clue will be assigned to the first... When choosing a repair shop, the first The relative clue probability of the nth candidate repair shop, i.e., the probability of the nth... Under the assumption of a certain type of allocation, the first The weighting of each candidate repair shop in receiving leads; the larger this value, the higher the weighting. The higher the probability that a candidate repair shop will be selected under this allocation scheme, the more likely it is to be given priority.

[0050] For example, for the clue distribution matrix P The first row of the clue distribution matrix P contains elements (2, 2, 5), whose sum is 9. Therefore, the normalized first row contains elements (2 / 9, 2 / 9, 5 / 9). This process is repeated for each row of the clue distribution matrix P, ultimately yielding the clue probability distribution matrix. .

[0051] In this application, the lead probability distribution matrix can intuitively reflect the lead carrying probability of each candidate repair shop after being assumed to be assigned a new target lead, providing a quantitative basis for subsequent output probability distribution matching. This transformation method not only considers the current lead load of each candidate repair shop, but also predicts the possible trend of future lead allocation in the form of probability, making the comprehensive evaluation more scientific and comprehensive.

[0052] Step S32: Based on each target output value, obtain the output value probability distribution vector corresponding to the set of candidate repair shops.

[0053] In this embodiment, firstly, based on the target output value corresponding to each candidate repair shop in the corresponding candidate repair shop set, the target output value distribution vector corresponding to the candidate repair shop set is obtained. Then, using the proportion of each target output value to the total target output value of the corresponding candidate repair shop set as weights, the original target output value distribution vector is normalized into a probability distribution form, so that the sum of all elements is 1. This statistically represents the relative contribution potential of each candidate repair shop to the target output value resources. This transformation not only eliminates the difference in absolute numerical dimensions but also highlights the structural differences in the output value achievement capabilities of different repair shops.

[0054] Specifically, such as Figure 6 As shown, step S32 includes the following sub-steps: Step S321: Based on each target output value, obtain the target output value distribution vector corresponding to the candidate repair shop set.

[0055] In this embodiment, for the candidate repair shop set, a first... The target output value of each candidate repair shop is denoted as: Therefore, it can be constructed based on this. Initial output distribution vector of dimension At the same time, introduce the smallest positive number (To avoid the influence of zero values ​​on calculations), the initial output distribution vector... Add very small positive numbers Then, the target output value distribution vector corresponding to the candidate repair shop set can be obtained. ( (A vector consisting entirely of 1s).

[0056] For example, if the candidate repair shop set has 3 candidate repair shops, their target output values ​​are respectively , , Minimal positive number Taking 0.01, the corresponding initial output distribution vector is: , Add very small positive numbers The target output distribution vector is as follows .

[0057] Step S322: Perform output probability transformation on the target output distribution vector to obtain the output probability distribution vector.

[0058] In this embodiment, the target output value distribution vector corresponding to the candidate repair shop set is... Each element in the vector is divided by the sum of all elements in the target vector, and then normalized so that the sum of all elements is 1, thus obtaining the output probability distribution vector Q. Specifically, for the target output distribution vector... Divide each element in by The sum of all elements in the vector yields the corresponding output probability distribution vector Q, and the nth element of the output probability distribution vector Q is the nth element of the vector Q. Each element can be represented as: ; in, ; Indicates the first The relative contribution probability of each candidate repair shop to the target output value; the larger this value, the more significant the contribution. The stronger the candidate repair shop's ability to achieve its target output value, the greater its responsibility.

[0059] For example, for the target output distribution vector The total target output value is Then the first element of the normalized output probability distribution vector Q is The second element is The third element is .

[0060] In this application, the output value probability distribution vector can intuitively reflect the relative contribution capacity of each candidate repair shop to the target output value resources, providing a key basis for subsequent matching with the clue probability distribution matrix and calculating the comprehensive fit score. This transformation method fully considers the target output value of each candidate repair shop and reflects its potential ability to achieve output value in a probabilistic form, making the comprehensive evaluation more in line with actual business needs. At the same time, it can also ensure that subsequent allocation decisions always revolve around the output value target, which helps to dynamically match the actual output value of each candidate repair shop with the target output value, improves the target output value achievement rate, and enhances the overall operational efficiency of claims business.

[0061] Furthermore, the introduction of the clue probability distribution matrix and the output probability distribution vector effectively avoids the problem of some candidate repair shops being overloaded with clues while other candidate repair shops are idle, thus achieving a balanced allocation of repair shop resources and improving the overall resource utilization rate.

[0062] Step S33: Match the clue probability distribution matrix with the output probability distribution vector to obtain the comprehensive fit score.

[0063] In this embodiment, the clue probability distribution vector of each candidate repair shop is first obtained based on the clue probability distribution matrix of the corresponding candidate repair shop set. Then, the distance between the vector and the output probability distribution vector obtained in step S32 is evaluated, the matching degree between the two is calculated, and the comprehensive fit score of each candidate repair shop is obtained.

[0064] Specifically, such as Figure 7 As shown, step S33 includes the following sub-steps: Step S331: Based on the clue probability distribution matrix, obtain the clue probability distribution vector corresponding to each candidate repair shop in the candidate repair shop set.

[0065] Among them, the first of the clue probability distribution matrix The line indicates that the target clue is assigned to the first When there are n candidate repair shops, the probability distribution of clues corresponding to each candidate repair shop is as follows: The probability distribution vector of clues corresponding to each candidate repair shop reflects the relative probability of each candidate repair shop accepting the target clue under this allocation scheme.

[0066] For example, the probability distribution matrix of clues In the diagram, the first row vector (2 / 9, 2 / 9, 5 / 9) represents the probability distribution of each candidate repair shop accepting the target clue when it is assigned to the first candidate repair shop. Specifically, when the first candidate repair shop accepts the target clue, the probabilities of the first, second, and third candidate repair shops accepting the target clue are 2 / 9, 2 / 9, and 5 / 9, respectively. Similarly, the second row vector (1 / 9, 3 / 9, 5 / 9) represents the probability distribution of each candidate repair shop accepting the target clue when it is assigned to the second candidate repair shop. Specifically, when the second candidate repair shop accepts the target clue, the probabilities of the first, second, and third candidate repair shops accepting the target clue are 1 / 9, 3 / 9, and 5 / 9, respectively. The vector (1 / 9, 2 / 9, 6 / 9) in the third row represents the probability distribution vector of the clues for the third candidate repair shop when the target clue is assigned to the third candidate repair shop. Specifically, when the third candidate repair shop accepts the target clue, the probabilities of the first, second, and third candidate repair shops accepting the target clue are 1 / 9, 2 / 9, and 6 / 9, respectively.

[0067] Step S332: Perform distance evaluation between the probability distribution vector of each clue and the corresponding output probability distribution vector to obtain the comprehensive fit score of each candidate repair shop in the corresponding candidate repair shop set.

[0068] It should be noted that distance assessment is a mathematical method for quantifying the differences between vectors. It can be implemented in the form of Euclidean distance, cosine similarity, or KL divergence to obtain the degree of matching between the lead distribution and output value distribution of each candidate repair shop, reflecting the level of synergistic matching between the lead acceptance capacity and output value contribution potential of the candidate repair shop.

[0069] In this embodiment, the Euclidean distance is first calculated between the clue probability distribution vector and the output probability distribution vector corresponding to each candidate repair shop in the candidate repair shop set to obtain the vector distance of the corresponding candidate repair shop. Based on the corresponding vector distance, each vector distance can be normalized and inversely mapped to a comprehensive fit score in the range of 0 to 1. The higher the comprehensive fit score, the more priority is given to recommending and assigning it.

[0070] Among them, the The probability distribution vector of clues for each candidate repair shop (clue probability distribution matrix) The The vector distance between the row vector and the output probability distribution vector Q can be expressed as: ; in, ; Indicates the first The vector distances corresponding to each candidate repair shop Indicates the first The relative contribution probability of each candidate repair shop to the target output value; This indicates that the target clue will be assigned to the first... When choosing a repair shop, the first The relative clue probability of each candidate repair shop.

[0071] Furthermore, normalization can be performed based on the reciprocal of the vector distance to obtain the comprehensive fit score for each candidate repair shop. The higher the comprehensive fit score, the more closely the clue probability distribution vector and the output probability distribution vector of the candidate repair shop match, and the more significant its synergistic advantages in clue conversion efficiency and output stability.

[0072] For example, if the candidate repair shop set has 3 candidate repair shops, and the corresponding output probability distribution vector Q is (0.2, 0.35, 0.45), when the target clue is assigned to the first candidate repair shop, its corresponding clue probability distribution vector is (2 / 9, 2 / 9, 5 / 9). Then the vector distance corresponding to the first candidate repair shop is... After normalization, the overall fit score is approximately 0.816. When the target clue is assigned to the second candidate repair shop, its corresponding clue probability distribution vector is (1 / 9, 3 / 9, 5 / 9). Therefore, the vector distance to the second candidate repair shop is... After normalization, the overall fit score is approximately 0.842. When the target clue is assigned to the third candidate repair shop, its corresponding clue probability distribution vector is (1 / 9, 2 / 9, 6 / 9). Therefore, the vector distance to the third candidate repair shop is... After normalization, the overall fit score is approximately 0.789.

[0073] It should be noted that the smaller the vector distance between candidate repair shops, the stronger the synergy between the candidate repair shop's ability to accept clues and its ability to contribute to the target output value, and the higher the corresponding comprehensive fit score. Therefore, when the clue probability distribution vector and the output value probability distribution vector of a candidate repair shop are highly coincident, the corresponding vector distance approaches 0, and the normalized comprehensive fit score is close to 1, making it the optimal allocation target. Conversely, if the two distributions are significantly different, the comprehensive fit score will rapidly decay.

[0074] In other embodiments, for other candidate repair shop sets, the same method is used to construct the corresponding clue probability distribution matrix and output probability distribution vector, and the comprehensive fit score of the candidate repair shops in each candidate repair shop set is calculated accordingly. This will not be elaborated further here.

[0075] In other embodiments, other methods can be used to obtain the comprehensive fit score of each candidate repair shop, depending on actual needs. For example, distance metrics based on cosine similarity or KL divergence can be used. The former focuses more on the consistency of distribution direction and is suitable for scenarios where clues and output structure trends are highly correlated; the latter is sensitive to subtle shifts in probability quality distribution and is suitable for scenarios where distribution patterns differ significantly but directional trends are consistent.

[0076] In other embodiments, the initial lead volume and target output value of each candidate repair shop can be linearly weighted and summed using weighted coefficients. Different weighting coefficients α and β are assigned to each candidate repair shop's lead taking capacity and output value contribution capacity, respectively, directly obtaining the comprehensive fit score for each candidate repair shop. This makes the comprehensive fit score more accurately reflect business priorities. For example, when focusing on rapid conversion, α > β; when emphasizing long-term output value, β > α, thereby achieving dynamic weighted allocation of lead resources under the guidance of business strategy.

[0077] In this application, step S3 utilizes the initial lead volume and target output value of each candidate repair shop and uses the corresponding comprehensive evaluation method to obtain the comprehensive fit score of each candidate repair shop, so as to evaluate the degree of synergistic matching of each candidate repair shop in terms of lead acceptance capacity and target output value contribution capacity, and provide a quantitative basis for the subsequent selection and allocation of target repair shops.

[0078] Step S4: Based on the overall fit score, determine the candidate repair shop with the highest overall fit score from all candidate repair shops as the target repair shop, and assign the target clue to the target repair shop.

[0079] It should be noted that the target repair shop was selected from all candidate repair shops, and is the one with the highest comprehensive suitability score. It is the only candidate repair shop selected to perform this target lead allocation task. The target repair shop has the dual advantages of sufficient lead acceptance capacity and target output value contribution capacity, ensuring that after receiving the target lead, the target repair shop has the hard power of real-time response, resource scheduling and process initiation, and can rely on its technical qualifications, historical performance rate and customer reputation to achieve repair quality standards, controllable timeliness, and customer satisfaction ranking among the top in the region.

[0080] In this embodiment, the comprehensive matching scores of the obtained candidate repair shops are sorted, and the candidate repair shop with the highest comprehensive matching score is selected as the target repair shop. The target clue is then assigned to the target repair shop. After the target clue is assigned, the clue status is updated to "assigned" and the target repair shop is included in the maintenance resource scheduling system for task assignment, so as to ensure that the target repair shop can respond in a timely manner and complete the corresponding vehicle maintenance task with high quality.

[0081] For example, if there are 3 candidate repair shops in the candidate repair shop set, the first candidate repair shop has a comprehensive suitability score of 0.816, the second candidate repair shop has a comprehensive suitability score of 0.842, and the third candidate repair shop has a comprehensive suitability score of 0.789. These 3 candidate repair shops are sorted by their comprehensive suitability scores, and the second candidate repair shop (with the highest comprehensive suitability score of 0.842) is identified as the target repair shop. The target lead is then assigned to this target repair shop, and the lead status is updated to "assigned". The work order assignment process is initiated in the maintenance resource scheduling system, and a structured work order containing key information such as vehicle damage details, damage assessment amount, and estimated repair cycle is pushed in real time. A service timeliness countdown is automatically triggered to ensure that the target repair shop completes the order acceptance response within 2 hours and initiates the first repair procedure within 24 hours, allowing the corresponding target repair shop to promptly accept and execute the accident repair task for the corresponding vehicle.

[0082] In this application, ranking the candidate repair shops by their comprehensive suitability scores and selecting the one with the highest score as the target repair shop is a key decision point for accurate lead allocation. Selecting the candidate repair shop that best matches the business objectives from among the candidate repair shops ensures that the target leads are allocated to the optimal candidate repair shop. This ensures that the lead allocation strategy is both accurate and feasible under the guidance of business objectives. Consequently, after receiving the target leads, the target repair shop can promptly accept and execute the accident repair task for the corresponding vehicle.

[0083] It should be noted that the above method embodiment mainly adapts, sorts and optimizes each candidate repair shop in the candidate repair shop set. However, in actual business, it is also necessary to take into account the regional coverage balance and the timeliness of emergency response. In step S4, it may not be possible to select a target repair shop that can be assigned to the target clue. Therefore, it is also necessary to conduct a dynamic capability assessment of the target repair shop to obtain the dynamic capability score of the target repair shop. Then, based on the dynamic capability score, it can be determined whether the target repair shop has the ability to accept calls in real time.

[0084] Therefore, this method also includes the following steps: performing a dynamic capability assessment on the target repair shop to obtain its dynamic capability score. If its dynamic capability score is lower than the score threshold, a secondary matching mechanism is triggered, returning to step S2, re-selecting and allocating a set of candidate repair shops with the second highest priority from the remaining set of repair shops, and obtaining the initial lead quantity and target output value corresponding to each candidate repair shop in the set of candidate repair shops with the second highest priority allocation. Then, based on the updated initial lead quantity and target output value data, steps S3 to S4 are re-executed until the optimal target repair shop with the qualified dynamic capability is selected, ensuring that the lead allocation of the target repair shop is both in line with its lead taking capacity and matches its output value contribution potential.

[0085] For example, when determining the core repair shop set with set index 0 as the candidate repair shop set, if the dynamic capability scores of all candidate repair shops in the core repair shop set are lower than the score threshold, the set index is automatically incremented to 1, the suboptimal core repair shop set is activated, and the filtering logic of step S2 is continued to be executed. The repair shop set with set index 1 is reselected as the candidate repair shop set, and steps S2 to S4 are continued until the optimal target repair shop with the dynamic capability meets the standard is found. This ensures that the clue allocation strategy still has strong robustness and scalability in complex and ever-changing business scenarios.

[0086] In this application, the above mechanism can ensure that when each candidate repair shop in the first-best candidate repair shop set is limited by dynamic factors such as real-time capacity, workstation occupancy or technician scheduling, it can seamlessly switch to the second-best candidate repair shop set, taking into account both allocation stability and business continuity. It ensures that the lead allocation is always anchored on the main repair shop with both excellent real-time capacity and fulfillment capabilities, significantly improving resource conversion efficiency and being able to cope with complex scenarios such as sudden accident peaks or regional maintenance demand fluctuations.

[0087] In this application, the above process combines dynamic grouping with probability distribution matching, which can quickly respond to changes in multi-dimensional factors such as geographical distance, brand differences, and target output value, and achieve the selection of the optimal target repair shop in real time. This shortens the decision time for clue allocation and meets the needs of efficiency and real-time processing in accident claims scenarios.

[0088] In summary, the above method first obtains the allocation priorities of multiple initial repair shops based on the target leads to be assigned. Then, it filters candidate repair shops according to the allocation priorities and obtains the number of initial leads and target output value received by each candidate repair shop. Next, it comprehensively evaluates each candidate repair shop based on the number of initial leads and target output value to obtain a comprehensive suitability score for each candidate repair shop. Finally, it sorts and locks the optimal target repair shop based on the comprehensive suitability score, thus completing the accurate distribution of target leads. This method first prioritizes initial repair shops based on target leads, then selects the initial repair shops with higher allocation priorities as candidate repair shops. By dynamically linking the lead status of repair shops with target output value constraints, it achieves multi-dimensional evaluation and accurate screening of repair shops, significantly improving the scientific nature and effectiveness of lead allocation. This is conducive to the efficient scheduling of subsequent repair shop resources and the closed-loop optimization of claims services. It can cope with complex scenarios such as sudden accident peaks or regional maintenance demand fluctuations, providing insurance companies with a more efficient and intelligent repair shop lead allocation solution.

[0089] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0090] In one embodiment, a repair shop lead allocation device is provided, which corresponds one-to-one with the repair shop lead allocation method described in the above embodiments. For example... Figure 8 As shown, the repair shop lead allocation device includes a priority calculation module 101, a candidate screening module 102, a comprehensive evaluation module 103, and an allocation decision module 104. Detailed descriptions of each functional module are as follows: Priority calculation module 101 is used to obtain the allocation priority of multiple initial repair shops based on the target clues to be assigned.

[0091] The candidate screening module 102 is used to select multiple candidate repair shops from each initial repair shop according to the allocation priority, and to obtain the initial clue volume and target output value received by each candidate repair shop.

[0092] The comprehensive evaluation module 103 is used to conduct a comprehensive evaluation of each candidate repair shop based on the initial clue quantity and each target output value, and obtain the comprehensive suitability score of each candidate repair shop.

[0093] The allocation decision module 104 is used to determine the candidate repair shop with the highest comprehensive adaptation score from among the candidate repair shops based on the comprehensive adaptation score, and to allocate the target clue to the target repair shop.

[0094] Specific limitations regarding the repair shop lead allocation device can be found in the limitations of the repair shop lead allocation method described above, and will not be repeated here. Each module in the aforementioned repair shop lead allocation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0095] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a repair shop lead allocation method.

[0096] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the repair shop lead allocation method described in the above embodiment, for example... Figure 2 As shown in S1-S4, or Figures 3 to 7 As shown, to avoid repetition, it will not be described again here. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in this embodiment of the repair shop clue allocation device, for example... Figure 8 The functions of the priority calculation module 101, candidate screening module 102, comprehensive evaluation module 103, and allocation decision module 104 shown are not described again here to avoid repetition.

[0097] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program implements the repair shop clue allocation method described in the above embodiment, for example... Figure 2 As shown in S1-S4, or Figures 3 to 7 As shown, to avoid repetition, it will not be described again here. Alternatively, when the computer program is executed by the processor, it implements the functions of each module / unit in this embodiment of the repair shop clue allocation device, for example... Figure 8 The functions of the priority calculation module 101, candidate screening module 102, comprehensive evaluation module 103, and allocation decision module 104 shown are not described again here to avoid repetition. The computer-readable storage medium can be non-volatile or volatile.

[0098] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0099] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0100] It should be noted that any AI models, software tools, or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with the knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0101] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for allocating repair shop leads, characterized in that, Including the following steps: Based on the target clues to be assigned, obtain the assignment priority of multiple initial repair shops; According to the allocation priority, multiple candidate repair shops are selected from each of the initial repair shops, and the initial lead volume and target output value received by each candidate repair shop are obtained. Based on the initial clue quantity and target output value, a comprehensive evaluation is conducted on each candidate repair shop to obtain a comprehensive fit score for each candidate repair shop. Based on the comprehensive adaptation scores, the candidate repair shop with the highest comprehensive adaptation score is determined as the target repair shop, and the target clue is assigned to the target repair shop.

2. The repair shop clue allocation method as described in claim 1, characterized in that, The process of obtaining the allocation priority of multiple initial repair shops based on the target clues to be assigned includes: Based on the target clues, obtain the relative distances between each initial repair shop and the accident site; Based on the type of each initial repair shop and the corresponding relative distance, the initial repair shops are grouped according to the corresponding distance division strategy to obtain multiple repair shop sets; The allocation priority of each initial repair shop is determined based on each set of repair shops.

3. The repair shop lead allocation method as described in claim 2, characterized in that, Each set of repair shops corresponds to an allocation priority. The step of selecting multiple candidate repair shops from each of the initial repair shops according to the allocation priority, and obtaining the initial lead volume and target output value received by each candidate repair shop, includes: Based on each set of repair shops, the set of repair shops with the highest corresponding allocation priority is selected as the candidate set of repair shops; Extract each initial repair shop from the candidate repair shop set as a candidate repair shop, and obtain the initial clue quantity and target output value of each candidate repair shop.

4. The repair shop clue allocation method as described in claim 3, characterized in that, The step of comprehensively evaluating each candidate repair shop based on the initial clue quantity and the target output value to obtain a comprehensive suitability score for each candidate repair shop includes: Based on the initial number of clues, obtain the clue probability distribution matrix corresponding to the candidate repair shop set; Based on each of the target output values, obtain the output value probability distribution vector corresponding to the candidate repair shop set; The clue probability distribution matrix is ​​matched with the output probability distribution vector to obtain the comprehensive adaptation score for each.

5. The repair shop lead allocation method as described in claim 4, characterized in that, The step of obtaining the clue probability distribution matrix corresponding to the candidate repair shop set based on each of the initial clue quantities includes: Based on the initial clue quantity, obtain the clue distribution matrix corresponding to the candidate repair shop set; The clue distribution matrix is ​​transformed into a clue probability distribution matrix.

6. The repair shop lead allocation method as described in claim 4, characterized in that, The step of obtaining the output probability distribution vector corresponding to the candidate repair shop set based on each of the target output values ​​includes: Based on each of the target output values, obtain the target output value distribution vector corresponding to the candidate repair shop set; The output probability distribution vector is obtained by performing an output probability transformation on the target output distribution vector.

7. The repair shop lead allocation method as described in claim 4, characterized in that, The step of matching the clue probability distribution matrix with the output probability distribution vector to obtain the comprehensive fit score includes: Based on the clue probability distribution matrix, obtain the clue probability distribution vector corresponding to each candidate repair shop in the candidate repair shop set; The distance between each clue probability distribution vector and the corresponding output probability distribution vector is evaluated to obtain the comprehensive fit score of each candidate repair shop in the candidate repair shop set.

8. A repair shop clue distribution device, characterized in that, include: The priority calculation module is used to obtain the allocation priority of multiple initial repair shops based on the target clues to be assigned; The candidate screening module is used to select multiple candidate repair shops from each of the initial repair shops according to the allocation priority, and to obtain the initial lead volume and target output value received by each of the candidate repair shops. The comprehensive evaluation module is used to comprehensively evaluate each candidate repair shop based on the initial clue quantity and the target output value, and obtain the comprehensive fit score of each candidate repair shop. The allocation decision module is used to determine the candidate repair shop with the highest comprehensive adaptation score from the candidate repair shops as the target repair shop based on the comprehensive adaptation score, and to allocate the target clue to the target repair shop.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the repair shop clue allocation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the repair shop clue allocation method as described in any one of claims 1 to 7.

Images