Intelligent classification system and method for truck driver appeal work orders
By using multi-field keyword matching and time threshold settings, the proportion of urgent work orders and priority coefficients are calculated, and work order levels are dynamically classified. This solves the problems of low efficiency in work order classification and unreasonable resource allocation in existing technologies, and achieves accurate classification and efficient processing of work orders.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- RES INST OF HIGHWAY MINIST OF TRANSPORT
- Filing Date
- 2026-01-26
- Publication Date
- 2026-05-12
AI Technical Summary
Existing work order classification methods rely on manual labeling or simple keyword matching, resulting in low efficiency, susceptibility to subjective factors, difficulty in adapting to the rapid processing needs of massive work orders, unreasonable resource allocation, and easy delays in urgent requests, which affects truck driver satisfaction and industry management efficiency.
By collecting work order data from the past year, performing multi-field keyword matching and setting time thresholds, calculating the proportion of urgent work orders and priority coefficients, and adjusting based on overdue status, we dynamically classify work orders into Category I and Category II, and prioritize their processing according to their processing coefficients.
It enables precise classification and sorting of work orders, ensuring timely response to urgent requests, optimizing resource allocation, and improving work order processing efficiency and industry service quality.
Smart Images
Figure CN122019779A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent classification technology, specifically to an intelligent classification system and method for truck driver complaint work orders. Background Technology
[0002] With the rapid development of the road freight industry, the number of truck drivers has continued to expand, leading to a surge in their demands regarding road access, cargo loading and unloading, vehicle maintenance, and rights protection. This has resulted in an explosive growth in the number of work orders related to these demands. As the core vehicle for handling truck drivers' requests and coordinating solutions, the efficiency of work orders directly impacts the driver's travel experience and the service quality of the freight industry. Currently, freight management platforms need to process massive amounts of work orders from different regions and scenarios, with complex and diverse textual descriptions covering various specific issues. Therefore, quickly and accurately classifying work orders and prioritizing urgent and important requests has become a critical issue that the industry urgently needs to address.
[0003] Existing work order classification methods largely rely on manual labeling or simple keyword matching, lacking quantitative assessment of the urgency and importance of work orders. Manual classification is inefficient, susceptible to subjective factors, and unsuitable for the rapid processing demands of massive volumes of work orders. Traditional keyword matching only achieves basic classification and cannot combine multi-dimensional information such as work order processing time and the proportion of similar work orders to determine priority. This often results in urgent work orders being delayed while secondary work orders consume significant resources, leading to unreasonable allocation of work order processing resources, untimely resolution of core requests, and impacting truck driver satisfaction and industry management efficiency. To optimize work order processing workflows and improve classification accuracy and resource allocation rationality, an intelligent method for classifying truck driver request work orders is urgently needed. Summary of the Invention
[0004] The purpose of this invention is to provide an intelligent classification system and method for truck driver complaint work orders, thereby solving the above-mentioned technical problems.
[0005] The objective of this invention can be achieved through the following technical solutions:
[0006] A method for intelligently classifying truck driver complaint work orders includes the following steps:
[0007] Collect work orders from the past year and obtain the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents a preset number of years.
[0008] The collected work order data is matched with keywords based on the work order request text, including the following steps:
[0009] Get the number of characters N of the keyword, get the i-th to i+N-th fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed.
[0010] As a further aspect of the present invention, it includes:
[0011] Work orders categorized with the same keywords are grouped into the same type of work order. A preset time threshold Ts is set, and the following operations are performed on work orders of the same type:
[0012] Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order.
[0013] Calculate the proportion of urgent work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of urgent work orders and general work orders of the same type, respectively, and M represents the total number of work orders of the same type in the reported work orders.
[0014] As a further aspect of the present invention, it includes:
[0015] Calculate the priority coefficient Y for similar work orders: Y = λ1 × B1 + λ2 × B2 + M / M all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1.
[0016] The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps:
[0017] Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all .
[0018] As a further aspect of the present invention, it includes:
[0019] The detection cycle H is preset and new work orders are periodically acquired. Keyword comparison is performed on the new work orders and they are classified into similar categories. Based on the correction coefficient corresponding to the work orders of the same category, the work orders are classified into Category I and Category II.
[0020] Perform the following operations sequentially on Category I and Category II work orders:
[0021] Obtain the correction coefficient X1 corresponding to a type of work order and the reported duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders.
[0022] As a further aspect of the present invention, it includes:
[0023] Get the maximum number of work orders processed within the current detection cycle H (Max), get the number of Class I work orders and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.
[0024] As a further aspect of the present invention: the method for classifying work orders into Class I and Class II based on the correction coefficient corresponding to the same type of work order includes:
[0025] Work orders of the same type are sorted in descending order of correction coefficient X. The top 50% of work orders of the same type are classified as Class I work orders, and the remaining work orders of the same type are classified as Class II work orders.
[0026] As a further aspect of the present invention: if the number of newly reported work orders within the detection period H is less than or equal to the maximum work order processing volume Max, then processing is performed directly, ignoring the above processing steps.
[0027] A truck driver complaint order intelligent classification system includes:
[0028] Data summary module: Collects work orders from the past year I and obtains the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents the preset number of years.
[0029] The collected work order data is matched with keywords based on the work order request text, including the following steps:
[0030] Get the number of characters N of the keyword, get the i to i+Nth fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed.
[0031] Preprocessing module: Work orders categorized with the same keywords are grouped into similar work orders. A preset time threshold Ts is set, and the following operations are performed on similar work orders:
[0032] Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order.
[0033] Calculate the proportion of emergency work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of emergency work orders and general work orders in the same type of work orders, respectively, and M represents the total number of work orders of the same type in the reported work orders.
[0034] Correction module: Calculates the priority coefficient Y = λ1 × B1 + λ2 × B2 + M / M for similar work orders. all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1.
[0035] The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps:
[0036] Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all ;
[0037] Sorting module: Preset detection period H and periodically obtain newly reported work orders, perform keyword comparison on the newly reported work orders and classify them into similar categories, and classify them into first-class and second-class work orders based on the correction coefficient of similar work orders;
[0038] Perform the following operations sequentially on Category I and Category II work orders:
[0039] Obtain the correction coefficient X1 corresponding to a type of work order and the reporting duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders.
[0040] Processing module: Obtain the maximum number of work orders processed within the current detection cycle H (Max), obtain the number of Class I and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.
[0041] The beneficial effects of this invention are as follows: This method effectively solves the industry pain points of low efficiency in classifying truck driver complaint work orders, ambiguous priority determination, and unreasonable resource allocation, and has significant application value. It achieves accurate initial classification of work orders through multi-field keyword matching, overcoming the limitations of manual and simple keyword matching, and adapting to the need for rapid processing of massive cross-scenario work orders. By combining processing time, the proportion of urgent work orders, and other multi-dimensional data to calculate priority coefficients and correct for overdue situations, it achieves quantitative assessment of work order priority, avoiding delays in urgent requests. Based on the correction coefficient, work orders are classified into levels, and resources are dynamically allocated by combining processing coefficient ranking and cycle processing volume to ensure that high-priority work orders are processed first. It can also flexibly adapt to the volume of work orders in different detection cycles, optimizing resource utilization. The overall process realizes intelligent management of work order classification, sorting, and resource allocation, significantly improving work order processing efficiency and accuracy, ensuring timely response to drivers' core requests, and improving industry service quality and management efficiency. Attached Figure Description
[0042] The invention will now be further described with reference to the accompanying drawings.
[0043] Figure 1 This is a flowchart illustrating an intelligent classification method for truck driver complaint work orders according to the present invention.
[0044] Figure 2 This is a schematic diagram of the structure of an intelligent classification system for truck driver complaint work orders according to the present invention. Detailed Implementation
[0045] 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 embodiments of the present invention, and not all embodiments. 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.
[0046] Please see Figure 1 As shown, this invention is an intelligent classification method for truck driver complaint work orders, comprising the following steps:
[0047] Collect work orders from the past year and obtain the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents a preset number of years.
[0048] The collected work order data is matched with keywords based on the work order request text, including the following steps:
[0049] Get the number of characters N of the keyword, get the i-th to i+N-th fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed.
[0050] It should be noted that the work orders from truck drivers over the past year are collected, and the core data of the corresponding work orders is extracted simultaneously. The work order data must include at least the work order request date r1 and the work order request text. During the collection process, duplicate work orders, invalid work orders, and work orders with severe field omissions must be removed to ensure data quality. Subsequently, the collected valid work order data is classified by keyword matching based on the work order request text. First, the number of characters N of each keyword in the preset keyword library is obtained. Then, the i-th to i+N-th fields in the work order request text are extracted in sequence as comparison fields (i starts from 1 and increments sequentially until the entire text is traversed). The comparison fields are compared one by one with the keywords in the keyword library. If the comparison field is completely consistent with a certain keyword, the work order is classified into the category to which the corresponding keyword belongs. The above field extraction and comparison operations are repeated until the comprehensive comparison of the entire work order request text is completed. Note that the keyword library needs to be updated regularly to adapt to new request types. The comparison process should retain comparison traces for subsequent traceability, and the same work order can correspond to multiple keyword categories to achieve accurate classification.
[0051] A preferred embodiment of this invention includes:
[0052] Work orders categorized with the same keywords are grouped into the same type of work order. A preset time threshold Ts is set, and the following operations are performed on work orders of the same type:
[0053] Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order.
[0054] Calculate the proportion of urgent work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of urgent work orders and general work orders of the same type, respectively, and M represents the total number of work orders of the same type in the reported work orders.
[0055] It is worth noting that all work orders categorized under the same keyword after keyword matching are collectively recorded as similar work orders. Simultaneously, a reasonable time threshold Ts is set, taking into account the industry average processing time for work orders, the urgency level of driver requests, and the management platform's handling capacity. Based on this threshold, the urgency level of each similar work order is determined and data is statistically analyzed. First, the actual processing time T for each similar work order is calculated using the formula T = t2 - t1, where t2 represents the actual completion date of the work order, and t1 corresponds to the work order reporting date. If the calculated processing time T ≥ Ts, it indicates that the work order has a lower timeliness requirement and is recorded as a general work order; if the processing time T < Ts, it indicates that the work order requires a rapid response and is recorded as an urgent work order. Subsequently, the number of emergency work orders (M1), the number of general work orders (M2), and the total number of work orders of the same type (M=M1+M2, excluding canceled and suspended abnormal work orders) are counted separately. Then, the proportion of emergency work orders (B1=M1 / M) and the proportion of general work orders (B2=M2 / M) are calculated. Note that the time threshold (Ts) needs to be optimized and adjusted regularly according to the work order processing efficiency. Date statistics should avoid the impact of holidays, system failures, and other special circumstances. The proportion calculation results should be retained to two decimal places. At the same time, the statistical dimensions and basis should be recorded simultaneously to ensure data traceability and consistent statistical standards for work orders of the same type.
[0056] A preferred embodiment of this invention includes:
[0057] Calculate the priority coefficient Y for similar work orders: Y = λ1 × B1 + λ2 × B2 + M / M all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1.
[0058] The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps:
[0059] Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all .
[0060] Understandably, by setting differentiated weights, the core impact of urgent work orders is highlighted. Simultaneously, by combining this with the proportion of similar work orders within the overall work order list, a scientific and quantitative assessment of work order importance is achieved, overcoming the subjectivity and bias of traditional assessments. Subsequent adjustments to the priority results based on overdue work order data can effectively avoid assessment biases caused by overdue work orders, accurately pinpointing the actual priority level of similar work orders. This provides reliable data support for subsequent work order grading and dynamic allocation of processing resources, helping to prioritize high-value and highly urgent demands, optimize the work order handling priority system, improve overall processing efficiency and resource utilization, and effectively ensure that truck drivers' core demands are responded to in a timely manner.
[0061] A preferred embodiment of this invention includes:
[0062] The detection cycle H is preset and new work orders are periodically acquired. Keyword comparison is performed on the new work orders and they are classified into similar categories. Based on the correction coefficient corresponding to the work orders of the same category, the work orders are classified into Category I and Category II.
[0063] Perform the following operations sequentially on Category I and Category II work orders:
[0064] Obtain the correction coefficient X1 corresponding to a type of work order and the reported duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders.
[0065] It is important to note that by setting a preset detection cycle H and periodically capturing new work orders, and after keyword comparison and classification, work orders are divided into first- and second-class categories based on corresponding correction coefficients. Then, by combining the correction coefficients with the work order reporting time and the detection cycle, a processing coefficient is calculated and the work orders are ranked, enabling dynamic handling of new work orders. This approach prioritizes high-priority work orders, avoids backlog of requests, improves response timeliness, and provides a basis for resource allocation. It is crucial that the detection cycle H be reasonably set to suit the volume of work orders, the keyword comparison criteria remain consistent with those mentioned earlier, the time calculation is accurate to the hour, and the ranking results are updated in real time to ensure smooth integration with subsequent processing procedures.
[0066] A preferred embodiment of this invention includes:
[0067] Get the maximum number of work orders processed within the current detection cycle H (Max), get the number of Class I work orders and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.
[0068] It should be noted that, based on the maximum number of work orders processed within the testing cycle, and according to the proportion of first- and second-class work orders, the work orders with the highest processing coefficients in each category are prioritized for processing. The core significance of this approach is to achieve dynamic and adaptive allocation of work order processing resources. This avoids the skew of resources towards low-priority work orders, ensuring that limited processing capacity is concentrated on serving high-value requests, thereby improving resource utilization and processing efficiency. Simultaneously, it connects with the previously mentioned ranking results, forming a closed-loop process of "tiering – ranking – on-demand allocation," avoiding the backlog of requests or imbalances in processing, ensuring that truck drivers' important requests are addressed quickly, and optimizing the work order processing system and industry service quality.
[0069] In a preferred embodiment of this invention, the method for classifying work orders into Class I and Class II based on the correction coefficient corresponding to work orders of the same category includes:
[0070] Work orders of the same type are sorted in descending order of correction coefficient X. The top 50% of work orders of the same type are classified as Class I work orders, and the remaining work orders of the same type are classified as Class II work orders.
[0071] In a preferred embodiment, if the number of newly reported work orders within the detection period H is less than or equal to the maximum work order processing volume Max, then processing is performed directly, ignoring the above processing steps.
[0072] A truck driver complaint order intelligent classification system includes:
[0073] Data summary module: Collects work orders from the past year I and obtains the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents the preset number of years.
[0074] The collected work order data is matched with keywords based on the work order request text, including the following steps:
[0075] Get the number of characters N of the keyword, get the i to i+Nth fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed.
[0076] Preprocessing module: Work orders categorized with the same keywords are grouped into similar work orders. A preset time threshold Ts is set, and the following operations are performed on similar work orders:
[0077] Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order.
[0078] Calculate the proportion of emergency work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of emergency work orders and general work orders in the same type of work orders, respectively, and M represents the total number of work orders of the same type in the reported work orders.
[0079] Correction module: Calculates the priority coefficient Y = λ1 × B1 + λ2 × B2 + M / M for similar work orders. all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1.
[0080] The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps:
[0081] Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all ;
[0082] Sorting module: Preset detection period H and periodically obtain newly reported work orders, perform keyword comparison on the newly reported work orders and classify them into similar categories, and classify them into first-class and second-class work orders based on the correction coefficient of similar work orders;
[0083] Perform the following operations sequentially on Category I and Category II work orders:
[0084] Obtain the correction coefficient X1 corresponding to a type of work order and the reporting duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders.
[0085] Processing module: Obtain the maximum number of work orders processed within the current detection cycle H (Max), obtain the number of Class I and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.
[0086] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.
Claims
1. A method for intelligent classification of truck driver complaint work orders, characterized in that, Includes the following steps: Collect work orders from the past year and obtain the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents a preset number of years. The collected work order data is matched with keywords based on the work order request text, including the following steps: Get the number of characters N of the keyword, get the i-th to i+N-th fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed.
2. The intelligent classification method for truck driver complaint work orders according to claim 1, characterized in that, include: Work orders categorized with the same keywords are grouped into the same type of work order. A preset time threshold Ts is set, and the following operations are performed on work orders of the same type: Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order. Calculate the proportion of urgent work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of urgent work orders and general work orders of the same type, respectively, and M represents the total number of work orders of the same type in the reported work orders.
3. The intelligent classification method for truck driver complaint work orders according to claim 2, characterized in that, include: Calculate the priority coefficient Y for similar work orders: Y = λ1 × B1 + λ2 × B2 + M / M all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1. The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps: Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all .
4. The intelligent classification method for truck driver complaint work orders according to claim 3, characterized in that, include: The detection cycle H is preset and new work orders are periodically acquired. Keyword comparison is performed on the new work orders and they are classified into similar categories. Based on the correction coefficient corresponding to the work orders of the same category, the work orders are classified into Category I and Category II. Perform the following operations sequentially on Class I and Class II work orders: Obtain the correction coefficient X1 corresponding to a type of work order and the reported duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders.
5. The intelligent classification method for truck driver complaint work orders according to claim 4, characterized in that, include: Get the maximum number of work orders processed within the current detection cycle H (Max), get the number of Class I work orders and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.
6. The intelligent classification method for truck driver complaint work orders according to claim 4, characterized in that, The methods for classifying work orders into Category I and Category II based on the correction coefficient corresponding to the same type of work order include: Work orders of the same type are sorted in descending order of correction coefficient X. The top 50% of work orders of the same type are classified as Class I work orders, and the remaining work orders of the same type are classified as Class II work orders.
7. The intelligent classification method for truck driver complaint work orders according to claim 5, characterized in that, If the number of newly reported work orders within the detection period H is less than or equal to the maximum work order processing volume Max, then the work orders will be processed directly, ignoring the above processing steps.
8. A truck driver complaint work order intelligent classification system, characterized in that, include: Data summary module: Collects work orders from the past year I and obtains the corresponding work order data. The work order data includes the work order request date r1 and the work order request text, where I represents the preset number of years. The collected work order data is matched with keywords based on the work order request text, including the following steps: Get the number of characters N of the keyword, get the i to i+Nth fields of the work order request text and record them as the fields to be compared, compare the fields to be compared with the keyword, if the fields to be compared are the same as the keyword, then the work order belongs to the category corresponding to the keyword, repeat the above operation until the comparison of the entire work order request text is completed. Preprocessing module: Work orders categorized with the same keywords are grouped into similar work orders. A preset time threshold Ts is set, and the following operations are performed on similar work orders: Calculate the processing time T = t2 - t1 for similar work orders, where t2 represents the completion date of the work order. If the processing time T ≥ Ts, it is recorded as a general work order; if the processing time T < Ts, it is recorded as an emergency work order. Calculate the proportion of emergency work orders B1 = M1 / M and the proportion of general work orders B2 = M2 / M, where M1 and M2 represent the number of emergency work orders and general work orders in the same type of work orders, respectively, and M represents the total number of work orders of the same type in the reported work orders. Correction module: Calculates the priority coefficient Y = λ1 × B1 + λ2 × B2 + M / M for similar work orders. all , of which M all λ1 represents the total number of work orders reported in the past year, λ2 represents the preset first weight threshold, and λ1 > λ2 > 1. The priority coefficient Y of similar work orders is adjusted to obtain the adjustment coefficient X, which includes the following steps: Work orders of the same type with a processing time T≥2Ts are classified as overdue work orders. The number of overdue work orders Q is obtained, and the correction factor X=YQ / M is calculated. all ; Sorting module: Preset detection period H and periodically obtain newly reported work orders, perform keyword comparison on the newly reported work orders and classify them into similar categories, and classify them into first-class and second-class work orders based on the correction coefficient of similar work orders; Perform the following operations sequentially on Class I and Class II work orders: Obtain the correction coefficient X1 corresponding to a type of work order and the reporting duration time1 of a type of work order. Calculate the processing coefficient C = X1 × time1 / H. Sort the type of work orders in descending order of processing coefficient C. Repeat the above operation for type of work orders. Processing module: Obtain the maximum number of work orders processed within the current detection cycle H (Max), obtain the number of Class I and Class II work orders (I1 and I2), and prioritize the processing of Class I work orders with the larger processing coefficient C of [I1 / (I1+I2)×Max] pieces and Class II work orders with the larger processing coefficient C of [I2 / (I1+I2)×Max] pieces.