Certificate reading task management method based on automatic classification of card reading work orders

Abstract

The invention discloses a reading and urging task management method based on automatic classification of reading and urging work orders, relates to the technical field of corpus recognition and classification, and solves the problems that in the prior art, the processing of reading and urging appeal work orders depends on manpower, the efficiency is low, and the processing efficiency is high. And when an automatic identification technology is adopted, multi-dimensional analysis is not carried out in combination with user characteristics, so that the identification precision is not high. According to the method, the accepted content of the historical copying and urging work order can be analyzed based on the copying and urging attribution framework, the regular expression technology is adopted to clean the text and remove words such as common words which are not helpful for classification, accurate word segmentation of complaint content is achieved by building a word segmentation professional word bank and combining with the Chinese natural language word segmentation technology, and the complaint content can be accurately classified. The method comprises the following steps: performing classification modeling on the reading and urging work orders by using an SVM algorithm, calculating the optimal classification of texts, and finally establishing an alarm threshold value based on each classification and attribution to realize management of batch reading and urging appeal work orders.

Description

Technical Field

[0001] The present invention relates to the technical field of corpus recognition and classification, and in particular to a copying and reminder task management method based on automatic classification of copying and reminder work orders. Background Art

[0002] In the digital age, the importance of power grids or power companies receiving and processing work orders for demand collection and reminder requests has become increasingly prominent. Work order data not only records the electricity consumption demands of power customers, but also directly reflects the customer's satisfaction and demand for power supply services. Accurate and efficient processing of these work orders is crucial to improving customer service quality, optimizing power resource allocation, and preventing and resolving power failures. Through natural language processing technology, key information can be extracted from the work order text and converted into structured data for easy analysis and processing, thereby achieving rapid identification and processing of fault types. In addition, the formulation of intelligent response and maintenance response plans also depends on the accurate understanding and classification of work order data.

[0003] Although natural language processing technology plays an important role in work order analysis, the processing of work orders for copying and demanding requests is still mostly manual and inefficient. Manual processing is not only time-consuming and labor-intensive, but also easily affected by subjective factors, resulting in limited consistency and accuracy of processing results. In addition, even if automatic recognition technology is used, it is often not combined with user characteristics for multi-dimensional analysis, which limits the improvement of recognition accuracy.

[0004] In view of this, a copying and reminder task management method based on automatic classification of copying and reminder work orders is needed. Summary of the invention

[0005] In view of the problems in the prior art of processing copy and reminder request work orders, the efficiency is low when relying on manual labor, and the recognition accuracy is low when using automatic recognition technology without combining user characteristics for multi-dimensional analysis. The present invention provides a copy and reminder task management method based on automatic classification of copy and reminder work orders, which can analyze the acceptance content of historical copy and reminder work orders based on the copy and reminder attribution framework, use regular expression technology to clean the text, remove common words and other words that are not helpful for classification, and achieve accurate word segmentation of complaint content by building a professional word segmentation vocabulary combined with Chinese natural language word segmentation technology. The SVM algorithm is used to classify and model the copy and reminder work orders, calculate the best classification of the text, and finally establish an alarm threshold based on each classification and attribution to achieve batch copy and reminder request work order management. The specific technical scheme is as follows: A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders comprises the following steps: Collect historical copy request work order texts and perform data preprocessing, and use the existing user load data and electricity consumption data in the marketing system as user features; Build the attribution classification label of copy reminder guidance, build a professional vocabulary, and further build a professional classification vocabulary, and classify the copy reminder work orders based on the professional classification vocabulary; The types of work orders and the reasons for the work orders are used as labels to mark the work orders, and the SVM algorithm is used to train the model, so that the types of work orders and the reasons for the work orders can be identified through the features of the work orders and users. Establish an alarm quantity threshold for each type of copy-reminder work order and each copy-reminder work order attribution, collect copy-reminder work orders and user characteristics within the collection period, automatically identify the copy-reminder work order type and copy-reminder work order attribution, and then perform statistics, and issue an alarm based on the alarm quantity threshold.

[0006] Preferably, the types of work orders include complaint work orders, report work orders, opinion and suggestion work orders, business application work orders and inquiry and consultation work orders; the causes of the power grid copy and reminder work orders include service attitude problems, business ability problems, power quality problems, power application problems, frequent power outages and voltage abnormalities, and power outage and restoration and fault repair problems; the professional classification vocabulary includes a complaint work word library, a report work word library, an opinion and suggestion work word library, a business application work word library, a business application work word library and an inquiry and consultation work word library.

[0007] Preferably, the construction process of the professional classification vocabulary is as follows: First, determine the business vocabulary, which is a collection of all professional terms in this profession. ,in They correspond to R professional classification word libraries, including complaint word library, report word library, opinion and suggestion word library, business application word library, business application word library and query and consultation word library; The professional classification vocabulary is the words in the set of professional words that appear in the corresponding copying and reminder request work order type, and these words are the professional classification vocabulary of this type of copying and reminder request work order.

[0008] Preferably, the data preprocessing includes: Punctuation removal: remove all symbols to reduce the size of training data; Stop word and rare word removal: Collect prepositions and greetings in the complaint work orders to create a stop word library, and remove the corresponding words in the complaint text based on the stop word library to achieve the purpose of text cleaning; rare words are words that only exist in a few work orders. Replace rare words with other synonyms to increase word frequency, or directly delete them to improve model iteration efficiency; Disambiguation conversion: Use TF-IDF as the weight to judge the frequency of some homophones in the text description and complete the disambiguation conversion; Idiom Removal: Remove common texts from different text categories to improve model iteration efficiency.

[0009] Preferably, the classification of the copy reminder work order is specifically as follows: extract the words contained in the business vocabulary in the copy reminder work order, and then compare the proportion of these words in various professional classification vocabulary libraries, and take the copy reminder work order type corresponding to the professional classification vocabulary library with the largest proportion as the final output copy reminder work order type.

[0010] Preferably, the specific steps of using the SVM algorithm to train the model are as follows: S01: Data collection: historical copy request work order text, and the user load data and electricity consumption data already in the marketing system as user characteristics; S02: Data preprocessing: Clean the collected data, including removing missing values, processing outliers, and unifying data formats; S01: Feature Engineering: Extract features that are helpful for classification, including the user's historical work order records, the urgency of the work order, and seasonal factors, and normalize the features to eliminate the impact of different dimensions; S04: Model training: Use the SVM algorithm to train the model; S05: Model parameter adjustment: Use cross-validation and grid search to optimize model parameters, including penalty coefficient C and kernel function parameter gamma; S06: Model evaluation: Evaluate the performance of the model on the test set, including using accuracy, recall, and F1 score; observe the confusion matrix to understand the performance of the model on different categories; S07: Model application: Apply the trained model to new work order data to automatically identify the type and attribution of work orders; S08: Result interpretation and feedback: Explain the model’s prediction results, collect feedback and continuously optimize the model based on the model’s performance in actual applications.

[0011] Preferably, the method further comprises the following steps: By calling the API interface of the classification model, the classification results are obtained, and the corresponding work order type interface of the front-end business system is adaptively modified to ultimately achieve the purpose of model result query, screening and quality inspection.

[0012] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned method for managing copying and reminder tasks for automatic classification of copying and reminder work orders.

[0013] A processor is used to run a program, wherein when the program is run, the method for managing copying and reminder tasks for automatically classifying copying and reminder work orders as described above is executed.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention collects historical copying reminder request work order texts and performs data preprocessing, and uses the existing user load data and electricity consumption data in the marketing system as user features; then builds copying reminder guidance attribution classification labels, builds a professional vocabulary and further builds a professional classification vocabulary, and classifies the copying reminder work orders based on the professional classification vocabulary; uses the copying reminder work order type and the copying reminder work order attribution as labels to mark the copying reminder work order, and uses the SVM algorithm to train the model, so that the copying reminder work order type and the copying reminder work order attribution can be finally identified through the copying reminder work order and user features; finally, establishes the alarm quantity threshold of each copying reminder work order attribution of each type of copying reminder work order, collects the copying reminder work order and user features within the period, automatically identifies the copying reminder work order type and the copying reminder work order attribution, and then performs statistics, and issues an alarm based on the alarm quantity threshold, so as to realize the management of batch copying reminder request work orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0016] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0019] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0020] It should be further understood that the term "and / or" used in the specification and appended claims of the present invention refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0021] In one embodiment of the present invention, a management method for collection and reminder tasks based on automatic classification of collection and reminder work orders is provided, as Figure 1 shown, including the following steps: Step 1: Collect historical 95598 collection and reminder demand work order texts, and use the existing user load data and power consumption data in the marketing system as user characteristics; Among them, the collection and reminder demand work order text includes acceptance content, first-level classification, second-level classification, third-level classification, handling situation, etc. Professional vocabulary can be directly obtained from the acceptance content, first-level classification, second-level classification, and third-level classification parts, and the attribution of the collection and reminder demand work order can be directly obtained from the handling situation part.

[0022] Step 2: Perform data preprocessing.

[0023] The data preprocessing includes: (1) Remove punctuation marks Because it does not add any additional information to the text data. Therefore, all the removed symbols will help reduce the size of the training data and improve the model training performance.

[0024] (2) Remove stop words and rare words Stop words refer to words whose information is not helpful for the classification of the model and may even cause certain misleading, and should be deleted from the text data. In this project, prepositions, greetings and other words in the collection and reminder demand work orders are collected to create a stop word library, and the corresponding words in the complaint text are removed according to the stop word library to achieve the purpose of cleaning the text.

[0025] Rare words refer to words that only exist in a few work orders. Due to their rarity, low-frequency words have extremely limited improvement on the model performance. The rare words can be replaced with other synonyms to increase the word frequency, or directly deleted to improve the model iteration efficiency.

[0026] (3) Disambiguation conversion Some homophonic misspelled words in the text description are discriminated by frequency with TF-IDF as the weight to complete the disambiguation conversion. For example: disambiguation processing and conversion of words such as "accompanying childbirth" and "compensation length" are carried out, and they are converted to "compensation".

[0027] (4) Remove idiomatic expressions Texts like "Customer called to report" and "Please ask the power company to handle it in time" are common texts in different text categories. They have high word frequency but are not helpful for classification. They should be removed to improve the efficiency of model iteration.

[0028] Step 3: Based on the attribution results of historical copying and reminder work orders, design and build copying and reminder guidance attribution classification labels, and build a professional vocabulary and further build a professional classification vocabulary.

[0029] The attribution results are obtained through the data generated during the processing of historical copy and reminder work orders, that is, the reasons for the formation of copy and reminder work orders, and the feedback from the corresponding executors. The types of copy and reminder work orders are as follows: Complaint work orders: including complaints about power outages, power supply quality, and grid construction, etc.

[0030] Reporting work orders: transfer of clues related to work style problems, and reports from the public on damage and endangerment of power facilities.

[0031] Opinion (suggestion) work orders: Customers’ opinions and suggestions on power supply services, facility maintenance, etc.

[0032] Business application work orders: involving power engineering design consultation, electricity safety inspection, power equipment procurement, etc.

[0033] Inquiry and consultation tickets: Customers’ questions or demands regarding electricity services, such as electricity bill inquiries, power outage notifications, meter installation, electricity consumption recommendations, etc. The attribution of power grid work orders mainly includes the following aspects: Service attitude issues: Meter readers and payment collectors have poor service attitudes, question problems raised by customers, and have a bad attitude when handling complaints.

[0034] Business capability issues: Complaints regarding business capabilities in power supply services and facility maintenance, such as issuing power outage information when repairing lines and equipment.

[0035] Power quality issues: including voltage instability, voltage deviation, frequency deviation and other power supply quality issues.

[0036] Electricity installation problems: After the user applies for electricity installation, the power supply company does not accept the application or does not complete the meter installation and power connection within the prescribed time limit.

[0037] Frequent power outages and abnormal voltage problems: The weak distribution network structure and overloaded lines in some areas lead to frequent power outages, low voltage and other power quality problems.

[0038] Issues with power outage and restoration and fault repair: Some power supply companies do not have standardized power outage and restoration management, and fault repairs are not timely.

[0039] Determine the type and attribution of the copy and reminder work order through historical processing feedback, and mark the type and attribution of each collected copy and reminder work order.

[0040] In this embodiment, the above-mentioned construction of professional word library can be constructed by using text segmentation technology and expert business experience to construct a copy reminder business word library to determine the model support. The professional classification word library includes a complaint word library, a report word library, an opinion and suggestion word library, a business application word library, a business application word library and a query and consultation word library.

[0041] The formation process of the professional classification word library is as follows: first determine the business word library, which is a collection of all professional words in this profession (related to power grid copying and reminder). The professional classification word library is the words in the collection of professional words that appear in the corresponding copying and reminder request work order type. These words are the professional classification word library of this type of copying and reminder request work order. For example: the collection of all professional words ,in They correspond to R professional classification word libraries, including complaint word library, report word library, suggestion word library, business application word library, business application word library and query and consultation word library; among them, complaint word library and report word library , , and so on, the professional terms contained in various lexicons may have overlapping parts or independent parts.

[0042] Step 4: Classify the copy reminder work orders based on the professional classification vocabulary.

[0043] The classification standard is: extract professional terms from the copy and reminder work order, that is, extract the words in the business vocabulary contained in the copy and reminder work order, and then compare the proportion of these words in various professional classification vocabulary, and take the copy and reminder work order type corresponding to the professional classification vocabulary with the largest proportion as the final output copy and reminder work order type.

[0044] Step 5: Use the type of work order and the attribution of the work order as labels to mark the work order, and use the SVM algorithm to train the model, so that the type of work order and the attribution of the work order can be identified through the work order and user characteristics.

[0045] The details are as follows: S01: Data collection: Collect grid work order data, including user characteristics (such as user age, gender, electricity usage category, etc.), work order types (such as complaints, reports, opinions, etc.), and work order attribution (such as service attitude issues, business capability issues, etc.).

[0046] S02: Data preprocessing: Clean the collected data, including removing missing values, handling outliers, unifying data formats, etc. Encode categorical variables, such as using One-Hot Encoding to process categorical data in user features.

[0047] S01: Feature Engineering: Extract features that are helpful for classification, which may include the user's historical work order records, the urgency of the work order, seasonal factors, etc. Normalize the features to eliminate the impact of different dimensions.

[0048] S04: Model training: Use the SVM algorithm to train the model. SVM is a supervised learning algorithm suitable for classification problems. Its core idea is to find an optimal hyperplane in the feature space to maximize the interval between different categories. Choose a suitable kernel function, such as linear kernel (`kernel='linear'`), radial basis kernel (`kernel='rbf'`), or polynomial kernel (`kernel='poly'`), and adjust the parameters to obtain the best performance.

[0049] S05: Model parameter adjustment: Use cross validation (such as 5-fold cross validation) and grid search (GridSearchCV) to optimize model parameters, such as the penalty coefficient C, kernel function parameter gamma, etc.

[0050] S06: Model evaluation: Evaluate the performance of the model on the test set, using metrics such as accuracy, recall, F1 score, etc. Observe the confusion matrix to understand the performance of the model on different categories.

[0051] S07: Model application: Apply the trained model to new work order data to automatically identify the type and attribution of work orders.

[0052] S08: Result interpretation and feedback: Explain the model’s prediction results so that grid personnel can understand and take appropriate measures. Collect feedback and continuously optimize the model based on the model’s performance in actual applications.

[0053] Through the above steps, a model based on the SVM algorithm can be constructed to automatically identify the types and attributions of power grid collection work orders, thereby improving the efficiency and accuracy of work order processing.

[0054] Step 6: Establish the alarm quantity threshold for each type of work order, collect the work orders and user characteristics within the cycle, automatically identify the types of work orders and the attribution of work orders, and then perform statistics, and issue an alarm based on the alarm quantity threshold. For example, if the service attitude problem of complaint work orders is identified 5 times (five work orders have this problem within the cycle), an alarm will be issued.

[0055] Step 7: Obtain the classification results by calling the classification model's API interface, and adaptively transform the corresponding work order type interface of the front-end business system to ultimately achieve the purpose of model result query, screening and quality inspection.

[0056] In summary, the present invention collects historical copying and reminder request work order texts and performs data preprocessing, and uses the existing user load data and electricity consumption data in the marketing system as user features; then builds copying and reminder guidance attribution classification labels, and constructs a professional vocabulary and further constructs a professional classification vocabulary, and classifies copying and reminder work orders based on the professional classification vocabulary; uses the copying and reminder work order type and copying and reminder work order attribution as labels to mark the copying and reminder work orders, and uses the SVM algorithm to train the model, so that the copying and reminder work order type and copying and reminder work order attribution can be finally identified through the copying and reminder work order and user features; finally, establishes the alarm quantity threshold of each copying and reminder work order attribution of each type of copying and reminder work order, collects the copying and reminder work order and user features within the period, automatically identifies the copying and reminder work order type and copying and reminder work order attribution, and then performs statistics, and issues an alarm based on the alarm quantity threshold, so as to realize the management of batch copying and reminder request work orders.

[0057] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0058] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0059] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0060] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders, characterized in that: The following steps are involved: Collect historical copy request work order texts and perform data preprocessing, and use the existing user load data and electricity consumption data in the marketing system as user features; Build the attribution classification label of copy reminder guidance, build a professional vocabulary and further build a professional classification vocabulary, and then classify the copy reminder work orders based on the professional classification vocabulary; The types of work orders and the reasons for the work orders are used as labels to mark the work orders, and the SVM algorithm is used to train the model, so that the types of work orders and the reasons for the work orders can be identified through the features of the work orders and users. Establish an alarm quantity threshold for each type of copy-reminder work order and each copy-reminder work order attribution, collect copy-reminder work orders and user characteristics within the collection period, automatically identify the copy-reminder work order type and copy-reminder work order attribution, and then perform statistics, and issue an alarm based on the alarm quantity threshold.

2. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 1, characterized in that: The types of work orders include complaint work orders, report work orders, opinion and suggestion work orders, business application work orders and inquiry and consultation work orders. The causes of the power grid copy and reminder work orders include service attitude problems, business ability problems, power quality problems, power application problems, frequent power outages and voltage abnormalities, and power outage and fault repair problems. The professional classification vocabulary includes a complaint work word library, a report work word library, an opinion and suggestion work word library, a business application work word library, and an inquiry and consultation work word library.

3. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 2, characterized in that: The construction process of the professional classification vocabulary is as follows: First, determine the business vocabulary, which is a collection of all professional terms in this profession. ,in They correspond to R professional classification word libraries, including complaint word library, report word library, opinion and suggestion word library, business application word library, business application word library and query and consultation word library; The professional classification vocabulary is the words in the set of professional words that appear in the corresponding copying and reminder request work order type, and these words are the professional classification vocabulary of this type of copying and reminder request work order.

4. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 1, characterized in that: The data preprocessing includes: Punctuation removal: remove all symbols to reduce the size of training data; Stop word and rare word removal: Collect prepositions and greetings in the complaint work orders to create a stop word library, and remove the corresponding words in the complaint text based on the stop word library to achieve the purpose of text cleaning; rare words are words that only exist in a few work orders. Replace rare words with other synonyms to increase word frequency, or directly delete them to improve model iteration efficiency; Disambiguation conversion: Use TF-IDF as the weight to judge the frequency of some homophones in the text description and complete the disambiguation conversion; Idiom Removal: Remove common texts from different text categories to improve model iteration efficiency.

5. The method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 1, characterized in that: The specific classification of the copy reminder work order is as follows: extract the words contained in the business vocabulary in the copy reminder work order, and then compare the proportion of these words in various professional classification vocabulary libraries, and take the copy reminder work order type corresponding to the professional classification vocabulary library with the largest proportion as the final output copy reminder work order type.

6. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 1, characterized in that: The specific steps of training the model using the SVM algorithm are as follows: S01: Data collection: historical copy request work order text, and the user load data and electricity consumption data already in the marketing system as user characteristics; S02: Data preprocessing: Clean the collected data, including removing missing values, processing outliers, and unifying data formats; S01: Feature Engineering: Extract features that are helpful for classification, including the user's historical work order records, the urgency of the work order, and seasonal factors, and normalize the features to eliminate the impact of different dimensions; S04: Model training: Use the SVM algorithm to train the model; S05: Model parameter adjustment: Use cross-validation and grid search to optimize model parameters, including penalty coefficient C and kernel function parameter gamma; S06: Model evaluation: Evaluate the performance of the model on the test set, including using accuracy, recall, and F1 score; observe the confusion matrix to understand the performance of the model on different categories; S07: Model application: Apply the trained model to new work order data to automatically identify the type and attribution of work orders; S08: Result interpretation and feedback: Explain the model’s prediction results, collect feedback and continuously optimize the model based on the model’s performance in actual applications.

7. A method for managing copying and reminder tasks based on automatic classification of copying and reminder work orders according to claim 1, characterized in that: The following steps are also included: By calling the API interface of the classification model, the classification results are obtained, and the corresponding work order type interface of the front-end business system is adaptively modified to ultimately achieve the purpose of model result query, screening and quality inspection.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the copying task management method for automatically classifying copying work orders as described in any one of claims 1 to 7.

9. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the method for managing copying and reminder tasks by automatically classifying copying and reminder work orders as described in any one of claims 1 to 7.

Images