Work order automatic distribution system and method applying government affair service hotline and medium

Abstract

The invention discloses an automatic work order distribution system and method applying a government affair service hotline and a medium, relates to the technical field of government affair information processing, and solves the technical problems that in the prior art, after work orders are distributed to corresponding departments, the work orders are randomly distributed to workers for processing, so that task load distribution is unbalanced, and the work orders cannot be processed in time. The method comprises the steps of obtaining call audio of citizens, and generating a hotline work order according to the call audio of the citizens; classifying the hotline work orders to obtain processing departments corresponding to the hotline work orders; acquiring work information of workers of corresponding departments, and distributing the hotline work order to the workers for processing according to the work information; processing of the hotline work order is monitored in real time, and real-time progress information is obtained; feeding back the work order according to the real-time progress information; the work order processing efficiency is improved, and the satisfaction degree and happiness of citizens are improved.

Description

Technical Field

[0001] The present invention belongs to the field of government information processing and relates to a work order automatic distribution technology, specifically a work order automatic distribution system, method and medium using a government service hotline. Background Art

[0002] A work order is a formal record of a government service hotline receiving public appeals, detailing key information such as the issues raised by the public, contact information, and the content of the appeal. Work orders can be used to clearly track the progress and results of each appeal, ensuring that the public's problems are properly resolved. The automatic allocation system can quickly identify and allocate work orders in a short period of time, reducing the time required for manual allocation, thereby speeding up response times and addressing public issues in a timely manner. The automatic allocation system can intelligently allocate work orders based on factors such as the staff's current workload, professional skills, and experience, ensuring that resources are reasonably allocated and improving overall work efficiency.

[0003] The prior art (invention patent application with publication number CN117474219A) discloses a method for intelligent dispatching of citizen hotline work orders, including: S1: establishing an intelligent dispatching system, pre-processing the work order data obtained from the citizen hotline through the system, dividing it according to different data types and marking it with labels, and marking similar department work orders with the same label; S2: special work order screening, directly dispatching work orders containing special characters; S3: text rule mining dispatching, by establishing text rules, extracting classified work orders according to text rules, and recording all rules to form a rule library; S4: training The system model is used to dispatch orders. For those that do not conform to the dispatching methods in S2 and S3, the dispatching system model is trained to dispatch them separately. S5: Intelligent dispatching. According to the work order classification number, the system automatically selects any one of the dispatching methods S2, S3 and S4 according to the type of work order to dispatch the order. This realizes the intelligent dispatching of work orders, eliminating the need for manual allocation and improving dispatching efficiency. However, after the actual work orders are assigned to the corresponding departments, they will be randomly dispatched to staff for processing, which will cause the workload of some staff to be overloaded and some staff to have no workload. This will result in an uneven distribution of tasks and an inability to process work orders in a timely manner.

[0004] The present invention provides a system, method and medium for automatically distributing work orders using a government service hotline to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an automatic work order distribution system, method and medium using a government service hotline, which is used to solve the technical problem in the prior art that after the work order is allocated to the corresponding department, it will be randomly distributed to staff for processing, which will cause the workload of some staff to be overloaded and some staff to have no workload; resulting in uneven distribution of tasks and inability to process the work order in a timely manner.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a work order automatic distribution system using a government service hotline, comprising: a work order distribution module, and a work order generation module and a work order feedback module connected thereto;

[0007] Work order generation module: used to obtain citizens' call audio and generate hotline work orders based on the citizens' call audio;

[0008] Work order allocation module: used to classify hotline work orders and obtain the corresponding processing department of the hotline work orders; obtain the work information of the staff in the corresponding department, and assign the hotline work orders to the staff for processing based on the work information;

[0009] Work order feedback module: used to monitor the processing of hotline work orders in real time and obtain real-time progress information; and provide feedback on the work orders based on the real-time progress information.

[0010] Preferably, generating a hotline work order based on the citizen's call audio includes:

[0011] Retrieve the audio of citizens' calls and pre-process the audio of citizens' calls; the pre-processing includes: noise reduction, filtering and echo removal;

[0012] Use speech recognition technology to convert the pre-processed call audio into call text; use natural language processing technology to extract feature information from the call text, and integrate the feature information to obtain a hotline work order.

[0013] Preferably, the extracting characteristic information from the call text using natural language processing technology includes:

[0014] Retrieve the conversation text of citizens to obtain the characteristic types of hotline work orders; divide the conversation text into several words; remove stop words from the several words according to the stop word library to obtain characteristic words; the characteristic types of hotline work orders include: address, time, problem type, and urgency;

[0015] Using a word embedding model, the feature words and the address, time, and problem type of the hotline ticket are converted into feature vectors and type vectors, respectively. The cosine similarity between the feature vectors and the type vectors is calculated, and the feature word corresponding to the feature vector with the highest cosine similarity is extracted as the initial information for the corresponding feature type.

[0016] Extract the event time in the call text to obtain the time urgency table; match the event time with the time urgency table to obtain the corresponding urgency level; integrate the initial information with the urgency level to obtain feature information.

[0017] The present invention divides the call text, removes stop words, and obtains characteristic words; matches the characteristic words with the characteristic types of hotline work orders to obtain corresponding initial information; extracts event time in the call text, and matches the corresponding urgency according to the event time; integrates the initial information with the urgency to obtain characteristic information; provides a basis for the generation of work orders, lays the foundation for the classification of work orders, and understands the needs of citizens.

[0018] Preferably, the hotline work orders are classified to obtain the processing departments corresponding to the hotline work orders, including:

[0019] Retrieve the hotline work order and use it as the allocation input sequence; call the work order allocation model;

[0020] Input the allocation input sequence into the work order allocation model to obtain the corresponding type label; match the corresponding processing department according to the type label; wherein the work order allocation model is constructed based on the artificial intelligence model.

[0021] The present invention uses a work order allocation model to analyze work orders and obtain corresponding type tags; matches the processing departments corresponding to the work orders according to the type tags; and can automatically allocate work orders, which is beneficial to improving the accuracy and speed of work order allocation.

[0022] Preferably, the work order allocation model is constructed based on an artificial intelligence model, including:

[0023] Obtain historical hotline work orders and corresponding type labels; use the historical hotline work orders and corresponding type labels as labeled input data and standard output data, respectively, and integrate them into a standard data set;

[0024] Divide the standard data set into training set, validation set and test set according to the set ratio; use the training set to train the artificial intelligence model; use the validation set to adjust the internal parameters of the artificial intelligence model; use the test set to test the artificial intelligence model and obtain test indicators;

[0025] Determine whether the test indicator is greater than the indicator threshold; if yes, mark the trained artificial intelligence model as the work order allocation model; if not, retrain the artificial intelligence model; wherein the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

[0026] It should be noted that a large amount of data is required in the process of training artificial intelligence models. When the amount of historical data is insufficient, a large amount of standard data can be obtained through data simulation; the simulated data and historical data are interspersed and integrated into a standard data set; the test indicators include: accuracy, recall rate, F1 score and stability; the indicator threshold is set through experimental simulation; the division ratio of the standard data set is set by expert assessment, and when the artificial intelligence model needs to be retrained, the division ratio needs to be reassessed.

[0027] The present invention utilizes historical data to train the artificial intelligence model, which can improve the accuracy of the distribution of the trained artificial intelligence model.

[0028] Preferably, allocating the hotline work order to a staff member for processing according to the work information includes:

[0029] Retrieve the work information of several staff members; the work information includes the remaining workload percentage, praise rate, and processing efficiency;

[0030] Scoring several staff members based on their work information; sorting the staff members from highest to lowest according to their scores to obtain an allocation ranking table; and allocating hotline work orders to corresponding staff members based on the allocation ranking table.

[0031] It should be noted that when hotline work orders are assigned according to the assignment ranking table, they are assigned in the order of the assignment ranking table. When the last staff member is assigned, the staff members are re-ranked by score.

[0032] The present invention scores staff members according to their work information, sorts the staff members from high to low according to their scores, and obtains an allocation sorting table; allocates hotline work orders to corresponding staff members according to the allocation sorting table; and can give priority to assigning staff members with high scores, which is beneficial to improving the efficiency and quality of hotline work order processing.

[0033] Preferably, the scoring of several staff members based on their work information includes:

[0034] Retrieve the work information of several staff members; mark the remaining workload ratio, praise rate, and processing efficiency in the work information as SGi, HPi, and CXi respectively;

[0035] By formula Calculate the score of the corresponding staff member; where α and β are proportional coefficients greater than 0; i represents the staff member's number; i = 1, 2, ..., n, where n is a positive integer; and the value range of k is (0, 1).

[0036] It should be noted that the proportional coefficient is set according to the actual situation. When the remaining workload of all staff members exceeds the workload threshold, the value of α is set to a larger value and the value of β is set to a smaller value; priority is given to the remaining workload of the staff; when the remaining workload of the staff does not exceed the workload threshold, the value of α is set to a smaller value and the value of β is set to a larger value; the workload threshold is set according to the average workload completed by the staff within the set time period.

[0037] Preferably, providing feedback on the work order based on real-time progress information includes:

[0038] Retrieve the real-time progress information of the hotline work order. If the real-time progress information indicates completion, send the completion information to the corresponding citizen. Otherwise, fit the real-time progress information into a completion curve. Based on the completion curve, predict the completion time of the processing to obtain the predicted completion time.

[0039] Get the set expected completion time and calculate the difference between the predicted completion time and the expected completion time; when the difference is greater than the difference threshold, notify the corresponding management personnel to increase the number of people dispatched for processing; otherwise, continue to monitor the real-time progress of the work order.

[0040] It should be noted that the real-time progress information is expressed as a percentage of completion; for example, when the completion degree of a hotline work order is 50%, the real-time progress information of the corresponding hotline work order is marked as 50%.

[0041] The present invention monitors the real-time progress of hotline work orders and feeds back the processed work orders to the corresponding citizens; when the real-time progress lags behind expectations, additional personnel are assigned to handle the work orders; ensuring that the work orders can be completed on schedule is conducive to improving the satisfaction and happiness of citizens.

[0042] A second aspect of the present invention provides a method for automatically allocating work orders using a government service hotline, comprising:

[0043] Step S1: Obtain the citizen's call audio and generate a hotline work order based on the citizen's call audio;

[0044] Step S2: Classify the hotline work orders and obtain the processing departments corresponding to the hotline work orders;

[0045] Step S3: Obtain the work information of the staff of the corresponding department and assign the hotline work order to the staff for processing according to the work information;

[0046] Step S4: Monitor the processing of the hotline work order in real time to obtain real-time progress information; and provide feedback on the work order based on the real-time progress information.

[0047] The third aspect of the present invention provides a medium for automatically distributing work orders using a government service hotline, wherein the medium stores computer program instructions, and the above-mentioned method steps are implemented when the program instructions are executed by a processor.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. The present invention pre-processes the audio of citizens' calls, which can remove noise in the call audio, make the call audio clearer, and facilitate subsequent language recognition; uses natural language processing technology to extract feature information in the call text, which can facilitate the extraction of citizens' needs and lay the foundation for the subsequent allocation of work orders; divides the call text, removes stop words, and obtains feature words; matches the feature words with the feature types of hotline work orders to obtain corresponding initial information; uses the work order allocation model to analyze the work order to obtain the corresponding type label; matches the processing department corresponding to the work order according to the type label; can automatically allocate work orders, which is conducive to improving the accuracy and rate of work order allocation; uses historical data to train the artificial intelligence model, which can improve the accuracy of the allocation of the trained artificial intelligence model.

[0050] 2. The present invention assigns hotline work orders to corresponding staff members according to an allocation ranking table; it can give priority to assigning staff members with high scores, which is conducive to improving the efficiency and quality of hotline work order processing; it monitors the real-time progress of hotline work orders and feeds back the completed work orders to the corresponding citizens; when the real-time progress lags behind expectations, it increases the number of people to handle the work orders; it ensures that the work orders can be completed on time, which is conducive to improving the satisfaction and happiness of citizens. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0052] Figure 1 A schematic diagram of the overall steps of the system of the present invention;

[0053] Figure 2 A schematic diagram of the steps for allocating work order departments according to the present invention;

[0054] Figure 3 A schematic diagram of the staff allocation and work order feedback steps of the present invention;

[0055] Figure 4 Schematic diagram of the specific steps of the method of the present invention. DETAILED DESCRIPTION

[0056] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] See also Figure 1 , a first aspect of the present invention provides an automatic work order distribution system using a government service hotline, comprising: a work order distribution module, and a work order generation module and a work order feedback module connected thereto;

[0058] Work order generation module: used to obtain citizens' call audio and generate hotline work orders based on the citizens' call audio;

[0059] Work order allocation module: used to classify hotline work orders and obtain the corresponding processing department of the hotline work orders; obtain the work information of the staff in the corresponding department, and assign the hotline work orders to the staff for processing based on the work information;

[0060] Work order feedback module: used to monitor the processing of hotline work orders in real time and obtain real-time progress information; and provide feedback on the work orders based on the real-time progress information.

[0061] See also Figure 2 , obtain the citizens' call audio, and pre-process the citizens' call audio; wherein the pre-processing includes: noise reduction, filtering and echo removal processing;

[0062] Use speech recognition technology to convert pre-processed call audio into call text; obtain the characteristic types of hotline work tickets; divide the call text into several words; remove stop words from several words based on a stop word library to obtain characteristic words; the characteristic types of hotline work tickets include: address, time, problem type, and urgency;

[0063] The word vector model is used to convert the feature words and the address, time and problem type of the hotline ticket into feature vectors and type vectors respectively. The cosine similarity between the feature vectors and the type vectors is calculated respectively, and the feature words corresponding to the feature vector with the highest cosine similarity are extracted as the initial information of the corresponding feature type.

[0064] It should be noted that the specific steps for work order feature matching using the word vector model are as follows:

[0065] 1. Extract high-frequency feature words based on historical work order data and establish a structured feature labeling system. Use pre-trained models or domain self-training models to ensure that semantic representations match business scenarios.

[0066] 2. Standardize address formats; perform weighted averaging of word vectors for segmented addresses; convert timestamps into vector combinations for business periods; and perform layered vector concatenation for multi-level classification labels (e.g., "network problem → broadband → disconnection").

[0067] 3. Use cosine similarity to calculate the degree of match between the feature vector and the type vector; dynamically adjust the weights; set a minimum similarity threshold; and return the three candidate feature words with the highest similarity for subsequent manual verification or model fusion decision-making. For multi-round work order conversations, introduce an attention mechanism to dynamically adjust the vector weights of historical conversations.

[0068] For example: processing an e-commerce logistics complaint work order; input work order description: "The express shows that it has been signed for, but the customer has not received it"; feature matching process: address vector: vec("Pudong New Area, Shanghai"); time vector: vec("2023-11-11 22:30") (night time during the promotion); problem vector: vec("logistics anomaly"); output result: matching feature word "delivery delay" (similarity 0.89), triggering the automatic compensation process.

[0069] Extract the event time from the call text to obtain a time urgency table; match the event time with the time urgency table to obtain the corresponding urgency level; integrate the initial information with the urgency level to obtain feature information, and integrate the feature information to obtain a hotline work order; use the hotline work order as the allocation input sequence; call the work order allocation model; input the allocation input sequence into the work order allocation model to obtain the corresponding type label; match the corresponding processing department based on the type label; where the type label is set to a positive integer.

[0070] For example, suppose we obtain an audio clip of a citizen's phone call and convert it into text: "I want to complain about the construction work starting at 8:00 a.m. every day near a certain primary school in the High-tech Zone. The noise is too loud and affects the students' classes. I hope to handle it as soon as possible."

[0071] The hotline ticket generated based on the call text is:

[0072] Address: Near a certain primary school in the High-tech Zone; Time: 8:00 a.m. every day; Problem type: Construction noise; Urgency: 80%;

[0073] The hotline work order is analyzed using the work order allocation model, and the corresponding type label is 6; then, the corresponding department matched according to the type label is: High-tech Zone Ecological Environment Management Bureau.

[0074] It is worth noting that the work order allocation model is built based on an artificial intelligence model, including:

[0075] Obtain historical hotline work orders and corresponding type labels; use the historical hotline work orders and corresponding type labels as labeled input data and standard output data, respectively, and integrate them into a standard data set;

[0076] Divide the standard data set into training set, validation set and test set according to the set ratio; use the training set to train the artificial intelligence model; use the validation set to adjust the internal parameters of the artificial intelligence model; use the test set to test the artificial intelligence model and obtain test indicators;

[0077] Determine whether the test indicator is greater than the indicator threshold; if yes, mark the trained artificial intelligence model as the work order allocation model; if not, retrain the artificial intelligence model; wherein the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

[0078] It should be noted that a large amount of data is required in the process of training artificial intelligence models. When the amount of historical data is insufficient, a large amount of standard data can be obtained through data simulation; the simulated data and historical data are interspersed and integrated into a standard data set; the test indicators include: accuracy, recall rate, F1 score and stability; the indicator threshold is set through experimental simulation; the division ratio of the standard data set is set by expert assessment, and when the artificial intelligence model needs to be retrained, the division ratio needs to be reassessed.

[0079] It should be noted that preprocessing the audio of citizens' calls can improve the clarity of the audio, which is beneficial to improving the accuracy of recognition when the audio is recognized; converting the call audio into call text, dividing the call text to obtain several words; removing stop words in several words to obtain several feature words; analyzing the matching degree of several feature words and the feature types of hotline work orders to obtain preliminary information; analyzing the urgency according to the event time in the call text; integrating the preliminary information and the urgency to obtain a hotline work order, which can extract the content of the call text and generate key information; allocating the hotline work order to the corresponding department based on the key information utilization model; allocating departments based on the call text, and analyzing key information is beneficial to reducing the amount of data processing and improving the efficiency of data processing.

[0080] See also Figure 3 , obtain the work information of the corresponding department staff; the work information includes the remaining workload ratio, praise rate and processing efficiency; the remaining workload ratio, praise rate and processing efficiency in the work information are marked as SGi, HPi and CXi respectively; through the formula Calculate the corresponding staff member's score; where α and β are proportional coefficients greater than 0; i represents the staff member's number; i = 1, 2, ..., n, where n is a positive integer; and k ranges from (0, 1). Sort the staff members by their scores from highest to lowest to obtain an allocation ranking table; assign the hotline ticket to the corresponding staff member according to the allocation ranking table.

[0081] It should be noted that the remaining workload ratio SGi refers to the percentage of the remaining workload of the staff member numbered i to the remaining workload of the entire department; the higher the remaining workload ratio of the staff member, the higher the The smaller the value of , the smaller the corresponding staff score; when the staff's praise rate or processing efficiency is higher, The larger the value of , the greater the score of the corresponding staff member.

[0082] The processing of hotline work orders is monitored in real time to obtain real-time progress information. When the real-time progress information indicates completion, the completion information is sent to the corresponding citizen. Otherwise, the real-time progress information is fitted into a completion curve. The completion time of the processing is predicted based on the completion curve to obtain the predicted completion time.

[0083] Get the set expected completion time and calculate the difference between the predicted completion time and the expected completion time; when the difference is greater than the difference threshold, notify the corresponding management personnel to increase the number of people dispatched for processing; otherwise, continue to monitor the real-time progress of the work order.

[0084] For example, suppose you are providing feedback on three hotline tickets and want to obtain real-time progress information for each of them:

[0085] Hotline work order A: Real-time progress information: 100% completed; then the information that work order A is processed is sent to the corresponding citizen;

[0086] Hotline Work Order B: Real-time progress information 20%; predicted completion time is 10 days later, expected completion time is 5 days later, and the difference between the predicted and expected completion times is 5 days, exceeding the difference threshold of 3 days; additional personnel will be assigned to handle the work order;

[0087] Hotline work order C: Real-time progress information 80%; predicted completion time 1 day later, expected completion time 2 days later, then the real-time progress of the hotline work order is continuously monitored. When hotline work order C is completed, the completion information will be fed back to the corresponding citizen.

[0088] It should be noted that the work information of the staff is obtained, the staff score is calculated based on the work information of the staff, and the staff is sorted according to the staff score to obtain an allocation ranking table; the hotline work order is assigned to the corresponding staff according to the allocation ranking table, the staff can be comprehensively evaluated, and the staff with high scores are given priority to handle the hotline work order; the processing of the hotline work order is monitored in real time, and the hotline work order is fed back based on the real-time progress information. For hotline work orders that are behind schedule, the number of people dispatched to handle them is increased to ensure that the hotline work order can be completed on schedule, which is conducive to improving the satisfaction and happiness of citizens and improving the quality of life of citizens.

[0089] See also Figure 4 The second embodiment of the present invention provides a method for automatically allocating work orders using a government service hotline, comprising:

[0090] Step S1: Obtain the citizen's call audio and generate a hotline work order based on the citizen's call audio;

[0091] Step S2: Classify the hotline work orders and obtain the processing departments corresponding to the hotline work orders;

[0092] Step S3: Obtain the work information of the staff of the corresponding department and assign the hotline work order to the staff for processing according to the work information;

[0093] Step S4: Monitor the processing of the hotline work order in real time to obtain real-time progress information; and provide feedback on the work order based on the real-time progress information.

[0094] The third embodiment of the present invention provides a medium for automatically distributing work orders using a government service hotline, on which computer program instructions are stored. When the program instructions are executed by a processor, the above-mentioned method steps are implemented.

[0095] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0096] The working principle of the present invention is as follows: the present invention obtains the audio of citizens' calls and generates a hotline work order based on the audio of citizens' calls; classifies the hotline work orders to obtain the processing departments corresponding to the hotline work orders; obtains the work information of the staff of the corresponding department, and assigns the hotline work orders to the staff for processing based on the work information; monitors the processing of the hotline work orders in real time to obtain real-time progress information; and provides feedback on the work orders based on the real-time progress information.

[0097] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. Apply the automatic work order distribution system of the government service hotline, which is characterized by: include: Work order allocation module, and the connected work order generation module and work order feedback module; Work order generation module: used to obtain citizens' call audio and generate hotline work orders based on the citizens' call audio; Work order allocation module: used to classify hotline work orders and obtain the corresponding processing departments of hotline work orders; Obtain the work information of the staff in the corresponding department and assign the hotline work order to the staff for processing based on the work information; Work order feedback module: used to monitor the processing of hotline work orders in real time and obtain real-time progress information; and provide feedback on the work orders based on the real-time progress information.

2. The automatic work order distribution system using the government service hotline according to claim 1 is characterized in that: The generation of a hotline work order based on the citizen's call audio includes: Retrieve the audio of citizens' calls and pre-process the audio of citizens' calls; the pre-processing includes: noise reduction, filtering and echo removal; Use speech recognition technology to convert the pre-processed call audio into call text; use natural language processing technology to extract feature information from the call text, and integrate the feature information to obtain a hotline work order.

3. The automatic work order distribution system using the government service hotline according to claim 2 is characterized in that: The method of extracting characteristic information from the call text using natural language processing technology includes: Retrieve the conversation text of citizens to obtain the characteristic types of hotline work orders; divide the conversation text into several words; remove stop words from the several words according to the stop word library to obtain characteristic words; the characteristic types of hotline work orders include: address, time, problem type, and urgency; Using a word embedding model, the feature words and the address, time, and problem type of the hotline ticket are converted into feature vectors and type vectors, respectively. The cosine similarity between the feature vectors and the type vectors is calculated, and the feature word corresponding to the feature vector with the highest cosine similarity is extracted as the initial information for the corresponding feature type. Extract the event time in the call text to obtain the time urgency table; match the event time with the time urgency table to obtain the corresponding urgency level; integrate the initial information with the urgency level to obtain feature information.

4. The automatic work order distribution system using the government service hotline according to claim 1 is characterized in that: The hotline work orders are classified to obtain the processing departments corresponding to the hotline work orders, including: Retrieve the hotline work order and use it as the allocation input sequence; call the work order allocation model; Input the allocation input sequence into the work order allocation model to obtain the corresponding type label; match the corresponding processing department according to the type label; wherein the work order allocation model is constructed based on the artificial intelligence model.

5. The automatic work order distribution system using the government service hotline according to claim 4 is characterized in that: The work order allocation model is built based on an artificial intelligence model and includes: Obtain historical hotline work orders and corresponding type labels; use the historical hotline work orders and corresponding type labels as labeled input data and standard output data, respectively, and integrate them into a standard data set; Divide the standard data set into training set, validation set and test set according to the set ratio; use the training set to train the artificial intelligence model; use the validation set to adjust the internal parameters of the artificial intelligence model; use the test set to test the artificial intelligence model and obtain test indicators; Determine whether the test indicator is greater than the indicator threshold; if yes, mark the trained artificial intelligence model as the work order allocation model; if not, retrain the artificial intelligence model; wherein the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

6. The automatic work order distribution system using the government service hotline according to claim 1 is characterized in that: Allocating the hotline work order to staff for processing based on the work information includes: Retrieve the work information of several staff members; the work information includes the remaining workload percentage, praise rate, and processing efficiency; Scoring several staff members based on their work information; sorting the staff members from highest to lowest according to their scores to obtain an allocation ranking table; and allocating hotline work orders to corresponding staff members based on the allocation ranking table.

7. The automatic work order distribution system using the government service hotline according to claim 6 is characterized in that: The scoring of several staff members based on their work information includes: Retrieve the work information of several staff members; mark the remaining workload ratio, praise rate, and processing efficiency in the work information as SGi, HPi, and CXi respectively; By formula Calculate the score of the corresponding staff member; where α and β are proportional coefficients greater than 0; i represents the staff member's number; i = 1, 2, ..., n, where n is a positive integer; and the value range of k is (0, 1).

8. The automatic work order distribution system using the government service hotline according to claim 1 is characterized in that: Providing feedback on the work order based on real-time progress information includes: Retrieve the real-time progress information of the hotline work order. If the real-time progress information indicates completion, send the completion information to the corresponding citizen. Otherwise, fit the real-time progress information into a completion curve. Based on the completion curve, predict the completion time of the processing to obtain the predicted completion time. Get the set expected completion time and calculate the difference between the predicted completion time and the expected completion time; when the difference is greater than the difference threshold, notify the corresponding management personnel to increase the number of people dispatched for processing; otherwise, continue to monitor the real-time progress of the work order.

9. A method for automatically allocating work orders using a government service hotline, applied to a system for automatically allocating work orders using a government service hotline according to any one of claims 1 to 8, characterized in that: include: Step S1: Obtain the citizen's call audio and generate a hotline work order based on the citizen's call audio; Step S2: Classify the hotline work orders and obtain the processing departments corresponding to the hotline work orders; Step S3: Obtain the work information of the staff of the corresponding department and assign the hotline work order to the staff for processing according to the work information; Step S4: Monitor the processing of the hotline work order in real time to obtain real-time progress information; and provide feedback on the work order based on the real-time progress information.

10. A medium for automatically distributing work orders using a government service hotline, applied to the method for automatically distributing work orders using a government service hotline according to claim 9, characterized in that: The medium stores computer program instructions, which implement the above method steps when executed by a processor.

Images