Cloud service order dispatching system based on multi-algorithm model
Patent Information
- Application Number
- CN202411149042.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-22
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
[0004]然而上述技术虽然通过两个模块的设立方便了工作人员对工单的快速了解和把控工单的处理时间,但在I T行业由于下属部门较多,需要多个部门处理不同的具体问题,这使得在派单方面需要投入巨大的人力成本,不仅要求派单人员的专业性,且人工派单也易收到外部因素影响,出现派单错误的情况,且I T公司下属的多个部门不免存在使用频次不高的情况,而人力难以察觉这些数据,这就容易造成公司部门设立冗余,造成运营成本的巨大提升;
[0039] The present invention obtains a task text set by receiving a business consultation text set from a client and performing preprocessing, and outputs an allocation result by collecting a historical efficiency set and establishing a task allocation model based on the task text set and the historical efficiency set. Compared with traditional manual dispatch, it not only greatly reduces the need for professional personnel, but also It also effectively reduces labor costs. Through the department efficiency set, the efficiency of each department can be intuitively viewed, and the needs and efficiency of the department can be more intuitively understood, which greatly reduces the company's operating costs. Overall, the present invention has advantages It has the obvious advantages of high single efficiency, good feedback effect and clear visual representation of the company's status.
Smart Images

Figure CN119005622A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of information technology, and more specifically, to a cloud service dispatching system based on a multi-algorithm model. Background Art
[0002] IT, or Internet technology, refers to an information technology developed and established on the basis of computer technology. Internet technology connects different devices to each other through the wide area network of computer networks, speeds up the transmission of information and broadens the channels for obtaining information, promotes the development of various software applications, and changes the way people live and learn. The widespread application of Internet technology is a sign of entering the information society. With the rapid development of my country's economy, the Internet has gradually entered thousands of households and become an indispensable part of people's lives. Internet technology has also developed rapidly. Today's Internet technology companies are blooming everywhere. When facing customer needs, most of them are still in the more traditional mode of manual classification and then dispatching orders in terms of classification and processing of needs. This not only puts forward requirements on the professionalism of dispatchers, but also requires companies to set up dispatching positions and corresponding staff, thereby increasing labor costs;
[0003] The patent application publication number CN110443501 discloses an intelligent dispatching system, which displays work order information through a display module, allowing staff to understand the actual status of the work order. By comparing the effective delivery time of the module with the actual delivery time, it is convenient for staff to quickly process the work order and control the time.
[0004] However, although the above technology facilitates the staff to quickly understand the work order and control the processing time of the work order through the establishment of two modules, in the IT industry, due to the large number of subordinate departments, multiple departments are required to handle different specific issues, which requires huge manpower costs in dispatching orders. Not only does it require the professionalism of dispatchers, but manual dispatching is also easily affected by external factors, resulting in dispatch errors. In addition, the multiple departments under the IT company inevitably have low usage frequencies, and it is difficult for human resources to detect these data, which easily leads to redundant department establishment in the company and a huge increase in operating costs.
[0005] In view of this, the present invention proposes a cloud service dispatching system based on a multi-algorithm model to solve the above problems. Summary of the invention
[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: comprising:
[0007] A task receiving module, used to receive a business consultation text set from a client;
[0008] The task processing module is used to pre-process the business consulting text set to obtain the task text set;
[0009] Furthermore, the method of preprocessing the business consulting text set includes extracting keywords from sentences in the business consulting text set to obtain key texts, and performing keyword classification on semantic texts to obtain a task text set;
[0010] Furthermore, the methods for extracting keywords include:
[0011] Perform structural analysis on sentences in the business consulting text set to obtain the syntactic structure of the sentences. The structural analysis includes segmenting the sentences into N phrases and identifying the nouns in the phrases.
[0012] Collect M sentences, manually annotate the sentences with noun tags, obtain the syntactic structure, calculate the basic structure of the syntactic structure, establish a regression model between the basic structure and the manually annotated noun tags, make the output result of the regression model the same as the tag, and obtain a keyword extraction model, the basic framework of the keyword extraction model is a TF-IDF model or a TextRank model;
[0013] Pack the nouns output by the keyword extraction model to obtain key texts;
[0014] Furthermore, the method of keyword classification includes presetting a category label set {1, 2, 3, ...K}, converting nouns in the key text into B noun vectors, using a multinomial logistic regression model to output a category result, comparing the category result with a preset category label set, obtaining a category after keyword classification, and packaging the category after keyword classification to obtain a task text set;
[0015] A monitoring execution module is used to collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, and monitor the task execution to obtain a first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score.
[0016] Furthermore, the method of establishing the task allocation model includes:
[0017] Collect the P group task text set as the sample set, and divide the sample set into 70% P training set, 15% P test set and 15% P validation set;
[0018] A classification model is established based on the training set. The input layer of the classification model is used to input the sample set, and the output layer is used to output the classification result. The classification model uses any one of the logistic regression algorithm, support vector machine and recurrent neural network;
[0019] The mathematical expression of the classification model is Get the classification value Yx, where e is a constant, zx is the probability that x belongs to the kth label in the model sample, and K is the total number of categories;
[0020] When the classification value is less than When , it indicates that the sample does not belong to category k. When the classification value is greater than When , it indicates that the sample belongs to category k;
[0021] Through C rounds of iterations, the classification model tends to output stable results and obtains the task classification model;
[0022] Pack the classification values and get the distribution results;
[0023] Furthermore, the method of monitoring the task execution status to obtain the first efficiency set includes:
[0024] Mark the time of issuing the task as T1, and the time of receiving the submitted task as T2, calculate the time difference, and get the time required to complete an order;
[0025] Count the number of tasks issued in one day for each tag in the preset category tag set, marked as D1, and get the number of tasks completed in one day, marked as D2;
[0026] D1, D2 and the category label are packaged to obtain the first efficiency set;
[0027] A service feedback module, used to obtain user feedback and the first efficiency set to obtain a current efficiency set;
[0028] Furthermore, the methods for obtaining the current efficiency set include:
[0029] The star rating of each feedback is counted separately, marked as D3g, where g includes 1, 2, 3, 4 and 5, to obtain the first efficiency set, and the category labels, D1, D2 and D3g in the first efficiency set are matched to obtain the second efficiency set;
[0030] Taking one day as the benchmark, the second efficiency sets of all IT departments are counted and packaged to obtain the current efficiency set;
[0031] Furthermore, the ways to obtain the department efficiency set include:
[0032] Based on the current efficiency set, obtain the category labels, D1, D2 and D3g of a single department. Take the label of a single department as the benchmark and substitute it into the calculation formula: Get the department efficiency value As, where Kn is the total task volume for one day, and M1, M2 and M3 are the preset weight factors;
[0033] Taking one day as the benchmark, collect the department efficiency values of all departments to obtain all efficiency values. When the department efficiency value is 0, it is not included in all efficiency values. All efficiency values are packaged based on the preset time of one day, one week or one quarter to obtain the department efficiency set.
[0034] Furthermore, a cloud service dispatching method based on a multi-algorithm model includes: S1: receiving a business consultation text set from a client;
[0035] S2: Preprocess the consultation text set to obtain the task text set;
[0036] S3: Collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, output the allocation results, and monitor the task execution to obtain the first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score.
[0037] S4: Obtain user feedback and the first efficiency set, obtain the current efficiency set, and perform analysis to obtain the department efficiency set.
[0038] The technical effects and advantages of the present invention are as follows:
[0039] The present invention receives a business consultation text set from a client and preprocesses it to obtain a task text set. It collects a historical efficiency set and establishes a task allocation model based on the task text set and the historical efficiency set to output an allocation result. Compared with traditional manual dispatching, it not only greatly reduces the demand for personnel expertise, but also effectively reduces labor costs. Through the department efficiency set, the efficiency of each department can be intuitively viewed, and the needs and efficiency of the department can be more intuitively understood, which greatly reduces the company's operating costs. Generally speaking, the present invention has the significant advantages of high dispatching efficiency, good feedback effect and clear and intuitive display of the company's status. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 It is a schematic diagram of a cloud service dispatching system based on a multi-algorithm model of the present invention;
[0041] Figure 2 Schematic diagram of the cloud service dispatching method based on a multi-algorithm model of the present invention. DETAILED DESCRIPTION
[0042] 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 only 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.
[0043] Example 1
[0044] See also Figure 1 As shown, the cloud service dispatching system based on the multi-algorithm model described in this embodiment includes:
[0045] A task receiving module, used to receive a business consultation text set from a client;
[0046] The task processing module is used to pre-process the business consulting text set to obtain the task text set;
[0047] Furthermore, the method of preprocessing the business consulting text set includes extracting keywords from sentences in the business consulting text set to obtain key texts, and performing keyword classification on semantic texts to obtain a task text set;
[0048] Further, the methods for extracting keywords include:
[0049] Perform structural analysis on sentences in the business consulting text set to obtain the syntactic structure of the sentences. The structural analysis includes segmenting the sentences into N phrases and identifying the nouns in the phrases.
[0050] Collect M sentences, manually annotate the sentences with noun tags, obtain the syntactic structure, calculate the basic structure of the syntactic structure, establish a regression model between the basic structure and the manually annotated noun tags, make the output result of the regression model the same as the tag, and obtain a keyword extraction model, the basic framework of the keyword extraction model is a TF-IDF model or a TextRank model;
[0051] Pack the nouns output by the keyword extraction model to obtain key texts;
[0052] Furthermore, the method of keyword classification includes presetting a category label set {1, 2, 3, ...K}, converting nouns in the key text into B noun vectors, using a multinomial logistic regression model to output a category result, comparing the category result with a preset category label set, obtaining a category after keyword classification, and packaging the category after keyword classification to obtain a task text set;
[0053] It should be explained that the digital labels in the preset category label set represent different responsible departments in an IT company, and K is the total number of categories, that is, the total number of departments in an IT company that need to dispatch orders;
[0054] A monitoring execution module is used to collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, and monitor the task execution to obtain a first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score.
[0055] Further, the method of establishing the task allocation model includes:
[0056] Collect the P group task text set as the sample set, and divide the sample set into 70% P training set, 15% P test set and 15% P validation set;
[0057] A classification model is established based on the training set. The input layer of the classification model is used to input the sample set, and the output layer is used to output the classification result. The classification model uses any one of the logistic regression algorithm, support vector machine and recurrent neural network;
[0058] The mathematical expression of the classification model is Get the classification value Yx, where e is a constant, zx is the probability that x belongs to the kth label in the model sample, and K is the total number of categories;
[0059] When the classification value is less than When , it indicates that the sample does not belong to category k. When the classification value is greater than When , it indicates that the sample belongs to category k;
[0060] Through C rounds of iterations, the classification model tends to output stable results and obtains the task classification model;
[0061] Pack the classification values and get the distribution results;
[0062] Furthermore, the method of monitoring the task execution status to obtain the first efficiency set includes:
[0063] Mark the time of issuing the task as T1, and the time of receiving the submitted task as T2, calculate the time difference, and get the time required to complete an order;
[0064] Count the number of tasks issued in one day for each tag in the preset category tag set, marked as D1, and get the number of tasks completed in one day, marked as D2;
[0065] D1, D2 and the category label are packaged to obtain the first efficiency set;
[0066] A service feedback module, used to obtain user feedback and the first efficiency set to obtain a current efficiency set;
[0067] Further, the methods of obtaining the current efficiency set include:
[0068] The star rating of each feedback is counted separately, marked as D3g, where g includes 1, 2, 3, 4 and 5, to obtain the first efficiency set, and the category labels, D1, D2 and D3g in the first efficiency set are matched to obtain the second efficiency set;
[0069] Taking one day as the benchmark, the second efficiency sets of all IT departments are counted and packaged to obtain the current efficiency set;
[0070] Furthermore, the department efficiency set is obtained by:
[0071] Based on the current efficiency set, obtain the category labels, D1, D2 and D3g of a single department. Take the label of a single department as the benchmark and substitute it into the calculation formula: Get the department efficiency value As, where Kn is the total task volume for one day, and M1, M2 and M3 are the preset weight factors;
[0072] Taking one day as the benchmark, collect the department efficiency values of all departments to obtain all efficiency values. When the department efficiency value is 0, it is not included in all efficiency values. All efficiency values are packaged based on the preset time of one day, one week or one quarter to obtain the department efficiency set.
[0073] This embodiment has the beneficial effect of receiving a business consultation text set from a client and preprocessing it to obtain a task text set. By collecting a historical efficiency set and establishing a task allocation model based on the task text set and the historical efficiency set, the allocation result is output. Compared with the traditional manual dispatching, it not only greatly reduces the demand for personnel expertise, but also effectively reduces labor costs. Through the department efficiency set, the efficiency of each department can be intuitively viewed, and the needs and efficiency of the department can be more intuitively understood, which greatly reduces the company's operating costs. In general, the present invention has the significant advantages of high dispatching efficiency, good feedback effect and clear intuitive display of the company's status.
[0074] Example 2
[0075] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides a method for detecting abnormalities in the power Internet of Things, including: S1: receiving a business consultation text set from a client;
[0076] S2: Preprocess the consultation text set to obtain the task text set;
[0077] S3: Collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, output the allocation results, and monitor the task execution to obtain the first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score.
[0078] S4: Obtain user feedback and the first efficiency set, obtain the current efficiency set, and perform analysis to obtain the department efficiency set.
[0079] Example 3
[0080] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the above-mentioned power Internet of Things anomaly detection method is implemented.
[0081] Since the electronic device introduced in this embodiment is an electronic device used to implement the cloud service dispatching system based on a multi-algorithm model in the embodiment of the present application, based on the cloud service dispatching system based on a multi-algorithm model introduced in the embodiment of the present application, the technical personnel in this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not introduced in detail here. As long as the technical personnel in this field implement the electronic device adopted by the cloud service dispatching system based on a multi-algorithm model in the embodiment of the present application, it belongs to the scope of protection of this application.
[0082] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.
[0083] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.
Claims
1. A cloud service dispatching system based on a multi-algorithm model, characterized in that: include: A task receiving module, used to receive a business consultation text set from a client; The task processing module is used to pre-process the business consulting text set to obtain the task text set; A monitoring execution module is used to collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, and monitor the task execution to obtain a first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score. The service feedback module is used to obtain user feedback and the first efficiency set to obtain the current efficiency set.
2. The cloud service dispatching system based on a multi-algorithm model according to claim 1, characterized in that: The method of preprocessing the business consulting text set includes extracting keywords from sentences in the business consulting text set to obtain key texts, and classifying semantic texts by keywords to obtain a task text set.
3. The cloud service dispatching system based on a multi-algorithm model according to claim 2, characterized in that: The method for extracting keywords includes: Perform structural analysis on sentences in the business consulting text set to obtain the syntactic structure of the sentences. The structural analysis includes segmenting the sentences into N phrases and identifying the nouns in the phrases. Collect M sentences, manually annotate the sentences with noun tags, obtain the syntactic structure, calculate the basic structure of the syntactic structure, establish a regression model between the basic structure and the manually annotated noun tags, make the output result of the regression model the same as the tag, and obtain a keyword extraction model, the basic framework of the keyword extraction model is a TF-IDF model or a TextRank model; The nouns output by the keyword extraction model are packaged to obtain key texts.
4. The cloud service dispatching system based on a multi-algorithm model according to claim 3, characterized in that: The method for keyword classification includes presetting a category label set {1, 2, 3, ... K}, converting the nouns in the key text into B noun vectors, using a polynomial logistic regression model to output category results, comparing the category results with the preset category label set, obtaining categories after keyword classification, and packaging the categories after keyword classification to obtain a task text set.
5. The cloud service dispatching system based on a multi-algorithm model according to claim 4, characterized in that: The method of establishing the task allocation model includes: Collect the P group task text set as the sample set, and divide the sample set into 70% P training set, 15% P test set and 15% P validation set; A classification model is established based on the training set. The input layer of the classification model is used to input the sample set, and the output layer is used to output the classification result. The classification model uses any one of the logistic regression algorithm, support vector machine and recurrent neural network; The mathematical expression of the classification model is Get the classification value Yx, where e is a constant, zx is the probability that x belongs to the kth label in the model sample, and K is the total number of categories; When the classification value is less than When , it indicates that the sample does not belong to category k. When the classification value is greater than When , it indicates that the sample belongs to category k; Through C rounds of iterations, the classification model tends to output stable results and obtains the task classification model; Pack the classification values and get the distribution results.
6. The cloud service dispatching system based on a multi-algorithm model according to claim 5, characterized in that: The method of monitoring the task execution status to obtain the first efficiency set includes: Mark the time of issuing the task as T1, and the time of receiving the submitted task as T2, calculate the time difference, and get the time required to complete an order; Count the number of tasks issued in one day for each tag in the preset category tag set, marked as D1, and get the number of tasks completed in one day, marked as D2; D1, D2 and the category label are packaged to obtain the first efficiency set.
7. The cloud service dispatching system based on a multi-algorithm model according to claim 6, characterized in that: Ways to get the current efficiency set include: The star rating of each feedback is counted separately, marked as D3g, where g includes 1, 2, 3, 4 and 5, to obtain the first efficiency set, and the category labels, D1, D2 and D3g in the first efficiency set are matched to obtain the second efficiency set; Taking one day as a benchmark, the second efficiency sets of all IT departments are counted and packaged to obtain the current efficiency set.
8. The cloud service dispatching system based on a multi-algorithm model according to claim 7, characterized in that: The ways to get the department efficiency set include: Based on the current efficiency set, obtain the category labels, D1, D2 and D3g of a single department. Take the label of a single department as the benchmark and substitute it into the calculation formula: Get the department efficiency value As, where Kn is the total task volume for one day, and M1, M2 and M3 are the preset weight factors; Taking one day as the benchmark, department efficiency values of all departments are collected to obtain all efficiency values. When the department efficiency value is 0, it is not included in all efficiency values. All efficiency values are packaged based on a preset time of one day, one week or one quarter to obtain a department efficiency set.
9. A cloud service dispatching method based on a multi-algorithm model, implemented according to the cloud service dispatching system based on a multi-algorithm model according to claims 1-8, characterized in that: S1: receiving a business consultation text set from a client; S2: Preprocess the consultation text set to obtain the task text set; S3: Collect historical efficiency sets, establish a task allocation model based on the historical efficiency sets and task text sets, output the allocation results, and monitor the task execution to obtain the first efficiency set. The historical efficiency set includes the time required to complete an order, the amount of tasks completed in one day, and the feedback score. S4: Obtain user feedback and the first efficiency set, obtain the current efficiency set, and perform analysis to obtain the department efficiency set.
Citation Information
Patent Citations
Power customer service work order processing system and method
CN110020777A
Intelligent work order assignment method based on machine learning and liability list identification
CN115759575A
Intelligent seat text processing method based on large model
CN117332066A
Customer service order sending method based on artificial intelligence
CN117391347A
Online customer service real-time distribution method and system
CN118170886A
Cited By
Electronic contract signing and order settlement integrated platform in zero work system
CN120746170A
