Enterprise service quality evaluation and optimization system and method

Abstract

The invention discloses an enterprise service quality evaluation and optimization system and method, and relates to the technical field of service quality evaluation, and the system comprises a prior data acquisition module, an abnormal evaluation group label generation module, an abnormal evaluation group processing module, a feedback terminal, a quality evaluation module, a Web set display end and a data warehouse. According to the method, before the service quality of the enterprise is evaluated, the abnormal evaluation of the enterprise is screened, then the screened abnormal evaluation is re-visited, the service quality of the enterprise is comprehensively evaluated in combination with the re-visited data, the reliability of the related data of the service quality evaluation of the enterprise is guaranteed, the accuracy of the service quality evaluation of the enterprise is improved, and the enterprise service quality evaluation efficiency is improved. According to the method, abnormal evaluation after screening is carried out, return visit is carried out according to a proportion, a return visit mode is confirmed, the feedback rate and return visit efficiency of evaluation users are improved, resource waste is avoided, and the competitiveness of the enterprises is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of service quality evaluation, and in particular to an enterprise service quality evaluation and optimization system and method. Background Art

[0002] In today's highly competitive market environment, the quality of corporate services has become one of the key indicators to measure the success of an enterprise. Service quality assessment is of strategic significance to the long-term development of an enterprise. High-quality services can win a good social reputation for an enterprise, enhance brand influence, and create favorable conditions for the enterprise's market expansion and new business development. Therefore, it is particularly important and necessary to conduct a comprehensive and objective assessment of the quality of corporate services.

[0003] Prior art, such as the invention application patent with announcement number: CN110472224B, discloses a method, device, computer equipment and storage medium for detecting service quality, wherein the method comprises: obtaining voice data during the service process. The voice data is converted into text to obtain a plurality of text data carrying role labels. Based on each of the text data and a pre-trained recognition model, a target semantic label and a first emotional label are obtained for each of the text data. The service quality is quantitatively detected based on the role label, target semantic label and first emotional label corresponding to each of the text data. By adopting the detection method in the embodiment, the quantitative detection result is more accurate, and there is no need to add a large number of keywords. Therefore, the maintenance cost of the detection system can be reduced and the user experience of the maintenance personnel can be improved.

[0004] Prior art, such as the invention application patent with announcement number: CN118134502B discloses an after-sales evaluation system for enterprise marketing consulting services, including a service information collection module, a service content collection module, a personnel information collection module, a data processing module and an information sending module. The service information collection module is used to collect service information, the service content collection module is used to collect service content information, and the personnel information collection module is used to collect service personnel information. The data processing module is used to process the service information, service content information and service personnel information to generate first evaluation information, second evaluation information and third evaluation information. The first evaluation information includes a-level evaluation information, b-level evaluation information and c-level evaluation information. The present invention can conduct a more comprehensive after-sales evaluation of enterprise marketing consulting services, further improve the quality of enterprise after-sales services, and thus improve the economic benefits of the enterprise.

[0005] Referring to the contents of the above scheme, it can be found that most of the existing technologies evaluate the service quality of enterprises through relevant parameters of service quality evaluation. However, in actual application, there is a phenomenon that the evaluation users are attracted by the additional conditions of the comments or guided by the staff and other situations to make high-quality evaluations. There is rarely any analysis at this level in the existing technology, which makes it difficult to ensure the reliability of the relevant data of the enterprise's service quality evaluation, reduce the accuracy of the enterprise's service quality evaluation, affect the scientific nature of the enterprise's decision-making, and hinder the continuous improvement of the enterprise. At the same time, the return visits to the evaluation users are also relatively blind, which reduces the efficiency of the return visits, causes waste of resources, and affects the competitiveness of the enterprise. Summary of the invention

[0006] The purpose of the present invention is to provide an enterprise service quality evaluation and optimization system and method, which solves the problems existing in the background technology.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions: The first aspect of the present invention provides an enterprise service quality evaluation and optimization system, including: a priori data acquisition module, used to obtain the prior evaluation data of the enterprise.

[0008] The abnormal evaluation group label generation module is used to generate the abnormal evaluation group label and the regular evaluation group label of the enterprise based on the prior evaluation data of the enterprise.

[0009] The abnormal evaluation group processing module is used to obtain the account ID corresponding to the abnormal evaluation group labels of each level from the abnormal evaluation group labels of the enterprise, and determine the number of evaluation extractions of the abnormal evaluation group labels of each level of the enterprise, obtain the extracted evaluations of the abnormal evaluation group labels of each level of the enterprise, and obtain the confirmation method of each account ID of the abnormal evaluation group labels of each level of the enterprise after processing.

[0010] The feedback terminal is used to provide feedback based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and collect the enterprise's abnormal feedback data.

[0011] The quality assessment module is used to evaluate the enterprise's service quality parameters based on the enterprise's abnormal feedback data and regular evaluation group labels, and obtain the enterprise's service quality optimization level after processing.

[0012] The Web integrated display terminal is used to display the enterprise's service quality parameters and service quality optimization level.

[0013] The second aspect of the present invention provides a method for executing the enterprise service quality evaluation and optimization system of the present invention, comprising: obtaining prior data, obtaining the enterprise's prior evaluation data from the enterprise operation platform.

[0014] Abnormal evaluation group label generation, based on the enterprise's prior evaluation data, generates the enterprise's abnormal evaluation group label and regular evaluation group label.

[0015] Abnormal evaluation group processing, obtain the account ID corresponding to the abnormal evaluation group label of each level from the abnormal evaluation group label of the enterprise, and determine the number of evaluation extractions of the abnormal evaluation group label of each level of the enterprise, randomly screen to obtain the extracted evaluations of the abnormal evaluation group label of each level of the enterprise, and obtain the confirmation method of each account ID of the abnormal evaluation group label of each level of the enterprise after processing.

[0016] Feedback processing is based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and the enterprise's abnormal feedback data is collected.

[0017] Quality assessment evaluates the enterprise's service quality parameters based on the enterprise's abnormal feedback data and regular evaluation group labels, and obtains the enterprise's service quality optimization level after processing.

[0018] Web integrated display processing displays the enterprise's service quality parameters and service quality optimization level.

[0019] The beneficial effects of the present invention are: (1) before evaluating the service quality of the enterprise, the present invention screens the abnormal evaluations of the enterprise, and then conducts a follow-up visit to the abnormal evaluations after the screening, and combines the data after the follow-up visit to comprehensively evaluate the service quality of the enterprise, thereby ensuring the reliability of the relevant data of the enterprise's service quality evaluation, improving the accuracy of the enterprise's service quality evaluation, and improving the scientific nature of the enterprise's decision-making, which is conducive to the continuous improvement of the enterprise.

[0020] (2) The present invention conducts return visits to abnormal evaluations after screening in proportion and confirms the return visit method, thereby increasing the feedback rate of evaluation users, improving the return visit efficiency, avoiding waste of resources, and improving the competitiveness of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0022] Figure 1 It is a schematic diagram of the system structure connection of the present invention.

[0023] Figure 2 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0024] 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.

[0025] Reference Figure 1 As shown, the first aspect of the present invention provides an enterprise service quality evaluation and optimization system including: a priori data acquisition module, an abnormal evaluation group label generation module, an abnormal evaluation group processing module, a feedback terminal, a quality evaluation module, a Web integrated display terminal and a data warehouse.

[0026] It should be noted that the prior data acquisition module is connected to the abnormal evaluation group label generation module, the abnormal evaluation group label generation module is respectively connected to the abnormal evaluation group processing module and the quality assessment module, the abnormal evaluation group processing module is connected to the feedback terminal, the feedback terminal is connected to the quality assessment module, the quality assessment module is connected to the Web display terminal, and the data warehouse is respectively connected to the abnormal evaluation group label generation module, the abnormal evaluation group processing module and the quality assessment module.

[0027] The prior data acquisition module is used to acquire the prior evaluation data of the enterprise from the enterprise operation platform.

[0028] In a specific embodiment of the present invention, the prior evaluation data includes characteristic parameters of each indicator, text of each evaluation, picture set, time point and account ID.

[0029] It should be noted that the characteristic parameters of each indicator are specifically obtained based on the type of enterprise. For example, if the type of enterprise is in the medical field, the characteristic parameters of each indicator include but are not limited to health indicators such as patient survival rate and disease cure rate. If the type of enterprise is in the environmental science field, the characteristic parameters of each indicator include but are not limited to environmental indicators such as air quality index and water quality index. If the type of enterprise is in the education field, the characteristic parameters of each indicator include but are not limited to educational outcome indicators such as student average grades, graduation rate and employment rate.

[0030] The abnormal evaluation group label generation module is used to generate the abnormal evaluation group label and the regular evaluation group label of the enterprise based on the prior evaluation data of the enterprise.

[0031] In a specific embodiment of the present invention, the abnormal evaluation group label and the regular evaluation group label of the enterprise are generated by: based on the text, picture and account ID of each evaluation in the prior evaluation data of the enterprise, the confidence level coefficient α of each evaluation of the enterprise is obtained through processing. _i, i is the number of each evaluation, i = 1, 2, ..., n.

[0032] If α _i ≥α′, then the evaluation is recorded as a regular evaluation group, where α′ is the confidence level convergence coefficient stored in the data warehouse.

[0033] If (α _i <α′)∧(α _i ∈β _m ), then this evaluation is recorded as the mth level group of abnormal evaluation, where β _m It is the confidence level coefficient interval corresponding to the mth level group label of abnormal evaluation stored in the data warehouse.

[0034] Summarize all regular evaluation groups to generate the regular evaluation group label of the enterprise, and summarize all levels of abnormal evaluation groups to generate the abnormal group label of the enterprise.

[0035] In a specific embodiment of the present invention, the confidence level coefficient α of each evaluation of the enterprise is _i The specific evaluation method is as follows: based on the text of each evaluation of the enterprise, the number of text keywords of each evaluation of the enterprise is summarized, and a text keyword set of each evaluation of the enterprise is generated, and the mode of the number of text keywords of the enterprise is obtained. Taking a certain evaluation as an example, the text overlap rate of this evaluation and the rest of the evaluations is evaluated. Combined with the number of text keywords SI and the mode SI′ of the number of text keywords in the enterprise's evaluation, the first confidence level coefficient of the evaluation is obtained through data processing. Where E _0 is the text keyword set of this evaluation, E′ is the keyword set corresponding to the enterprise stored in the data warehouse, and m is the number of the remaining evaluations, m=1,2,...,l.

[0036] Based on the picture sets of each evaluation of the enterprise, obtain the picture sets of the enterprise's evaluation, and evaluate the second confidence level coefficient αI of the enterprise's evaluation _1 .

[0037] It should be noted that the second confidence level coefficient αI of the evaluation of the evaluated enterprise is _1 The specific evaluation method is: based on several pictures in the picture set of the enterprise's evaluation and several pictures in the picture sets of other evaluations, through the image analysis tool, the similarity between each picture in the picture set of the enterprise's evaluation and each picture in the picture sets of other evaluations is output, and the similarity between the enterprise's evaluation and the picture sets of other evaluations is obtained by averaging, and the similarity is obtained by averaging again and then negating it to obtain the second confidence level coefficient of the enterprise's evaluation.

[0038] Based on the account ID of each evaluation of the enterprise, the corresponding evaluation of each account ID of the enterprise is mapped, and the time point T of each evaluation corresponding to each account ID of the enterprise is obtained _pj , get the evaluation frequency PI of each account id of the enterprise _p , after numerical processing, the confidence correction parameters of each account ID of the enterprise are obtained Where T′ and PI′ are the preset appropriate evaluation interval and evaluation frequency of the enterprise, p is the number of each account id, p=1,2,...,q, j is the number of each evaluation corresponding to the account id, j=1,2,...,k.

[0039] It should be noted that the evaluation frequency refers specifically to the total number of evaluations per unit time.

[0040] Based on the account ID of the enterprise's evaluation and the confidence correction parameter of each account ID, the confidence correction parameter γ of the enterprise's evaluation is matched _0 , and import the confidence correction parameter, the first confidence level coefficient and the second confidence level coefficient of the enterprise's evaluation into the confidence level evaluation model α _0 =γ _0 *(αI _0 *λ _1 +αI _1 *λ _2 ), where λ _1 , _2 The weight factors corresponding to the preset first confidence level coefficient and second confidence level coefficient are output as the confidence level coefficient of the enterprise's current evaluation. Similarly, the confidence level coefficients of each evaluation of the enterprise are obtained through numerical processing.

[0041] It should be noted that the weight factors corresponding to the preset first confidence level coefficient and second confidence level coefficient are specifically values ​​of 0-1, reflecting the degree of influence of the first confidence level coefficient and the second confidence level coefficient on the confidence level coefficient, and are specifically set by the company's supervisors.

[0042] The abnormal evaluation group processing module is used to obtain the account ID corresponding to the abnormal evaluation group labels of each level from the abnormal evaluation group labels of the enterprise, and determine the number of evaluation extractions of the abnormal evaluation group labels of each level of the enterprise, randomly screen and obtain the extracted evaluations of the abnormal evaluation group labels of each level of the enterprise, and obtain the confirmation method of each account ID of the abnormal evaluation group labels of each level of the enterprise after processing.

[0043] It should be noted that the specific method for determining the number of evaluation extractions for each level group label of abnormal evaluation of an enterprise is as follows:

[0044] The total number of evaluations of the enterprise is summarized and multiplied by the appropriate extraction ratio stored in the data warehouse to obtain the total number of appropriate extractions for the enterprise.

[0045] The number of evaluations of each level group label of the enterprise's abnormal evaluation is obtained, and it is multiplied by the total number of suitable extractions of the enterprise to obtain the number of evaluation extractions of each level group label of the enterprise's abnormal evaluation.

[0046] In a specific embodiment of the present invention, the confirmation method of each extracted evaluation of each level group label of the abnormal evaluation of the enterprise is as follows: obtaining historical return visit records from the enterprise operation platform, wherein the historical return visit records include the participation rate of the feedback audio of each account id and the participation rate of the feedback questionnaire, based on the account id of each extracted evaluation of each level group label of the abnormal evaluation of the enterprise, screening the participation rate Y1 of the feedback audio of each extracted evaluation of each level group label of the abnormal evaluation of the enterprise _hb and the participation rate of the feedback questionnaire Y2 _hb , and import it into the validation mode evaluation model Output the abnormal evaluation of the enterprise. Confirmation method A for each extracted evaluation of each level group label _hb , h is the number of the group label of each level of abnormal evaluation, h = 1, 2, ..., g, b is the number of each extracted evaluation, b = 1, 2, ..., d.

[0047] It should be noted that when A _hb The output is 001, which means that the enterprise's abnormal evaluation of the level group label is confirmed by the feedback audio. _hb The output is 002, which means that the confirmation method for the extracted evaluation of the abnormal evaluation of the enterprise for this level group label is the feedback questionnaire.

[0048] The feedback terminal is used to provide feedback based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and collect the enterprise's abnormal feedback data.

[0049] In a specific embodiment of the present invention, the abnormal feedback data of the enterprise includes a feedback data set of each abnormal evaluation, wherein the feedback data set is a feedback audio or a feedback questionnaire.

[0050] The present invention conducts return visits according to the proportion of abnormal evaluations after screening, and confirms the return visit method, thereby improving the feedback rate of evaluation users, improving the return visit efficiency, avoiding resource waste, and improving the competitiveness of enterprises.

[0051] The quality assessment module is used to assess the enterprise's service quality parameters based on the enterprise's abnormal feedback data and conventional evaluation group labels, and obtain the enterprise's service quality optimization level after processing.

[0052] In a specific embodiment of the present invention, the specific evaluation method of the enterprise's service quality parameter is: the characteristic parameter β of each indicator in the enterprise's prior evaluation data is _f Import into the first service quality parameter evaluation model and output the first service quality parameter of the enterprise Where β _ ' f is the preset reference characteristic parameter of the fth indicator, f is the indicator number, f=1,2,...,t.

[0053] The abnormal feedback data from the enterprise includes a feedback data set of each abnormal evaluation, wherein the feedback data set is a feedback audio or a feedback questionnaire, and the characteristic value of each abnormal evaluation of the enterprise is determined, and each abnormal evaluation with a characteristic value of 1 is eliminated to obtain each abnormal evaluation after elimination, which is recorded as each false-proof evaluation, and the type of each false-proof evaluation of the enterprise is determined.

[0054] Each regular evaluation is extracted from the regular evaluation group label of the enterprise, and the type of each regular evaluation of the enterprise is determined.

[0055] It should be noted that the method for confirming the type of each false evaluation of the enterprise is consistent to determine the type of each regular evaluation of the enterprise.

[0056] Based on the types of each conventional evaluation and each type of false evaluation of the enterprise, the types of each evaluation to be evaluated of the enterprise are summarized, and the types of each evaluation to be evaluated of the enterprise of each type are mapped to obtain the total number of evaluations to be evaluated of each type of the enterprise RI is summarized. _x , based on the enterprise's regular evaluation group labels, the total number of regular evaluation groups UI is obtained, and the total number of evaluations of the enterprise UI′ is summarized. After numerical processing, the enterprise's service quality parameters are obtained Where δ _x It is the service quality parameter of the unit evaluation quantity corresponding to the x-th type stored in the data warehouse, x is the number of each type, x = 1, 2, ..., y.

[0057] In a specific embodiment of the present invention, the specific method for determining the characteristic value of each abnormal evaluation of the enterprise is as follows: if the feedback data set of a certain abnormal evaluation of the enterprise is a feedback audio, then obtain several reply keywords of each question text and their corresponding interval lengths ti _rc , construct the keyword set Q of each question text _r , evaluate the characteristic value of the abnormal evaluation of the enterprise Where Q _0 is a keyword set of a question text, Q _'0, QI', HI' respectively represent the abnormal reply keyword set corresponding to a certain question text stored in the data warehouse, the threshold of the overlap rate of the feedback question text, and the risk convergence coefficient of the feedback statement. ∨ and ∧ are logical symbols or and, r is the number of each question text, r = 1, 2, ..., w, c is the number of each reply keyword, c = 1, 2, ..., s.

[0058] If the feedback data set of an abnormal evaluation of an enterprise is a feedback questionnaire, then each feedback option is obtained and compared one by one with each feedback abnormal option stored in the data warehouse. If a certain feedback option is the same as the corresponding abnormal option, the characteristic value of the abnormal evaluation is recorded as 1. If each feedback option is different from the corresponding abnormal option, the characteristic value of the abnormal evaluation is recorded as -1.

[0059] And so on, the characteristic value of each abnormal evaluation of the enterprise is determined.

[0060] In a specific embodiment of the present invention, the specific method for determining the type of each false evaluation of the enterprise is: based on the text keyword set and the picture set of each evaluation of the enterprise, obtain the text keyword set D of each false evaluation of the enterprise _B and picture sets, compare several pictures in the picture sets of each enterprise's de-fake evaluation to obtain the richness FI of the picture sets of each enterprise's de-fake evaluation _B , the quality coefficient of each false evaluation of the enterprise is obtained through numerical processing In the formula, D′ is a preset set of high-quality evaluation keywords, B is the number of each false evaluation, B=1,2,...,D,.

[0061] It should be noted that the specific method for obtaining the richness of the picture set of each de-false evaluation of the enterprise is as follows: based on a number of pictures in the picture set of each de-false evaluation of the enterprise, through an image analysis tool, the similarity between each picture of each de-false evaluation of the enterprise and the remaining pictures is output, and the similarity between each picture of each de-false evaluation of the enterprise and the remaining pictures is obtained, and the similarity between each picture of each de-false evaluation of the enterprise and the remaining pictures is obtained, and the similarity is again processed by the mean and then inverted to obtain the richness of the picture set of each de-false evaluation of the enterprise.

[0062] The evaluation quality coefficient of each false evaluation of the enterprise is compared with the evaluation quality coefficient intervals of various types stored in the data warehouse, and the types of each false evaluation of the enterprise are screened out.

[0063] It should be noted that the various types include excellent, medium and general, which are specifically uploaded and set by the company's staff, and the evaluation quality coefficient ranges of the various types are also specifically uploaded and set by the company's staff.

[0064] Before evaluating the service quality of an enterprise, the present invention screens the abnormal evaluations of the enterprise, and then conducts a return visit to the screened abnormal evaluations, and combines the data after the return visit to comprehensively evaluate the service quality of the enterprise, thereby ensuring the reliability of relevant data of the enterprise's service quality evaluation, improving the accuracy of the enterprise's service quality evaluation, and improving the scientific nature of the enterprise's decision-making, which is conducive to continuous improvement of the enterprise.

[0065] It should be noted that the specific evaluation method of the enterprise's service quality optimization level is: comparing the enterprise's service quality parameters with the service quality parameter ranges corresponding to each service quality optimization level in the data warehouse, and screening out the enterprise's service quality optimization level. The higher the service quality parameters, the lower the service quality optimization level, and the lower the service quality parameters, the higher the service quality optimization level.

[0066] The Web integrated display terminal is used to display the enterprise's service quality parameters and service quality optimization level.

[0067] Reference Figure 2 As shown, the second aspect of the present invention provides a method for executing the enterprise service quality evaluation and optimization system of the present invention, including: obtaining prior data, obtaining the enterprise's prior evaluation data from the enterprise operation platform.

[0068] Abnormal evaluation group label generation, based on the enterprise's prior evaluation data, generates the enterprise's abnormal evaluation group label and regular evaluation group label.

[0069] Abnormal evaluation group processing, obtain the account ID corresponding to the abnormal evaluation group label of each level from the abnormal evaluation group label of the enterprise, and determine the number of evaluation extractions of the abnormal evaluation group label of each level of the enterprise, randomly screen to obtain the extracted evaluations of the abnormal evaluation group label of each level of the enterprise, and obtain the confirmation method of each account ID of the abnormal evaluation group label of each level of the enterprise after processing.

[0070] Feedback processing is based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and the enterprise's abnormal feedback data is collected.

[0071] Quality assessment evaluates the enterprise's service quality parameters based on the enterprise's abnormal feedback data and regular evaluation group labels, and obtains the enterprise's service quality optimization level after processing.

[0072] Web integrated display processing displays the enterprise's service quality parameters and service quality optimization level.

[0073] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. An enterprise service quality evaluation and optimization system, characterized in that: include: A priori data acquisition module is used to obtain priori evaluation data of enterprises; An abnormal evaluation group label generation module is used to generate abnormal evaluation group labels and regular evaluation group labels for an enterprise based on the enterprise's prior evaluation data; The abnormal evaluation group processing module is used to obtain the account ID corresponding to each level of abnormal evaluation group labels from the abnormal evaluation group labels of the enterprise, and determine the number of evaluation extractions of each level of abnormal evaluation group labels of the enterprise, obtain each extracted evaluation of each level of abnormal evaluation group labels of the enterprise, and obtain the confirmation method of each account ID of each level of abnormal evaluation group labels of the enterprise after processing; The feedback terminal is used to provide feedback based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and collect the enterprise's abnormal feedback data; The quality assessment module is used to assess the enterprise's service quality parameters based on the enterprise's abnormal feedback data and regular evaluation group labels, and obtain the enterprise's service quality optimization level after processing; The Web integrated display terminal is used to display the enterprise's service quality parameters and service quality optimization level.

2. According to claim 1, a system for evaluating and optimizing enterprise service quality, characterized in that: The prior evaluation data includes characteristic parameters of each indicator, text of each evaluation, picture set, time point and account ID.

3. According to claim 1, the enterprise service quality evaluation and optimization system is characterized in that: The abnormal feedback data of the enterprise includes a feedback data set of each abnormal evaluation, wherein the feedback data set is a feedback audio or a feedback questionnaire.

4. The enterprise service quality evaluation and optimization system according to claim 2, characterized in that: The specific generation method of the abnormal evaluation group label and the regular evaluation group label of the enterprise is as follows: Based on the text, pictures and account ID of each evaluation in the enterprise's prior evaluation data, the confidence level coefficient α of each evaluation of the enterprise is obtained after processing i , i is the number of each evaluation, i = 1, 2, ..., n; If α _i ≥α′, then the evaluation is recorded as a regular evaluation group, where α′ is the confidence level convergence coefficient stored in the data warehouse; If (α _i <α′)∧(α _i ∈β _m ), then this evaluation is recorded as the mth level group of abnormal evaluation, where β _m is the confidence level coefficient interval corresponding to the mth level group label of abnormal evaluation stored in the data warehouse; Summarize all regular evaluation groups to generate the regular evaluation group label of the enterprise, and summarize all levels of abnormal evaluation groups to generate the abnormal group label of the enterprise.

5. The enterprise service quality evaluation and optimization system according to claim 4, characterized in that: The confidence level coefficient α of each evaluation of the enterprise _i , and its specific evaluation method is: Based on the text of each evaluation of the enterprise, the number of text keywords of each evaluation of the enterprise is summarized, and a text keyword set of each evaluation of the enterprise is generated to obtain the mode of the number of text keywords of the enterprise. Taking a certain evaluation as an example, the text overlap rate CI of this evaluation and the rest of the evaluations is evaluated. _m , and combined with the number of text keywords SI and the mode SI′ of the number of text keywords in the enterprise's evaluation, the first confidence level coefficient of the evaluation is obtained through data processing Where E _0 is the text keyword set of this evaluation, E′ is the keyword set corresponding to the enterprise stored in the data warehouse, and m is the number of the remaining evaluations, m = 1, 2, ..., l; Based on the picture sets of each evaluation of the enterprise, obtain the picture sets of the enterprise's evaluation, and evaluate the second confidence level coefficient αI of the enterprise's evaluation _1 ; Based on the account ID of each evaluation of the enterprise, the corresponding evaluation of each account ID of the enterprise is mapped, and the time point T of each evaluation corresponding to each account ID of the enterprise is obtained _pj , get the evaluation frequency PI of each account id of the enterprise _p , after numerical processing, the confidence correction parameters of each account ID of the enterprise are obtained Where T′ and PI′ are the preset appropriate evaluation interval and evaluation frequency of the enterprise, p is the number of each account id, p = 1, 2, ..., q, j is the number of each evaluation corresponding to the account id, j = 1, 2, ..., k; Based on the account ID of the enterprise's evaluation and the confidence correction parameter of each account ID, the confidence correction parameter γ of the enterprise's evaluation is matched _0 , and import the confidence correction parameter, the first confidence level coefficient and the second confidence level coefficient of the enterprise's evaluation into the confidence level evaluation model α _0 =γ _0 *(αI _0 *λ _1 +αI _1 *λ _2 ), where λ _1 , _2 The weight factors corresponding to the preset first confidence level coefficient and second confidence level coefficient are output as the confidence level coefficient of the enterprise's current evaluation. Similarly, the confidence level coefficients of each evaluation of the enterprise are obtained through numerical processing.

6. The enterprise service quality evaluation and optimization system according to claim 1, characterized in that: The specific confirmation method of each extracted evaluation of each level group label of the abnormal evaluation of the enterprise is as follows: Obtain historical return visit records from the enterprise operation platform, where the historical return visit records include the participation rate of feedback audio and feedback questionnaire of each account id, and the participation rate of feedback audio of each extracted evaluation of each level group label of the enterprise's abnormal evaluation based on the account id of each extracted evaluation of each level group label of the enterprise's abnormal evaluation. _hb and the participation rate of the feedback questionnaire Y2 _hb , and import it into the validation mode evaluation model Output the abnormal evaluation of the enterprise. Confirmation method A for each extracted evaluation of each level group label _hb , h is the number of the group label of each level of abnormal evaluation, h = 1, 2, ..., g, b is the number of each extracted evaluation, b = 1, 2, ..., d.

7. The enterprise service quality evaluation and optimization system according to claim 5, characterized in that: The specific evaluation method of the service quality parameters of the enterprise is as follows: The characteristic parameters β of each indicator in the enterprise's prior evaluation data _f Import into the first service quality parameter evaluation model and output the first service quality parameter of the enterprise Where β _ ' f is the preset reference characteristic parameter of the fth index, f is the index number, f = 1, 2, ..., t; The abnormal feedback data of the enterprise includes a feedback data set of each abnormal evaluation, wherein the feedback data set is a feedback audio or a feedback questionnaire, and the characteristic value of each abnormal evaluation of the enterprise is determined, and each abnormal evaluation with a characteristic value of 1 is eliminated to obtain each abnormal evaluation after elimination, which is recorded as each false-free evaluation, and the type of each false-free evaluation of the enterprise is determined; Extract each regular evaluation from the regular evaluation group label of the enterprise and determine the type of each regular evaluation of the enterprise; Based on the types of each conventional evaluation and each type of false evaluation of the enterprise, the types of each evaluation to be evaluated of the enterprise are summarized, and the types of each evaluation to be evaluated of the enterprise of each type are mapped to obtain the total number of evaluations to be evaluated of each type of the enterprise RI is summarized. _x , based on the enterprise's regular evaluation group labels, the total number of regular evaluation groups UI is obtained, and the total number of evaluations of the enterprise UI′ is summarized. After numerical processing, the enterprise's service quality parameters are obtained Where δ _x It is the service quality parameter of the unit evaluation quantity corresponding to the x-th type stored in the data warehouse, x is the number of each type, x = 1, 2, ..., y.

8. The enterprise service quality evaluation and optimization system according to claim 7, characterized in that: The specific method for determining the characteristic value of each abnormal evaluation of the enterprise is as follows: If the feedback data set of an abnormal evaluation of an enterprise is a feedback audio, then obtain several reply keywords and their corresponding interval time ti of each question text _rc , construct the keyword set Q of each question text _r , evaluate the characteristic value of the abnormal evaluation of the enterprise Where Q _0 is a keyword set of a question text, Q _ '0, QI', HI' respectively represent the abnormal reply keyword set corresponding to a certain question text stored in the data warehouse, the threshold of the overlap rate of the feedback question text, and the risk convergence coefficient of the feedback statement. ∨ and ∧ are logical symbols or and, r is the number of each question text, r = 1, 2, ..., w, c is the number of each reply keyword, c = 1, 2, ..., s; If the feedback data set of a certain abnormal evaluation of an enterprise is a feedback questionnaire, then obtain each feedback option and compare it with each feedback abnormal option stored in the data warehouse one by one. If a certain feedback option is the same as the corresponding abnormal option, then the characteristic value of this abnormal evaluation is recorded as 1. If each feedback option is different from the corresponding abnormal option, then the characteristic value of this abnormal evaluation is recorded as -1. And so on, the characteristic value of each abnormal evaluation of the enterprise is determined.

9. The enterprise service quality evaluation and optimization system according to claim 7, characterized in that: The specific method for determining the type of each false evaluation of the enterprise is as follows: Based on the text keyword set and picture set of each evaluation of the enterprise, obtain the text keyword set D of each false evaluation of the enterprise _B and picture sets, compare several pictures in the picture sets of each enterprise's de-fake evaluation to obtain the richness FI of the picture sets of each enterprise's de-fake evaluation _B , the quality coefficient of each false evaluation of the enterprise is obtained through numerical processing Where D′ is the preset high-quality evaluation keyword set, B is the number of each false evaluation, B = 1, 2, ..., D,; The evaluation quality coefficient of each false evaluation of the enterprise is compared with the evaluation quality coefficient intervals of various types stored in the data warehouse, and the types of each false evaluation of the enterprise are screened out.

10. A method for executing the enterprise service quality evaluation and optimization system according to any one of claims 1 to 9, characterized in that: include: Prior data acquisition: obtaining the enterprise's prior evaluation data from the enterprise operation platform; Abnormal evaluation group label generation: based on the enterprise's prior evaluation data, the abnormal evaluation group label and regular evaluation group label of the enterprise are generated; Abnormal evaluation group processing, obtain the account ID corresponding to each level group label of abnormal evaluation from the abnormal evaluation group label of the enterprise, and determine the number of evaluation extractions of each level group label of abnormal evaluation of the enterprise, randomly screen and obtain each extracted evaluation of each level group label of abnormal evaluation of the enterprise, and obtain the confirmation method of each account ID of each level group label of abnormal evaluation of the enterprise after processing; Feedback processing: feedback is provided based on the confirmation method of each account ID of each level group label of the enterprise's abnormal evaluation, and the abnormal feedback data of the enterprise is collected; Quality assessment, based on the abnormal feedback data and regular evaluation group labels of the enterprise, evaluates the enterprise's service quality parameters, and obtains the enterprise's service quality optimization level after processing; Web integrated display processing displays the enterprise's service quality parameters and service quality optimization level.

Images