A method for follow-up evaluation of community endowment services

Abstract

The application relates to the technical field of smart community management and big data processing, and discloses a follow-up visit evaluation method for community pension services, which comprises the following steps: screening key service quality influence feature weights according to the historical service records of pension service users; analyzing the key service quality influence feature weights by using a deep learning algorithm; matching and calculating the analysis results with real-time collected pension service user feedback data; and storing the evaluation scores corresponding to the calculation results in a hierarchical manner. By using the deep learning algorithm, subtle patterns and correlations are learned from complex data sets, so that more accurate and precise weights are provided for each service quality feature, the dynamic adaptability and long-term effectiveness of the service evaluation model are greatly improved, and the evaluation results can continuously reflect the real level of service quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of smart community management and big data processing, and in particular to a revisit evaluation method for community pension services. BACKGROUND

[0002] Community pension services are one of the key measures to address the challenge of population aging. With the development of social economy and the change of population structure, the demand for pension service quality is becoming higher and higher. Traditional community pension services mainly rely on manual experience to evaluate service quality, which has the disadvantages of low efficiency, strong subjectivity and difficulty in standardization.

[0003] In recent years, with the rapid development of information technology, especially the application of data mining and machine learning technology, the evaluation method of pension service quality has gradually developed towards intelligence and precision. For example, some researches use user feedback data and service records to build evaluation models, which can better reflect the actual effect of the service and user satisfaction.

[0004] However, the existing technology still has deficiencies in processing complex data and extracting key service quality influencing factors. The current evaluation system often ignores the weight distribution of service quality influencing features, resulting in evaluation results that cannot accurately reflect the true situation of the service.

[0005] In addition, the existing technology lacks an effective mechanism to combine historical service data with real-time user feedback, which limits the effectiveness and practicality of the evaluation system in actual application. Therefore, there is an urgent need for a new type of pension service evaluation method that can integrate historical and real-time data and improve evaluation accuracy and efficiency through intelligent methods. SUMMARY

[0006] The purpose of this section is to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, abstract and title. Such simplifications or omissions cannot be used to limit the scope of the present application.

[0007] In view of the above-mentioned existing problems, the present application is proposed.

[0008] Therefore, the technical problem solved by the present application is that the existing community pension service revisit evaluation technology has the problems of low data utilization rate and inaccurate evaluation.

[0009] To solve the above technical problems, the present application provides the following technical solutions: according to the historical service records of the pension service users, the key service quality influencing feature weights are screened out;

[0010] analyzing the key service quality influence feature weight by using a deep learning algorithm;

[0011] matching and calculating the analysis result with real-time collected pension service user feedback data, and storing the evaluation scores corresponding to the calculation results in grades.

[0012] As a preferred scheme of the method for follow-up evaluation of community pension services, the screening of the key service quality influence feature weight comprises:

[0013] Collecting historical service records of the pension service users, the historical service records at least containing service logs, user feedbacks and complaint records;

[0014] Cleaning and formatting the historical service records;

[0015] Using a principal component analysis algorithm to identify key features influencing service quality from the historical service records;

[0016] According to the influence degree of each feature on user satisfaction, assigning weights based on a regression analysis algorithm to obtain the key service quality influence feature weight.

[0017] As a preferred scheme of the method for follow-up evaluation of community pension services, finding the key features comprises:

[0018] Calculating the mean and standard deviation of X to obtain a standardized data matrix Z;

[0019] Calculating the covariance matrix C and solving the eigenvalues and corresponding eigenvectors of C;

[0020] According to the descending order of the eigenvalues, selecting the first m eigenvectors as the key features;

[0021] Wherein, X is an n*p data matrix, representing the cleaned and formatted historical service records, n is the number of records, and p is the number of feature elements.

[0022] As a preferred scheme of the method for follow-up evaluation of community pension services, the key features at least include service attitude, professional ability, response speed, service continuity and personalized service.

[0023] As a preferred scheme of the method for follow-up evaluation of community pension services, assigning weights based on a regression analysis algorithm can be represented by the following formula:

[0024]

[0025]

[0026] wherein, Z ij is the standardized value, t ik is the score of the i-th record on the k-th principal component, u jk is the score of the j-th feature element on the k-th principal component, w k is the weight of the k-th principal component, y i is the user satisfaction, is the error term.

[0027] As a preferred scheme of the revisit evaluation method for community pension services, the analysis can be represented by the following mathematical formula:

[0028]

[0029]

[0030] wherein, h i is the output of the i-th layer, W i is the weight matrix of the i-th layer, b i is the bias vector of the i-th layer, is the activation function, F is the output vector, representing the estimation of the weight of each feature.

[0031] As a preferred scheme of the revisit evaluation method for community pension services, the value range of the output vector F is [0, 1], wherein:

[0032] The closer the value is to 1, the greater the impact of the feature on the service quality;

[0033] The closer the value is to 0, the smaller the impact of the feature on the service quality.

[0034] As a preferred scheme of the revisit evaluation method for community pension services, the evaluation scores corresponding to the calculation results are stored in grades, including:

[0035] The real-time collected pension service user feedback data is normalized to make the data in the same order of magnitude;

[0036] The normalized feedback data is combined with the feature weight using a weighted summation algorithm to obtain the evaluation score of each element;

[0037] The evaluation score is stored in grades.

[0038] The beneficial effects of the present application: the present application learns subtle patterns and correlations from complex data sets through deep learning algorithms, thereby providing more refined and accurate weights for each quality of service feature, greatly improving the dynamic adaptability and long-term effectiveness of the service evaluation model, so that the evaluation results can continuously reflect the true level of service quality; not only can the service quality be more accurately evaluated, but also can provide a basis for service providers to improve services, thereby significantly improving the overall quality and efficiency of community care services. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0040] Figure 1 The flowchart of the method for follow-up evaluation of community care services shown in the present application. DETAILED DESCRIPTION

[0041] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present application, not all embodiments.

[0042] Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0043] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0044] According to the embodiments of the present application, combined with the flowchart shown in Figure 1 The method for follow-up evaluation of community care services, specifically comprising the following steps:

[0045] S1, according to the historical service records of the elderly service users, screening out the key service quality influence feature weight. It should be noted that in this step, screening out the key service quality influence feature weight includes:

[0046] Collecting the historical service records of the elderly service users, the historical service records at least containing service logs, user feedback, complaint records;

[0047] cleaning and formatting the historical service records;

[0048] identifying key features affecting service quality from the historical service records using a principal component analysis algorithm;

[0049] assigning weights based on a regression analysis algorithm according to the degree of influence of each feature on user satisfaction, to obtain key service quality impact feature weights.

[0050] Further, the key features include:

[0051] calculating the mean and standard deviation of X to obtain a standardized data matrix Z;

[0052] calculating the covariance matrix C and solving the eigenvalues and corresponding eigenvectors of C;

[0053] sorting the eigenvalues from large to small and selecting the first m eigenvectors as the key features;

[0054] where X is an n*p data matrix representing the cleaned and formatted historical service records, n is the number of records, and p is the number of feature elements.

[0055] As an example, the weights assigned based on the regression analysis algorithm can be represented by the following formula:

[0056]

[0057]

[0058] where Z is the standardized value, t is the score of the i-th record on the k-th principal component, u is the score of the j-th feature element on the k-th principal component, w is the weight of the k-th principal component, y is the user satisfaction, and e is the error term. ij ik jk k i

[0059] In optional embodiments, the key features include at least service attitude, professional ability, response speed, service continuity, and personalized service.

[0060] In optional embodiments, the historical service records include at least service logs, user feedback, complaint records, service effectiveness tracking, and service personnel change records.

[0061] As an example, the service logs are used to record the date, time, service personnel, and service content of each service.

[0062] ​​​​​​As an example, user feedback is used to record user's satisfaction evaluation of each service, and improvement suggestions.

[0063] As an example, complaint records are used to record detailed records of user's dissatisfaction or complaint events of the service.

[0064] As an example, service effect tracking is used to track and record user's health status, quality of life, etc. for a long time, and to evaluate the long-term effect of the service.

[0065] As an example, service personnel change record is used to record the replacement of service personnel, so as to evaluate the impact of personnel change on service quality.

[0066] Preferably, this step can accurately identify and screen out key features affecting service quality, such as response time, service attitude, professional ability, and assign corresponding weights to these features, to ensure that the evaluation and calculation process is more in line with the actual service situation, thereby improving the accuracy and effectiveness of the evaluation.

[0067] S2, using deep learning algorithm to analyze the weight of key service quality impact features. It should be noted that this step can be represented by the following mathematical formula:

[0068]

[0069]

[0070] where h i is the output of the i-th layer, W i is the weight matrix of the i-th layer, b i is the bias vector of the i-th layer, is the activation function, F is the output vector, which represents the estimation of each feature weight.

[0071] As an example, the value range of output vector F is [0, 1], wherein:

[0072] The closer the value is to 1, the greater the impact of the feature on service quality;

[0073] The closer the value is to 0, the smaller the impact of the feature on service quality.

[0074] Preferably, this step applies deep learning algorithm to further analyze the weight of key service quality impact features, which not only enhances the prediction and evaluation accuracy of the model, but also automatically adjusts and optimizes the weight distribution, greatly improves the dynamic adaptability and long-term effectiveness of the service evaluation model, so that the evaluation result can continuously reflect the true level of service quality.

[0075] S3, match and calculate the analysis result with the real-time collected pension service user feedback data, and store the corresponding evaluation score according to the calculation result. It should be pointed out that the step of storing the corresponding evaluation score according to the calculation result includes:

[0076] The real-time collected pension service user feedback data is normalized to make the data in the same order of magnitude;

[0077] The normalized feedback data is combined with the feature weight by using the weighted summation algorithm to obtain the evaluation score of each element;

[0078] The evaluation score is stored in a hierarchical manner.

[0079] As an example, the evaluation score can be represented by the following formula:

[0080]

[0081] By adding an exponential function to adjust the sensitivity of the evaluation score, the evaluation score calculation result is more in line with the actual situation, which can be represented by the following mathematical formula:

[0082]

[0083] Based on the adjusted V i The evaluation score is stored in a hierarchical manner, and the mathematical expression formula is as follows:

[0084]

[0085] Where S i is the original evaluation score of the i-th sample, k and c are the adjustment parameters of the exponential function, which are used to control the shape and position of the function, F j is the output vector of the j-th feature, R is the normalized user feedback data matrix, R ij is the original feedback of the i-th sample to the j-th feature, V i is the adjusted evaluation score, is the hierarchical threshold of the evaluation score.

[0086] As an example, the value range of V i is [0, 1], where a value close to 0 indicates poor service quality, and a value close to 1 indicates excellent service quality.

[0087] Preferably, the embodiment combines the analysis result with the real-time collected user feedback data, obtains the final service quality evaluation score through matching calculation, and stores the scores in a hierarchical manner, so that the evaluation result can reflect the current satisfaction and service experience of the user in real time, increasing the timeliness and practicality of the evaluation.

[0088] Preferably, the hierarchical storage method facilitates quick identification of different levels of service quality, and facilitates managers to make targeted service improvements or develop differentiated service strategies based on specific evaluation results, thereby improving overall service level and enhancing user satisfaction.

[0089] It should be further noted that the conventional method usually relies on a single type of service data (such as only user feedback or service logs), lacks multi-dimensional comprehensive analysis, and the weight of service quality characteristics is usually fixed and cannot be dynamically adjusted according to actual service changes, thereby ignoring the importance of evaluating user personalized needs.

[0090] In order to better verify the technical effects adopted in the method of the present application, the present embodiment selects to compare and test the conventional method and the method of the present application, compares the test results by scientific demonstration means, and verifies the real effects of the present method.

[0091] (1) Experimental conditions

[0092] Collect 1000 historical service records, each record containing service logs, user feedback, complaint records, service effect tracking and service personnel change records;

[0093] Five key features: service attitude, professional ability, response speed, service continuity and personalized service;

[0094] Use a high-performance computer with NVIDIA Tesla V100 GPU, equipped with Python 3.8 and TensorFlow2.4 for deep learning algorithm implementation;

[0095] Principal component analysis uses SciPy library, and deep learning model construction and training uses TensorFlow framework.

[0096] (2) Experimental process

[0097] Clean the data, including removing invalid records and formatting date and time;

[0098] Standardize the data to make different types of data quantifiable for comparison;

[0099] Use principal component analysis (PCA) to extract key features from historical service records;

[0100] Calculate the covariance matrix of the features, analyze the eigenvalues and eigenvectors, and select the eigenvectors with the highest contribution;

[0101] Use regression analysis algorithm to assign weights according to the influence of each feature on user satisfaction;

[0102] A deep learning algorithm is applied to further optimize the feature weights;

[0103] In combination with real-time collected user feedback data, a weighted summation algorithm is used to calculate the evaluation score of service quality;

[0104] The sensitivity of the evaluation score is adjusted by an exponential adjustment function to ensure that the evaluation result conforms to the actual situation;

[0105] The evaluation result is stored hierarchically to facilitate quick identification and response to changes in service quality.

[0106] (3) Experimental results

[0107] Table 1, comparison of score results

[0108] Test subjects Service attitude Professional ability Response speed Service continuity Personalized service User satisfaction Final score Traditional method 0.80 0.82 0.78 0.85 0.79 0.84 0.82 Method of the present invention 0.92 0.91 0.89 0.94 0.90 0.93 0.92

[0109] Referring to Table 1, it can be seen that the scores of each key feature of the method of the present application are higher than those of the traditional method, especially in the significant improvement in personalized service and response speed, which indicates that the method of the present application can more accurately identify and respond to the actual needs of users, effectively improving the overall quality of the elderly service and user satisfaction.

[0110] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, which should be covered by the scope of the claims of the present application.

Claims

1. A revisit evaluation method for community care services, characterized by, The method comprises the following steps: Screening key service quality impact feature weights according to the historical service records of the elderly service users; Collecting the historical service records of the elderly service users, which at least include service logs, user feedbacks, and complaint records; Cleaning and formatting the historical service records; Identifying key features that impact service quality from the historical service records by using principal component analysis algorithm; Assigning weights based on regression analysis algorithm according to the impact degree of each feature on user satisfaction, to obtain the key service quality impact feature weights; Analyzing the key service quality impact feature weights by using deep learning algorithm; the analysis can be represented by the following mathematical formula: wherein, is the output of the layer, is the weight matrix of the layer, is the bias vector of the layer, is the activation function, F is the output vector, and represents an estimate of the weight of each feature. The value range of the output vector F is [0, 1], wherein: The closer the value is to 1, the greater the impact of the feature on service quality; The closer the value is to 0, the smaller the impact of the feature on service quality; Matching and calculating the analysis results with real-time collected elderly service user feedback data, and storing the evaluation scores corresponding to the calculation results in grades.

2. The method of claim 1, wherein, Finding the key features includes: Calculating the mean and standard deviation of X to obtain the standardized data matrix Z; Calculating the covariance matrix C and solving the eigenvalues and corresponding eigenvectors of C; According to the descending order of the eigenvalues, selecting the first m eigenvectors as the key features; Wherein, X is an n*p data matrix, representing the cleaned and formatted historical service records, n is the number of records, and p is the number of feature elements.

3. The method of claim 2, wherein, The key features at least include service attitude, professional ability, response speed, service continuity, and personalized service.

4. The method of claim 1, wherein, Assigning weights based on regression analysis algorithm can be represented by the following formula: where Z ij is the standardized value, t ik is the score of the i-th record on the k-th principal component, u jk is the score of the j-th feature element on the k-th principal component, w k is the weight of the k-th principal component, y i is the user satisfaction, is the error term.

5. The method of claim 1, wherein, Storing the evaluation scores corresponding to the calculation results in grades includes: Normalizing the real-time collected elderly service user feedback data to make the data in the same order of magnitude; Combining the normalized feedback data with the feature weights by using weighted summation algorithm to obtain the evaluation score of each element; Storing the evaluation scores in grades.

Images