A business analysis method and system based on evaluation indicators

By constructing evaluation indicators and using natural language processing technology, combined with a pre-set model, power business assessment was conducted, solving the problem of low accuracy in single text matching assessment and achieving efficient and accurate business analysis.

CN123117876APending Publication Date: 2026-11-17GUANGDONG POWER GRID CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510962136.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-14
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing technologies, using a single text matching as an evaluation and analysis indicator is difficult to reflect the actual business needs of users, resulting in low evaluation accuracy and large errors, and failing to meet the diverse and personalized power service needs of users.

Method used

By acquiring business service information and user demand information, natural language processing and aggregation algorithms are used to construct evaluation indicators. Combined with a pre-set evaluation model, business scores are calculated, data tracking and process information acquisition are carried out, and a business evaluation report is generated.

Benefits of technology

It enables comprehensive evaluation based on the actual needs of different users, improving evaluation accuracy, simplifying the evaluation process, shortening processing time, and increasing evaluation and analysis efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117876A_ABST
    Figure CN123117876A_ABST
Patent Text Reader

Abstract

This invention discloses a business analysis method and system based on evaluation indicators. The method includes: acquiring business service information and evaluation indicator information, wherein the business service information is service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information; calling a preset evaluation model to calculate a business score value based on the business service information and the evaluation indicator information, and performing data tracking based on the numerical value of the business score value to obtain business process information; performing evaluation processing based on the time and value of the business process information to obtain a business evaluation result, and generating a business evaluation report based on the business evaluation result and the business score value. This invention performs comprehensive evaluation and analysis based on different evaluation indicators, which can meet the actual needs of different users, avoid the error of a single evaluation indicator, thereby reducing evaluation bias and improving evaluation accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the technical field of business forecasting and analysis, and in particular to a business analysis method and system based on evaluation indicators. Background Technology

[0002] As the global energy industry accelerates its digital transformation, the power sector is also facing unprecedented opportunities for change. The widespread application of technologies such as the Internet of Things, big data, and cloud computing provides strong technical support for intelligent management in the power industry. At the same time, users' demands for electricity services are becoming increasingly diversified and personalized, requiring power companies to continuously improve service quality to meet market demands.

[0003] As users' business service demands increase, different business service items need to be continuously updated. Therefore, after an update is completed, it needs to be evaluated and analyzed to determine whether the updated business services meet user needs. One common method is to use a matching model to identify the text of the updated business service items and the user's business requirement items. Then, a text matching algorithm is used to match the text of the service items with the text of the requirement items one by one. The matching results determine whether the business services meet different user needs and generate an analysis and evaluation report, which technical personnel can then use to make subsequent business adjustments.

[0004] However, the commonly used methods have the following technical problems: different users have different business needs, and the forms of services are constantly being updated. Using a single text matching as an evaluation and analysis indicator is difficult to reflect the actual situation of users and may deviate from the actual needs of users, thus reducing the accuracy of the evaluation and resulting in a large evaluation error. Summary of the Invention

[0005] This invention proposes a business analysis method and system based on evaluation indicators, which can solve one or more of the above-mentioned technical problems.

[0006] A first aspect of this invention provides a business analysis method based on evaluation indicators, characterized in that the method includes: The business service information and evaluation indicator information are obtained separately. The business service information is the service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information. The system calls a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information, and performs data tracking based on the value of the business score to obtain business process information. The business process information is evaluated based on time and numerical values ​​to obtain a business evaluation result, and a business evaluation report is generated based on the business evaluation result and the business score.

[0007] In one possible implementation of the first aspect, obtaining the evaluation index information includes: Obtain multiple user demand information, where each user demand information represents the business service needs of a different user; A natural language processing algorithm is used to split each user request information into multiple request texts. The multiple requirement texts are aggregated and classified using an aggregation algorithm to obtain multiple requirement categories, each requirement category including one or more requirement texts; The classification type of each of the aforementioned demand categories is determined, and evaluation index information is obtained.

[0008] In one possible implementation of the first aspect, the step of invoking a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information includes: A preset quality evaluation model is invoked to perform quality assessment processing on the business service information and the evaluation indicator information to obtain a quality evaluation value. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item and the category weight value of each demand category of the business service information; The business score is calculated using the category weight value and the service weight value.

[0009] In one possible implementation of the first aspect, the step of tracking data based on the numerical value of the business score to obtain business process information includes: If the business score is greater than the preset score, then read the project text and service time nodes of different business service items contained in the business service information. Use the tags of several project texts to search several corresponding business databases; Based on the service time node, business processing data corresponding to the time is extracted from each of the business databases to obtain business process information.

[0010] In one possible implementation of the first aspect, the evaluation process based on the time and value of the business process information to obtain the business evaluation result includes: The service duration is calculated based on the service time node, and the service value is extracted from the service processing data; The service duration is compared with the corresponding preset processing duration, and the service value is compared with the corresponding preset parameter range. The business evaluation result is determined based on the comparison results of the time comparison and the comparison structure of the numerical comparison.

[0011] In one possible implementation of the first aspect, the step of performing data tracking processing on the business database based on the numerical value of the business score to obtain business process information includes: If the business score is less than the preset score, then the preset Bayesian optimization model is invoked to generate business optimization information based on the business service information; Adjust the business service items based on the optimization information.

[0012] In one possible implementation of the first aspect, after the step of generating a business evaluation report based on the business evaluation results and the business score, the method further includes: In response to the display request, the business evaluation report is converted into a dynamic profile, and the dynamic profile is then visualized and displayed to technical personnel.

[0013] A second aspect of this invention provides a business analysis system based on evaluation metrics, the system comprising: The acquisition module is used to acquire business service information and evaluation indicator information respectively. The business service information is the service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information. The calculation module is used to call a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information, and to perform data tracking based on the value of the business score to obtain business process information. The analysis module is used to evaluate the business process information based on time and values ​​to obtain business evaluation results, and to generate a business evaluation report based on the business evaluation results and the business score.

[0014] In one possible implementation of the second aspect, obtaining the evaluation index information includes: Obtain multiple user demand information, where each user demand information represents the business service needs of a different user; A natural language processing algorithm is used to split each user request information into multiple request texts. The multiple requirement texts are aggregated and classified using an aggregation algorithm to obtain multiple requirement categories, each requirement category including one or more requirement texts; The classification type of each of the aforementioned demand categories is determined, and evaluation index information is obtained.

[0015] In one possible implementation of the second aspect, the step of invoking a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information includes: A preset quality evaluation model is invoked to perform quality assessment processing on the business service information and the evaluation indicator information to obtain a quality evaluation value. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item and the category weight value of each demand category of the business service information; The business score is calculated using the category weight value and the service weight value.

[0016] In one possible implementation of the second aspect, the step of tracking data based on the numerical value of the business score to obtain business process information includes: If the business score is greater than the preset score, then read the project text and service time nodes of different business service items contained in the business service information. Use the tags of several project texts to search several corresponding business databases; Based on the service time node, business processing data corresponding to the time is extracted from each of the business databases to obtain business process information.

[0017] In one possible implementation of the second aspect, the evaluation process based on the time and numerical values ​​of the business process information to obtain the business evaluation result includes: The service duration is calculated based on the service time node, and the service value is extracted from the service processing data; The service duration is compared with the corresponding preset processing duration, and the service value is compared with the corresponding preset parameter range. The business evaluation result is determined based on the comparison results of the time comparison and the comparison structure of the numerical comparison.

[0018] In one possible implementation of the second aspect, the step of performing data tracking processing on the business database based on the numerical value of the business score to obtain business process information includes: If the business score is less than the preset score, then the preset Bayesian optimization model is invoked to generate business optimization information based on the business service information; Adjust the business service items based on the optimization information.

[0019] In one possible implementation of the second aspect, the system further includes: The visualization module is used to respond to a display request after the step of generating a business evaluation report based on the business evaluation results and the business score, convert the business evaluation report into a dynamic profile, and visualize the dynamic profile for technical personnel.

[0020] Compared to existing technologies, the business analysis method and system based on evaluation indicators provided by this invention have the following advantages: This invention can acquire business service information and evaluation indicator information separately; it calls a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information, and performs data tracking based on the numerical value of the business score to obtain business process information; it performs evaluation processing based on the time and value of the business process information to obtain a business evaluation result, and generates a business evaluation report based on the business evaluation result and the business score. This invention performs comprehensive evaluation and analysis based on different evaluation indicators, which can meet the actual needs of different users, avoid the errors of a single evaluation indicator, thereby reducing evaluation bias and improving evaluation accuracy; moreover, different evaluation indicators are extracted from the needs information of different users, allowing for the evaluation of multiple users' needs at once, simplifying the process, shortening the evaluation processing time, and improving the efficiency of evaluation analysis. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a business analysis method based on evaluation indicators provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a business analysis system based on evaluation indicators provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] As the global energy industry accelerates its digital transformation, the power sector is also facing unprecedented opportunities for change. The widespread application of technologies such as the Internet of Things, big data, and cloud computing provides strong technical support for intelligent management in the power industry. At the same time, users' demands for electricity services are becoming increasingly diversified and personalized, requiring power companies to continuously improve service quality to meet market demands.

[0024] As users' business service demands increase, different business service items need to be continuously updated. Therefore, after an update is completed, it needs to be evaluated and analyzed to determine whether the updated business services meet user needs. One common method is to use a matching model to identify the text of the updated business service items and the user's business requirement items. Then, a text matching algorithm is used to match the text of the service items with the text of the requirement items one by one. The matching results determine whether the business services meet different user needs and generate an analysis and evaluation report, which technical personnel can then use to make subsequent business adjustments.

[0025] However, the commonly used methods have the following technical problems: different users have different business needs, and the forms of services are constantly being updated. Using a single text matching as an evaluation and analysis indicator is difficult to reflect the actual situation of users and may deviate from the actual needs of users, thus reducing the accuracy of the evaluation and resulting in a large evaluation error.

[0026] To address the aforementioned issues, the following specific embodiments will provide a detailed introduction and explanation of a business analysis method and system based on evaluation indicators provided in this application.

[0027] Reference Figure 1 The diagram shows a flowchart of a business analysis method based on evaluation indicators provided by an embodiment of the present invention.

[0028] In an optional embodiment, the business analysis method based on evaluation metrics is applicable to the data management platform of a business service enterprise, which can analyze and evaluate business services. Analyzing and evaluating through the platform simplifies the evaluation and analysis process and improves efficiency.

[0029] As an example, the business analysis method based on evaluation indicators may include: S11. Obtain business service information and evaluation indicator information respectively, wherein the business service information is service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information.

[0030] In one embodiment, business service information can be obtained, wherein the business service information is service information of different business service items.

[0031] It should be noted that each service can contain several service items, and each service item can contain different service information. For example, the service of reporting power outages may include: fault reporting, fault repair, repair cost settlement, equipment registration, and after-sales maintenance. Each service item contains different service content.

[0032] In one embodiment, evaluation indicator information can also be obtained, wherein the evaluation indicator information is an evaluation indicator of business services. This evaluation indicator can be constructed based on the obtained user business requirement information.

[0033] In one operational scenario, before providing a service to a user, the system can obtain the user's business needs information and then provide the service based on those needs. After completing the service, the system can obtain service information for that operation and, based on the user's business needs information, construct evaluation metrics. These metrics are then used to assess and analyze the service requirements to determine if the service met the user's needs, and the system can then optimize or modify the service based on the assessment results.

[0034] In another operational scenario, after updating a business service, updated business service information can be obtained. To determine whether the updated business service meets the user's requirements, the updated business service can be provided to different users. These different users can be users randomly selected by the enterprise from among its previously served users, and these previously served users were those who received the service before the update. Referring to the example above, if the updated business service is a power outage reporting service, multiple users can be randomly selected from those who previously requested power outage reporting services.

[0035] Simultaneously, the needs of randomly selected users can be obtained. Next, updated business services can be provided to users, specifically through simulated service provision. After the service is completed, business service information for this operation can be obtained. Furthermore, evaluation indicators can be constructed based on the different business needs of different users, resulting in evaluation indicator information. This evaluation indicator information is then used to assess and analyze the business needs to determine whether the service meets user requirements. Based on the evaluation results, the updated service can be optimized or modified.

[0036] Existing technologies involve calling models or algorithms to match service item information with each user's business needs one by one, which is time-consuming and inefficient. By acquiring the needs information of different users to construct evaluation indicators for evaluating and analyzing business services, each operation can be combined with multiple business needs information, which can shorten the processing time and improve the processing efficiency.

[0037] Based on this, combining different evaluation indicators for comprehensive assessment can avoid the bias caused by a single evaluation and analysis indicator. Moreover, different comprehensive indicators can meet the actual needs of users, thereby improving the accuracy of the assessment and reducing the assessment error.

[0038] In one embodiment, different users may have different needs. In order to construct different evaluation indicators based on the needs of different users, obtaining the evaluation indicator information, as an example, may include the following sub-steps: S111. Obtain multiple user demand information, where each user demand information represents the business service demand of a different user.

[0039] S112. Call a natural language processing algorithm to split each user requirement information into multiple requirement texts.

[0040] S113. Use an aggregation algorithm to aggregate and classify multiple requirement texts to obtain multiple requirement categories, each requirement category including one or more requirement texts.

[0041] S114. Determine the classification type for each of the aforementioned demand categories to obtain evaluation index information.

[0042] In one embodiment, demand information from different users for the same service can be obtained, resulting in multiple user demand information sets. Each user demand information set corresponds to one user, and each user demand information set corresponds to the same service. For example, for a power outage repair service, demand information from 10 users can be collected, resulting in 10 user demand information sets. Similarly, for a power outage query service, demand information from 20 users can be collected, resulting in 20 user demand information sets. For electricity and electricity billing processing, demand information from 15 users can be collected, resulting in 15 user demand information sets.

[0043] Next, a natural language processing algorithm can be called to split the text of each user's request information, resulting in multiple request texts.

[0044] In one embodiment, natural language processing technology can be used to preprocess the text of user demand information. The preprocessing includes text cleaning, word segmentation, part-of-speech tagging, syntactic analysis, stop word removal, semantic role labeling, etc. Then, a pre-trained machine learning model is used to identify and extract the preprocessed text to filter terms and texts related to the demand, thereby obtaining multiple demand texts.

[0045] In practice, text cleaning can clean the input raw text, including removing redundant whitespace characters, special symbols, non-text elements, etc., to ensure text quality.

[0046] Word segmentation can process cleaned text into words, dividing continuous text into individual words or phrases. This step can use dictionary matching, statistical methods, or deep learning methods to adapt to different languages ​​and domains.

[0047] Part-of-speech tagging can label the parts of speech of words after word segmentation, identifying the part of speech of each word (such as noun, verb, adjective, etc.), which helps to understand the grammatical function and meaning of words in the text of user needs information.

[0048] Syntactic analysis can further analyze the text syntactically to determine the dependencies between words, which can help the system understand the structure and meaning of sentences.

[0049] Removing stop words can eliminate common stop words (such as "and", "of", "is") during the analysis process. These words often do not carry actual meaning, and their removal helps reduce the complexity of the analysis.

[0050] Semantic role labeling can identify the main components of a sentence, such as subject, verb, and object, and understand the relationships between these components, which helps to accurately capture the meaning of the text. Through the above preprocessing, the input user requirement information can be transformed into a form that is easier to analyze and process for subsequent processing.

[0051] Next, a pre-trained machine learning model can be used to identify and extract features from the pre-processed text. This pre-trained machine learning model can be based on a Transformer model, specifically an attention-based extraction mechanism. The Transformer model, through its self-attention layer, considers the relationships and dependencies between words in the text, calculates the similarity between the query and the key, and further transforms these similarities into weights using a softmax function. Subsequently, a feedforward neural network layer can further process the output of the self-attention layer for feature extraction, thereby filtering relevant terms and text to obtain multiple requirement texts.

[0052] The above operations combine the outputs of the attention mechanism and the feedforward neural network layer to obtain a deep feature representation of the input text. These features capture the complex relationships between words in the text and can be used for subsequent classification processing. Furthermore, this feature extraction process captures deep semantic and structural information in the input text, providing a rich foundation for subsequent processing. In practical applications, the model structure and parameters may be adjusted according to specific needs and data.

[0053] Subsequently, aggregation algorithms can be used to aggregate and classify multiple requirement texts to obtain multiple requirement categories, each of which includes one or more requirement texts.

[0054] As described above, during the recognition process, text features can be extracted, and these features contain semantics. Aggregation algorithms such as K-means clustering, agglomerative clustering, or DBSCA can be used to aggregate and classify multiple demand texts. During this aggregation and classification process, the extracted features can be used as the clustering target, grouping texts with the same semantics together. This allows for the initial aggregation of electrical nodes in the graph theory model, resulting in multiple demand categories. Each demand category includes one or more demand texts.

[0055] Finally, the semantics of each of the aforementioned demand categories can be determined, and its semantics can be used as the classification type. Each type is an evaluation index, and evaluation index information can be obtained.

[0056] In an optional embodiment, there may be multiple requirement texts. After aggregation and classification, the first requirement category may contain 20 requirement texts, and the second requirement category may contain only 1 requirement text. For requirement categories with a small number of requirement texts, their category may not be related to requirements but rather a category of special texts. To eliminate requirement categories with a small number of requirement texts, after obtaining multiple requirement categories, the number of requirement texts in each requirement category can be determined. If the number is greater than a preset number, the requirement category can be eliminated.

[0057] S12. Call the preset evaluation model to calculate the business score value based on the business service information and the evaluation indicator information, and perform data tracking based on the value of the business score value to obtain business process information.

[0058] In one embodiment, after acquiring business service information and evaluation indicator information, the data management platform can invoke a preset evaluation model to calculate a business score based on the business service information and evaluation indicator information. The business service information is then evaluated using the indicators in the evaluation indicator information to obtain the business score. The magnitude of the business score can determine whether the content of each service item in the business service information meets the user's needs, thereby quantifying the actual service provided.

[0059] Next, based on the business score, data tracking can be performed on each service item of the business service information to obtain business process information. Through data tracking, the specific operational data and service data of the service items in this business service can be retrieved, thereby obtaining business process information. Furthermore, the tracked data can be used for further verification and subsequent evaluation and analysis to determine whether the business has errors and whether it needs to be optimized.

[0060] As analyzed above, the classification of demand categories involves multiple texts, and the data in different demand texts may be inconsistent and may contain abnormal data. To avoid the impact of abnormal data on the accuracy of the score calculation and to quantify the business service information and evaluation indicator information into corresponding numerical values, in an optional embodiment, the step of calling a preset evaluation model to calculate the business score based on the business service information and the evaluation indicator information may include the following sub-steps: S121. Call the preset quality evaluation model to perform quality evaluation processing on the business service information and the evaluation indicator information to obtain the quality evaluation value.

[0061] S122. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item of the business service information and the category weight value of each demand category.

[0062] S123. Calculate the business score using the category weight value and the service weight value.

[0063] In one embodiment, a preset quality evaluation model can be invoked to perform quality assessment processing on business service information and evaluation indicator information to obtain a quality assessment value.

[0064] The data quality of business service information and evaluation indicator information is determined by the quality evaluation value. If the data quality meets the requirements, the business service information and evaluation indicator information can be used to calculate the business score value.

[0065] In one operational mode, the preset quality assessment model can be an LSTM prediction model. In practice, sample data meeting quality requirements can be collected and processed (including data repair and normalization) to obtain normalized sample data. Specifically, prior time-point business service information and evaluation indicator information meeting quality requirements can be used. Feature extraction is performed on the normalized sample data to obtain highly relevant feature parameters. These highly relevant feature parameters are then divided into training and testing datasets. The LSTM model is trained using the training dataset, resulting in a trained LSTM model. The trained LSTM model is then tested using the testing dataset, yielding the optimal LSTM model. Finally, the optimal LSTM model is used to score the business service information and evaluation indicator information, and a quality assessment value is output.

[0066] If the quality assessment value is greater than the preset quality value, it means that the data quality of the business service information and evaluation indicator information meets the requirements, and the service weight value of each business service item and the category weight value of each demand category can be obtained.

[0067] The service weight value for each business service item can be a pre-set weight value. The more important the business service, the greater the weight value, and vice versa. The more common the business service, the smaller the weight value.

[0068] The category weight value for each requirement category can be the number of requirement texts in that category. The more requirement texts there are, the more important the user is to it, the more important the corresponding requirement evaluation metric is, and the larger the corresponding category weight value. Conversely, the fewer the requirement texts, the smaller the category weight value.

[0069] Finally, the business score can be calculated using category weight values ​​and service weight values.

[0070] Specifically, a one-way service score can be calculated using category weight values ​​and service weight values, and then the sum of multiple one-way service scores can be calculated to obtain the business score.

[0071] In one embodiment, the one-way service score can be calculated as follows: ; In the above formula, C i This represents the one-way service score for the i-th service item, where i = 1 - n, and n is the number of service items, P. j Let P be the category weight value of the j-th demand category. k Let P be the category weight value of the k-th demand category. l Let A be the category weight value of the l-th demand category. i The service ratio value for the i-th business service item.

[0072] The one-way service score for each business service can be calculated based on the above calculations. Finally, the sum of the n one-way service scores for the n business services is calculated to obtain the business score.

[0073] In one embodiment, the business score can be calculated as follows: ; In the above formula, S represents the business score.

[0074] In one embodiment, the service information is the information obtained after completing the current service. The above operation can determine whether the service meets the user's requirements by calculating a service score. To further evaluate the entire service, the entire service process needs to be verified. To verify the entire service process and determine whether there are any errors during processing, in one embodiment, the step of tracking data based on the value of the service score to obtain service process information may include the following sub-steps: S21. If the business score value is greater than the preset score value, then read the project text and service time nodes of different business service items contained in the business service information.

[0075] S22. Use the tags of several project texts to search several corresponding business databases.

[0076] S23. Based on the service time node, extract the corresponding time-related business processing data from each of the business databases to obtain business process information.

[0077] If the business score is higher than the preset score, it means that the service item corresponding to the business service can meet the user's needs and comply with the various indicators of the user's needs. To verify the business service process, data tracking can be performed to obtain various data involved in the business service process. Based on the actual operational data, it can be determined whether there are errors in the service process and whether further optimization is needed. To track various data, the project text and service time nodes of different business service items contained in the business service information can be read.

[0078] The project text can correspond to the different service information contained in each business service item. The service time node is the specific data interaction time node when the service is provided to the user.

[0079] In one embodiment, each business service item corresponds to a different data interaction operation. To distinguish the data of different operations, different databases can be set up, with each database storing the interaction data of one business service item. To determine the database corresponding to a project text, the tags of each project text can be obtained, and then the tags can be used for database matching to filter out several corresponding business databases. Each business database is used to store the operation data of one business service item. For example, the business service of power failure reporting may include: failure reporting items, failure repair items, repair cost settlement items, repair equipment registration items, and after-sales maintenance items, etc. There can be 5 business databases corresponding to the power failure reporting business service, and each business database corresponds to a tag.

[0080] Alternatively, if setting up multiple databases may increase business costs, different storage areas can be set up within a single database, with each storage area storing the operation data for one business service item.

[0081] Finally, business processing data for the corresponding time can be extracted from each business database based on the service time node to obtain business process information.

[0082] Since each service item's operation data is stored, it can be searched based on time points to obtain data on business processes during the interaction. Finally, by summarizing multiple data sets, business process information can be obtained.

[0083] In one embodiment, by calculating a business score, it is determined that the business service does not meet the user's requirements. Therefore, the business service needs to be optimized to meet the user's actual needs. As an example, the step of tracking data based on the magnitude of the business score to obtain business process information may include the following sub-steps: S24. If the business score is less than the preset score, then the preset Bayesian optimization model is invoked to generate business optimization information based on the business service information.

[0084] S25. Adjust the business service items according to the optimization information.

[0085] If the business score is lower than the preset score, it means that the service item corresponding to the business service cannot meet the user's needs. The indicators that do not meet user needs may be due to inadequate service processes or insufficient service content. To optimize the business service, a preset Bayesian optimization model can be invoked to construct business optimization information based on the business service information.

[0086] In one embodiment, a Bayesian optimization model can be pre-constructed. This Bayesian optimization model can be a Bayesian optimization process decision model, which provides an optimization decision scheme by combining deep neural networks and reinforcement learning techniques of the Bayesian optimization model.

[0087] Specifically, the construction of a Bayesian optimization model includes the following steps: The first step is to construct a Bayesian network or other probabilistic graphical model based on the service content and historical operational data to represent a prior probability model that reflects the relationship between process status and optimized service content.

[0088] Specifically, historical operational data of business processes can be collected and organized to construct the original dataset. Prior operational data of this business service can be obtained to build the original dataset. Simultaneously, based on the knowledge and experience of business personnel, key factors influencing business service processes and content, along with their causal relationships, can be identified. Combining this knowledge with data analysis results, a structure learning algorithm is used to automatically mine potential causal structures from the data, constructing an initial Bayesian network topology. Parameter learning is then performed on the initial Bayesian network, estimating network parameters from historical data to obtain a complete parameterized probabilistic model reflecting the process-service content relationship. A scoring function is designed to evaluate the predictive performance of the Bayesian network on the validation set, and a structure reduction strategy is used to optimize the network topology. The optimized Bayesian network undergoes expert review, with human-machine collaboration refining the causal relationships and conditional probability distributions within the network, forming the final prior probability model. A visual interface based on the prior probability model is developed to intuitively display the relationship between the process and service content, providing support for subsequent decision-making and reasoning.

[0089] The second step is to model the process of optimizing service content as a Markov decision process, defining the state space (process anomalies), discrete action space (service content options), and reward function (optimization objective).

[0090] Preferably, the state space represents the various possible process states in the business process, the action space refers to a series of operational behaviors in the service process (e.g., maintenance operation, automatic deduction, service reporting, service date adjustment, etc.), and the reward function is used to characterize the goal of the service, that is, to maximize the long-term cumulative reward of the expected action sequence. The reward function of this invention can take into account service efficiency, economic cost, and other optimization goals.

[0091] The third step is to use business service information as status data and perform standardization, missing value handling, and feature engineering.

[0092] The fourth step involves using a deep neural network to encode the high-dimensional raw state into a low-dimensional continuous state representation, which serves as the input to the reinforcement learning model. Preferably, the network can employ structures such as convolutional networks or gated recurrent units to extract features from anomalies and device states.

[0093] The fifth step is to train the reinforcement learning model in a simulated environment using a deep reinforcement learning algorithm.

[0094] Specifically, a simulated environment model is constructed to describe the dynamic evolution of business service processes and scenarios involving user complaints or unmet needs. For the continuous state space and discrete action space, a suitable deep reinforcement learning algorithm framework is selected, and the algorithm's network structure and training process are designed. The high-dimensional original state data is encoded using the deep neural network from step four to obtain a low-dimensional continuous state vector, which serves as the input to the reinforcement learning agent. Based on the discrete action space designed in step two, a fully connected output layer is selected to parameterize the optional service content schemes, serving as the output of the reinforcement learning agent. The reinforcement learning training process is initiated in the simulated environment, where the agent interacts with the environment to obtain state-action-reward sequences, and training data is cached using experience replay. Parameter updates are periodically performed according to the DQN algorithm's update strategy, allowing the reinforcement learning agent model to gradually optimize the business service content under the guidance of the reward function. Training metrics are monitored, and model performance is evaluated in a validation environment. Once the expected level is reached, the reinforcement learning agent model is saved.

[0095] The sixth step involves combining the prior probability model with the reinforcement learning model. When the business score is less than the preset score, Bayesian inference is performed on the business service information to output the optimal business service solution with the maximum expected reward, thus obtaining optimization information.

[0096] Finally, the optimization information can be used to replace the corresponding content of the business service information, thereby completing the optimization and adjustment process.

[0097] S13. Based on the time and value of the business process information, perform evaluation processing to obtain business evaluation results, and generate a business evaluation report based on the business evaluation results and the business score value.

[0098] In one embodiment, the business process information includes data on the interaction and processing of the business service, as well as the processing time, including its duration and time nodes. An evaluation result can be obtained based on the time and values ​​of the business process information. The evaluation may involve determining whether the operation's values ​​or processing time are abnormal. Finally, a business evaluation report is generated by combining the values, time, and score, allowing backend technical personnel to determine whether the business service needs adjustment and optimization.

[0099] In one embodiment, the process of evaluating the business process information based on time and numerical values ​​to obtain a business evaluation result may include the following steps: S131. Calculate the service duration based on the service time node, and extract the service value from the service processing data.

[0100] S132. Compare the service duration with the corresponding preset processing duration, and compare the service value with the corresponding preset parameter range.

[0101] S133. Determine the business evaluation result based on the comparison results of the time comparison and the comparison structure of the numerical comparison.

[0102] In one embodiment, there are multiple service time nodes. The difference between two adjacent service time nodes can be calculated to obtain multiple service durations. Simultaneously, business values ​​can be extracted from business processing data. These business values ​​can be interactive values ​​corresponding to the service content of a business service item. For example, the business service of power fault reporting may include: fault reporting items, fault repair items, repair fee settlement items, repair equipment registration items, and after-sales maintenance items, etc. The value of the fault reporting item is the fault parameter value, the value of the fault repair item is the value after repair, and the value of the repair fee settlement item is the fee paid, etc.

[0103] Next, the process can compare the service duration with the corresponding preset processing duration, and compare the business value with the corresponding preset parameter range. For example, the first service duration can be compared with its corresponding preset processing duration A, and the second service duration can be compared with its corresponding preset processing duration B. Similarly, the first business value can be compared with its corresponding preset parameter range C, and the second business value can be compared with its corresponding preset parameter range D.

[0104] Finally, the business evaluation result can be determined based on the comparison results of various durations and values. For example, if the comparison result of any value is outside the preset parameter range or the comparison result of any duration is greater than the preset processing time, the business evaluation result can be determined to be abnormal. Otherwise, it is normal.

[0105] For example, if the comparison results of 3-5 values ​​are outside the preset parameter range, or the comparison results of 3-5 time periods are greater than the preset processing time, the business assessment result can be determined as "normal". If the comparison results of 5-8 values ​​are outside the preset parameter range, or the comparison results of 5-8 time periods are greater than the preset processing time, the business assessment result can be determined as "medium". If the comparison results of 8-12 values ​​are outside the preset parameter range, or the comparison results of 8-12 time periods are greater than the preset processing time, the business assessment result can be determined as "severe". The specific assessment method can be adjusted according to actual needs, and this invention does not limit it. Finally, a business evaluation report can be constructed using the above-mentioned values, durations, business scores, and assessment results for technical personnel to refer to.

[0106] In one embodiment, the business evaluation report includes multiple parameters, and technicians need to understand many parameters when reviewing it. To allow technicians to view the data more intuitively, after the step of generating the business evaluation report based on the business evaluation results and the business score, the method may further include the following steps: S14. In response to the display request, convert the business evaluation report into a dynamic profile and visualize the dynamic profile to the technical personnel.

[0107] Specifically, the system can receive presentation requests from technical personnel and, in response, convert business evaluation reports into dynamic profiles. In one operational approach, text can be converted into labels. If a numerical value is associated with a corresponding label, the label's size can be adjusted based on the value. If the value is not associated with a label, different charts can be created for each value, and dynamic effects can be added to these charts and labels to form the dynamic profile. Finally, the dynamic profile can be visually presented to technical personnel.

[0108] In this embodiment, the present invention provides a business analysis method based on evaluation indicators. Its advantages are as follows: the present invention can acquire business service information and evaluation indicator information separately; it calls a preset evaluation model to calculate a business score value based on the business service information and the evaluation indicator information, and performs data tracking based on the numerical value of the business score value to obtain business process information; it performs evaluation processing based on the time and value of the business process information to obtain a business evaluation result, and generates a business evaluation report based on the business evaluation result and the business score value. Comprehensive evaluation and analysis based on different evaluation indicators can meet the actual needs of different users, avoid the error of a single evaluation indicator, thereby reducing evaluation bias and improving evaluation accuracy; moreover, different evaluation indicators are extracted from the needs information of different users, allowing for the evaluation of multiple users' needs at once, simplifying the process, shortening the evaluation processing time, and improving the efficiency of evaluation and analysis.

[0109] This invention also provides a business analysis system based on evaluation indicators, see [link to relevant documentation]. Figure 2 The diagram shows a schematic representation of a business analysis system based on evaluation metrics provided in an embodiment of the present invention.

[0110] As an example, the business analysis system based on evaluation indicators may include: The acquisition module 201 is used to acquire business service information and evaluation indicator information respectively, wherein the business service information is service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information. The calculation module 202 is used to call a preset evaluation model to calculate a business score value based on the business service information and the evaluation indicator information, and to perform data tracking based on the value of the business score value to obtain business process information; The analysis module 203 is used to perform evaluation processing based on the time and value of the business process information to obtain a business evaluation result, and to generate a business evaluation report based on the business evaluation result and the business score value.

[0111] Optionally, obtaining the evaluation index information includes: Obtain multiple user demand information, where each user demand information represents the business service needs of a different user; A natural language processing algorithm is used to split each user request information into multiple request texts. The multiple requirement texts are aggregated and classified using an aggregation algorithm to obtain multiple requirement categories, each requirement category including one or more requirement texts; The classification type of each of the aforementioned demand categories is determined, and evaluation index information is obtained.

[0112] Optionally, the step of calling a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information includes: A preset quality evaluation model is invoked to perform quality assessment processing on the business service information and the evaluation indicator information to obtain a quality evaluation value. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item and the category weight value of each demand category of the business service information; The business score is calculated using the category weight value and the service weight value.

[0113] Optionally, the step of tracking data based on the numerical value of the business score to obtain business process information includes: If the business score is greater than the preset score, then read the project text and service time nodes of different business service items contained in the business service information. Use the tags of several project texts to search several corresponding business databases; Based on the service time node, business processing data corresponding to the time is extracted from each of the business databases to obtain business process information.

[0114] Optionally, the step of obtaining a business evaluation result by evaluating the time and values ​​of the business process information includes: The service duration is calculated based on the service time node, and the service value is extracted from the service processing data; The service duration is compared with the corresponding preset processing duration, and the service value is compared with the corresponding preset parameter range. The business evaluation result is determined based on the comparison results of the time comparison and the comparison structure of the numerical comparison.

[0115] Optionally, the step of performing data tracking processing on the business database based on the numerical value of the business score to obtain business process information includes: If the business score is less than the preset score, then the preset Bayesian optimization model is invoked to generate business optimization information based on the business service information; Adjust the business service items based on the optimization information.

[0116] Optionally, the system further includes: The visualization module is used to respond to a display request after the step of generating a business evaluation report based on the business evaluation results and the business score, convert the business evaluation report into a dynamic profile, and visualize the dynamic profile for technical personnel.

[0117] Those skilled in the art will understand that, for ease of description and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0118] Furthermore, this application also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the business analysis method based on evaluation indicators as described in the above embodiments.

[0119] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a computer-executable program, which is used to cause a computer to perform the business analysis method based on evaluation indicators as described in the above embodiments.

[0120] In the description of the embodiments of the present invention, it should be noted that the terms "above," "below," etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. When an element such as a layer, region, or substrate is referred to as being "above" or "on top of" another element, it may be directly on the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly on" or "above" another element, there is no intermediate element. It should also be understood that when an element is referred to as being "below" or "under" another element, it may be directly below or under the other element, or there may be an intermediate element. Conversely, when an element is referred to as being "directly below" or "under" another element, there is no intermediate element. Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0121] Those skilled in the art will understand that embodiments of this application may also include computer program products. Therefore, this application may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), devices, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0123] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0125] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A business analysis method based on evaluation indicators, characterized in that, The method includes: The business service information and evaluation indicator information are obtained separately. The business service information is the service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information. The system calls a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information, and performs data tracking based on the value of the business score to obtain business process information. The business process information is evaluated based on time and numerical values ​​to obtain a business evaluation result, and a business evaluation report is generated based on the business evaluation result and the business score.

2. The business analysis method based on evaluation indicators according to claim 1, characterized in that, Obtaining the evaluation index information includes: Obtain multiple user demand information, where each user demand information represents the business service needs of a different user; A natural language processing algorithm is used to split each user request information into multiple request texts. The multiple requirement texts are aggregated and classified using an aggregation algorithm to obtain multiple requirement categories, each requirement category including one or more requirement texts; The classification type of each of the aforementioned demand categories is determined, and evaluation index information is obtained.

3. The business analysis method based on evaluation indicators according to claim 2, characterized in that, The step of calling a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information includes: A preset quality evaluation model is invoked to perform quality assessment processing on the business service information and the evaluation indicator information to obtain a quality evaluation value. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item and the category weight value of each demand category of the business service information; The business score is calculated using the category weight value and the service weight value.

4. The business analysis method based on evaluation indicators according to claim 1, characterized in that, The process of tracking data based on the numerical value of the business score to obtain business process information includes: If the business score is greater than the preset score, then read the project text and service time nodes of different business service items contained in the business service information. Use the tags of several project texts to search several corresponding business databases; Based on the service time node, business processing data corresponding to the time is extracted from each of the business databases to obtain business process information.

5. The business analysis method based on evaluation indicators according to claim 4, characterized in that, The evaluation result obtained by evaluating the time and values ​​based on the business process information includes: The service duration is calculated based on the service time node, and the service value is extracted from the service processing data; The service duration is compared with the corresponding preset processing duration, and the service value is compared with the corresponding preset parameter range. The business evaluation result is determined based on the comparison results of the time comparison and the comparison structure of the numerical comparison.

6. The business analysis method based on evaluation indicators according to claim 4, characterized in that, The process of tracking and processing data in the business database based on the numerical value of the business score to obtain business process information includes: If the business score is less than the preset score, then the preset Bayesian optimization model is invoked to generate business optimization information based on the business service information; Adjust the business service items based on the optimization information.

7. The business analysis method based on evaluation indicators according to any one of claims 1-6, characterized in that, After the step of generating a business evaluation report based on the business evaluation results and the business score, the method further includes: In response to the display request, the business evaluation report is converted into a dynamic profile, and the dynamic profile is then visualized and displayed to technical personnel.

8. A business analysis system based on evaluation indicators, characterized in that, The system includes: The acquisition module is used to acquire business service information and evaluation indicator information respectively. The business service information is the service information of different business service items, and the evaluation indicator information is an indicator constructed using user demand information. The calculation module is used to call a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information, and to perform data tracking based on the value of the business score to obtain business process information. The analysis module is used to evaluate the business process information based on time and values ​​to obtain business evaluation results, and to generate a business evaluation report based on the business evaluation results and the business score.

9. The business analysis system based on evaluation indicators according to claim 8, characterized in that, Obtaining the evaluation index information includes: Obtain multiple user demand information, where each user demand information represents the business service needs of a different user; A natural language processing algorithm is used to split each user request information into multiple request texts. The multiple requirement texts are aggregated and classified using an aggregation algorithm to obtain multiple requirement categories, each requirement category including one or more requirement texts; The classification type of each of the aforementioned demand categories is determined, and evaluation index information is obtained.

10. The business analysis system based on evaluation indicators according to claim 9, characterized in that, The step of calling a preset evaluation model to calculate a business score based on the business service information and the evaluation indicator information includes: A preset quality evaluation model is invoked to perform quality assessment processing on the business service information and the evaluation indicator information to obtain a quality evaluation value. If the quality assessment value is greater than the preset quality value, then obtain the service weight value of each business service item and the category weight value of each demand category of the business service information; The business score is calculated using the category weight value and the service weight value.