A service data pushing method, device, equipment and storage medium thereof

Abstract

This application belongs to the field of data processing technology and is applied to service information introduction scenarios. It relates to a service data push method, apparatus, device, and storage medium. The method involves acquiring service push data; performing semantic analysis to extract key data; iteratively modifying and optimizing the service push data based on the key data to obtain modified and optimized service push data; and evaluating the data until it meets preset push requirements before sending it to the target receiving end. Applying this service data push method to scenarios where financial sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual push, enabling both quality control for financial sales personnel and providing customers with more diverse service push data.

Description

Technical Field

[0001] This application relates to the field of data processing technology and is applied in the scenario of service information introduction, and in particular to a service data push method, apparatus, device and its storage medium. Background Technology

[0002] Currently, the common method for pushing business service data to customer service personnel is to provide professional script text and then directly replace the dialogue text between customer service personnel and customers with the professional script text, thereby realizing the push of business service data to customer service personnel.

[0003] However, this service data push method cannot evaluate the business service level of customer service personnel based on the dialogue text between customer service personnel and customers. As a result, the management cannot better control the business quality of different customer service personnel. Moreover, directly replacing it with professional script text results in a relatively simple and mechanical output, which makes it impossible to push diversified service data to customers. Summary of the Invention

[0004] The purpose of this application is to provide a service data push method, apparatus, device and storage medium to solve the problem that in the service data push scenario, the manager cannot better control the business quality of different customer service personnel, and cannot push diversified service data to customers.

[0005] To address the aforementioned technical problems, this application provides a service data push method, which employs the following technical solution:

[0006] A service data push method includes the following steps:

[0007] Step 201: Obtain service push data from the target server;

[0008] Step 202: Perform semantic analysis on the service push data to extract key data;

[0009] Step 203: Determine whether the key data contains non-compliant key data;

[0010] Step 204: If the key data does not contain any non-compliant key data, then the service push data is input into the preset service evaluation model.

[0011] Step 205: If the key data contains non-compliant key data, then the service push data is modified and optimized to obtain modified and optimized service push data, and the service push data is input into the service evaluation model.

[0012] Step 206: Evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model;

[0013] Step 207: If the current service push data does not meet the push requirements, then steps 205 to 206 are executed iteratively to modify, optimize, and evaluate the latest service push data until the current service push data meets the push requirements, at which point the iterative execution stops.

[0014] Step 208: If the current service push data meets the push requirements, then the current service push data is sent to the target receiving end.

[0015] Furthermore, before performing the step of semantic analysis on the service push data to extract key data, the method further includes:

[0016] A key data dictionary is constructed based on the experience of the prophets, which includes key data on violations, key data in colloquial language, and key data in standardized language.

[0017] Deploy the key data dictionary onto the pre-trained semantic analysis model;

[0018] The step of performing semantic analysis on the service push data to extract key data specifically includes:

[0019] The service push data is input into the semantic analysis model, wherein the semantic analysis model is a semantic analysis model based on natural language processing technology;

[0020] The semantic analysis model is used to perform semantic analysis on the service push data, and all key data contained in the service push data are filtered out based on the semantic analysis results.

[0021] After performing the step of semantic analysis on the service push data to extract key data, the method further includes:

[0022] Based on the aforementioned key data dictionary, all key data in the service push data are classified to obtain key data classification results.

[0023] Furthermore, before performing the step of modifying and optimizing the service push data if the key data contains non-compliant key data, the method further includes:

[0024] Obtain a preset speech text quality inspection dataset, wherein each quality inspection data in the speech text quality inspection dataset consists of a key value and a value value pair, wherein the key value represents the detected hit word, and the value value represents the modified and optimized word;

[0025] A gradient optimization strategy is generated based on the key and value values ​​of each quality inspection data point.

[0026] The gradient optimization strategy is deployed as optimization knowledge into a preset automatic optimization model, wherein the automatic optimization model includes an LLM-based text automatic optimization model.

[0027] Furthermore, the step of generating a gradient optimization strategy based on the key and value of each quality inspection data point specifically includes:

[0028] Different quality inspection data are sequentially obtained from the aforementioned script text quality inspection dataset and used as the current quality inspection data;

[0029] Semantic information encoding is performed on the key and value values ​​in the current quality inspection data to obtain the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data.

[0030] Based on the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data, calculate the semantic encoding distance between the value of the current quality inspection data and the key value of other quality inspection data.

[0031] Based on the semantic coding distance from smallest to largest, all quality inspection data are sorted in a gradient order to obtain the gradient ordering result.

[0032] The gradient optimization strategy is generated based on the gradient arrangement result, wherein the gradient optimization strategy includes: if the modified and optimized words obtained after the current optimization still do not meet the push requirements, then the next quality inspection data corresponding to the modified and optimized words is selected from the gradient arrangement result; and the modified and optimized words of the next quality inspection data are obtained as modified and optimized words for further modification and optimization.

[0033] Furthermore, the step of modifying and optimizing the service push data to obtain modified and optimized service push data if the key data contains non-compliant key data specifically includes:

[0034] The service push data is input into the automatic optimization model;

[0035] Based on the key data classification results, the non-compliant key data in the service push data was identified;

[0036] Based on the gradient optimization strategy, all key data with violations are iteratively modified and optimized;

[0037] The current iteration and optimization will cease once all non-compliant key data has been at least converted into conversational key data.

[0038] Replace each of the non-compliant key data with its corresponding modified key data in the service push data to generate the modified and optimized service push data.

[0039] The step of modifying and optimizing the service push data to obtain modified and optimized service push data further includes:

[0040] Input the service push data that does not meet the push requirements into the automatic optimization model;

[0041] Based on the key data classification results, the colloquial key data in the service push data was identified;

[0042] Based on the gradient optimization strategy, all key colloquial data are iteratively modified and optimized;

[0043] The current round of iterative transformation and optimization will cease once all colloquial key data has been at least converted into standardized key data.

[0044] Replace each of the colloquial key data points with its corresponding modified key data points in the service push data to generate the modified and optimized service push data.

[0045] Furthermore, before executing the step of evaluating whether the current service push data meets the preset push requirements based on the preset data scoring strategy in the service evaluation model, the method further includes:

[0046] Set all key values ​​and all value values ​​in the script text quality inspection dataset as service scoring fields and deploy them to the service evaluation model;

[0047] Set a service score for each service rating field according to the gradient ranking results;

[0048] The step of evaluating whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model specifically includes:

[0049] Identify all service rating fields contained in the current service push data;

[0050] The service scores corresponding to all the service rating fields are summed and averaged to obtain the average service score corresponding to the current service push data.

[0051] Determine whether the average service score exceeds a preset score threshold;

[0052] If the average service score does not exceed the score threshold, then the current service push data does not meet the push requirements;

[0053] If the average service score exceeds the score threshold, then the current service push data meets the push requirements.

[0054] Furthermore, after performing the step of sending the current service push data to the target receiving end, the method further includes:

[0055] Based on the preset modification and optimization log, count the total number of times the current service push data has been modified and optimized;

[0056] Determine whether the total number of times the current service push data has been modified and optimized exceeds the preset threshold.

[0057] If the total number of times the current service push data has been modified and optimized exceeds the threshold, feedback that the number of modifications and optimizations has exceeded the threshold will be sent to the target management terminal.

[0058] If the total number of times the current service push data has been modified and optimized does not exceed the threshold, then an incentive feedback is sent to the target server.

[0059] To address the aforementioned technical problems, this application also provides a service data push device, which employs the following technical solution:

[0060] A service data push device, comprising:

[0061] The service push data acquisition module is used to acquire service push data from the target server.

[0062] The key data extraction module is used to perform semantic analysis on the service push data and extract key data;

[0063] The violation determination module is used to determine whether the key data contains non-compliant key data;

[0064] The evaluation processing first branch module is used to input the service push data into a preset service evaluation model if the key data does not contain any non-compliant key data.

[0065] The second branch module for evaluation and processing is used to modify and optimize the service push data if the key data contains non-compliant key data, obtain modified and optimized service push data, and input the service push data into the service evaluation model.

[0066] The push satisfaction judgment module is used to evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model.

[0067] The iterative modification and optimization module is used to iteratively modify, optimize, and evaluate the latest service push data if the current service push data does not meet the push requirements, until the current service push data meets the push requirements, at which point the iterative execution stops.

[0068] The service push data sending module is used to send the current service push data to the target receiving end if the current service push data meets the push requirements.

[0069] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution:

[0070] A computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the service data push method described above.

[0071] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below:

[0072] A computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps of the service data push method described above.

[0073] Compared with the prior art, the embodiments of this application have the following main advantages:

[0074] The service data push method described in this application involves: acquiring service push data; performing semantic analysis to extract key data; iteratively modifying and optimizing the service push data based on the key data to obtain modified and optimized service push data; and evaluating the data until it meets preset push requirements before sending it to the target receiving end. Applying this service data push method to scenarios where financial sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual push, enabling both quality control for financial sales personnel and providing customers with more diverse service push data. Attached Figure Description

[0075] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0076] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;

[0077] Figure 2 This is a flowchart of an embodiment of the service data push method according to this application;

[0078] Figure 3 This is a flowchart of a specific embodiment of the gradient optimization strategy generation and deployment method in the service data push method described in this application;

[0079] Figure 4 yes Figure 3 A flowchart of a specific embodiment of step 302 shown;

[0080] Figure 5 yes Figure 2 A flowchart of a specific embodiment of step 206 shown;

[0081] Figure 6 This is a schematic diagram of a structure of an embodiment of the service data push device according to this application;

[0082] Figure 7 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0083] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0084] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0085] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0086] like Figure 1 As shown, system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. Network 104 serves as the medium for providing communication links between terminal devices 101, 102, and 103 and server 105. Network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0087] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0088] Terminal devices 101, 102, and 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 players (Moving Picture Experts Group Audio Layer IV), laptops, and desktop computers, etc.

[0089] Server 105 can be a server that provides various services, such as a backend server that supports the pages displayed on terminal devices 101, 102, and 103.

[0090] It should be noted that the service data push method provided in this application embodiment is generally executed by the terminal device, and correspondingly, the service data push device is generally set in the terminal device.

[0091] It should be understood that Figure 1The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0092] Continue to refer to Figure 2 A flowchart of an embodiment of the service data push method according to this application is shown. The service data push method includes the following steps:

[0093] Step 201: Obtain service push data from the target server.

[0094] In this embodiment, the target server includes a real-time dialogue input terminal for financial business sales personnel; the service push data includes dialogue data when financial business sales personnel communicate with the push target, and the push target refers to the customer who is having a dialogue with the financial business sales personnel.

[0095] By acquiring the service push data from the target server, namely the product promotion data or service introduction data entered by financial business sales personnel in the real-time dialogue input terminal, the service push data can be modified and optimized using the service data push method of this application before actual push. This not only enables business quality control for financial business sales personnel, but also makes the service push data received by customers more diversified.

[0096] Step 202: Perform semantic analysis on the service push data to extract key data.

[0097] In this embodiment, the key data includes preset business violation terms, false advertising terms, inappropriate sales terms, impolite terms, business keywords, high-frequency business terms, business summary information, business pricing information, business name information, and business action verbs, etc.; alternatively, the push intent can be composed of business action verbs and business name information from the service push data. Specifically, the push intent identification method includes: filtering business action verbs and business name information from the key data; identifying the push intent by concatenating the business action verbs and business name information, wherein the business action verbs are, for example: purchase, renewal, claims, etc., and the business name information is, for example: life insurance product name, health insurance product name, car insurance product name, etc. Including the push intent in the key data allows for both modification and optimization of the service push data based on the key data and the push intent itself, enabling the modification and optimization to combine multi-dimensional key data, resulting in a more standardized modification and optimization outcome.

[0098] Specifically, the key data includes, in terms of type, violation key data, colloquial key data, and standardized key data. Violation key data generally refers to the aforementioned business violation terms, false advertising terms, inappropriate sales terms, impolite terms, etc. Colloquial key data generally refers to non-professional terms in the service push data that are semantically consistent with the professional terms in the preset script text but are expressed differently. Standardized key data generally refers to words in the service push data that are consistent with the professional terms in the preset script text.

[0099] In this embodiment, before performing the step of semantic analysis on the service push data to extract key data, the method further includes: constructing a key data dictionary based on prophetic experience, wherein the key data dictionary includes violation key data, colloquial key data, and standardized key data; and deploying the key data dictionary to a pre-trained semantic analysis model.

[0100] Specifically, the key data dictionary constructed based on the experience of the prophet can be built by marking key data of violations, colloquial key data and standardized key data based on the service recommendation data of multiple financial business salespersons in history, and then constructing the key data dictionary based on the marking information.

[0101] In this embodiment, the step of performing semantic analysis on the service push data to extract key data specifically includes: inputting the service push data into the semantic analysis model, wherein the semantic analysis model is a semantic analysis model based on natural language processing technology; performing semantic analysis on the service push data through the semantic analysis model, and filtering out all key data contained in the service push data according to the semantic analysis results; specifically, filtering out all key data contained in the service push data according to the key data dictionary in the semantic analysis model.

[0102] In this embodiment, after performing the step of semantic analysis on the service push data and extracting key data, the method further includes: classifying all key data in the service push data based on the key data dictionary to obtain key data classification results.

[0103] Specifically, the classification process is mainly aimed at classifying the key data in the service push data into categories such as non-compliant key data, colloquial key data, and standardized key data.

[0104] By performing semantic analysis on the service push data, key data is extracted, which facilitates subsequent modification and optimization of different types of key data, making the modification and optimization process more scientific, reasonable, and intelligent.

[0105] Step 203: Determine whether the key data contains non-compliant key data.

[0106] Step 204: If the key data does not contain any non-compliant key data, then the service push data is input into the preset service evaluation model.

[0107] Step 205: If the key data contains non-compliant key data, then the service push data is modified and optimized to obtain modified and optimized service push data, and the service push data is input into the service evaluation model.

[0108] In this embodiment, before performing the step of modifying and optimizing the service push data if the key data contains non-compliant key data, the method further includes a gradient optimization strategy generation and deployment method.

[0109] Continue to refer to Figure 3 , Figure 3 This is a flowchart of a specific embodiment of the gradient optimization strategy generation and deployment method in the service data push method described in this application, including:

[0110] Step 301: Obtain a preset speech text quality inspection dataset, wherein each quality inspection data in the speech text quality inspection dataset consists of a key value and a value value pair, wherein the key value represents the detected hit word, and the value value represents the modified and optimized word;

[0111] Specifically, in actual quality inspection scenarios, the key data refers to the data to be inspected. The key data is input into the automatic optimization model, and the detection hit words corresponding to the key data are filtered out through query matching. Then, the modification and optimization words corresponding to the detection hit words are output to complete one modification and optimization.

[0112] Step 302: Generate a gradient optimization strategy based on the key and value values ​​of each quality inspection data point;

[0113] Continue to refer to Figure 4 , Figure 4 yes Figure 3 A flowchart of a specific embodiment of step 302 shown includes:

[0114] Step 401: Sequentially obtain different quality inspection data from the speech text quality inspection dataset and use them as the current quality inspection data;

[0115] Step 402: Semantic information encoding is performed on the key and value values ​​in the current quality inspection data to obtain the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data.

[0116] Specifically, the semantic information encoding of the key and value values ​​in the current quality inspection data can be achieved by pre-setting encoding values ​​for descriptive text commonly used by financial sales personnel when recommending service data. Then, the encoding values ​​are combined and accumulated to obtain the encoding and value corresponding to the key and value values ​​in the current quality inspection data as semantic information encoding.

[0117] Step 403: Based on the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data, calculate the semantic encoding distance between the value of the current quality inspection data and the key value of other quality inspection data.

[0118] Specifically, the difference between the encoding and value corresponding to the current quality inspection data's value and the encoding and value corresponding to the key values ​​of other quality inspection data is calculated, and this difference is used as the semantic encoding distance. By sequentially obtaining different quality inspection data in step 401 as the current quality inspection data, the semantic encoding distance between the key and value values ​​of all quality inspection data can be obtained. This facilitates subsequent gradient sorting of all quality inspection data by combining the semantic encoding distance between the key and value values ​​of all quality inspection data, thus obtaining the gradient sorting result.

[0119] Step 404: Based on the semantic coding distance from smallest to largest, perform gradient sorting on all quality inspection data to obtain the gradient sorting result;

[0120] Step 405: Generate the gradient optimization strategy based on the gradient arrangement result.

[0121] The gradient optimization strategy includes: if the modified and optimized words obtained after the current optimization still do not meet the push requirements, then the next quality inspection data corresponding to the modified and optimized words is selected from the gradient ranking results; the modified and optimized words of the next quality inspection data are obtained as modified and optimized words for further modification and optimization.

[0122] By generating the gradient optimization strategy, multiple gradient transformations and optimizations can be performed on a single key data input, enabling the automatic optimization model to achieve automated transformation and optimization, saving the manpower required for multiple inputs, and making it more automated and intelligent.

[0123] Step 303: Deploy the gradient optimization strategy as optimization knowledge into a preset automatic optimization model, wherein the automatic optimization model includes an LLM-based text automatic optimization model.

[0124] By deploying the gradient optimization strategy as optimization knowledge into a pre-defined automatic optimization model, multiple gradient transformations and optimizations can be performed on a single key data input. This enables the automatic optimization model to automate the transformation and optimization process, saving the manual effort of multiple inputs and making it more automated and intelligent. The LLM-based automatic text optimization model (LLM) refers to a large language model, a fundamental machine learning model that uses deep learning algorithms to process and understand natural language. These models are trained on large amounts of text data to learn patterns and entity relationships in language. Therefore, steps 301 to 303 can be implemented within the large language model. Specifically, the spoken text quality inspection dataset is used as training data to train the large language model, resulting in a trained large language model that serves as the automatic optimization model. The training objective is to generate the gradient optimization strategy and deploy it as optimization knowledge into the automatic optimization model, facilitating subsequent automatic optimization of key data using the gradient optimization strategy for automated multiple transformations and optimizations.

[0125] In this embodiment, the step of modifying and optimizing the service push data to obtain modified and optimized service push data if the key data contains non-compliant key data specifically includes: inputting the service push data into the automatic optimization model; identifying non-compliant key data in the service push data through the key data classification results; iteratively modifying and optimizing all non-compliant key data according to the gradient optimization strategy; stopping the current round of iterative modification and optimization when all non-compliant key data has been at least converted into colloquial key data; and replacing the modified key data corresponding to each of the non-compliant key data into the corresponding positions in the service push data to generate modified and optimized service push data.

[0126] Specifically, the aforementioned modification and optimization involves steps to modify and optimize the service push data when the detected key data contains non-compliant key data. A pre-set condition is in place to stop the current iteration of modification and optimization until all non-compliant key data is at least converted to colloquial key data. The number of iterations in this round of modification and optimization is determined by the gradient change of the gradient optimization strategy. Of course, different iteration termination conditions can also be set, such as the proportion of non-compliant key data being less than a certain value, or the proportion of colloquial key data being less than a certain value; these will not be described in detail here.

[0127] In this embodiment, the step of modifying and optimizing the service push data to obtain modified and optimized service push data further includes: inputting service push data that does not meet the push requirements into the automatic optimization model; identifying colloquial key data in the service push data through the key data classification results; iteratively modifying and optimizing all colloquial key data according to the gradient optimization strategy; stopping the current round of iterative modification and optimization until all colloquial key data has been at least converted into standardized key data; and replacing the modified key data corresponding to each of the colloquial key data with the corresponding positions in the service push data to generate modified and optimized service push data.

[0128] Specifically, the above-mentioned modification and optimization are steps taken after the push evaluation in step 206 to modify and optimize service push data that does not meet the push requirements. At this point, since the service push data no longer contains non-compliant key data, the termination condition for this round of iterative modification and optimization is raised. That is, the round of iterative modification and optimization will stop until all colloquial key data is converted into standardized key data. This ensures that the subsequently output modified and optimized service recommendation data only contains standardized key data, that is, only professional terms in the script text. This enables business quality control for financial business sales personnel and makes the service push data received by customers more diversified.

[0129] Step 206: Evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model.

[0130] In this embodiment, before executing the step of evaluating whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model, the method further includes: setting all key values ​​and all value values ​​in the speech text quality inspection dataset as service scoring fields and deploying them into the service evaluation model; and setting service scores for each service scoring field according to the gradient sorting results.

[0131] Specifically, since the gradient arrangement result is based on the semantic coding distance from smallest to largest, all quality inspection data are arranged in a gradient order. Combined with the gradient optimization strategy mentioned above, which includes: if the modified and optimized words obtained after the current optimization still do not meet the push requirements, the next quality inspection data corresponding to the modified and optimized words is selected from the gradient arrangement result; the modified and optimized words of the next quality inspection data are obtained as modified and optimized words again. This can be understood as the word changes for modification and optimization based on the gradient arrangement result following the rule of "first modifying and optimizing non-compliant key data into colloquial key data, and then modifying and optimizing colloquial key data into standardized key data". Therefore, the service score corresponding to each service rating field increases sequentially according to the gradient arrangement result. For example, the four quality inspection data in the speech text quality inspection dataset are (key1, value1), (key2, value2), (key3, value3), and (key4, value4), and the sequence relationship of the key and value values ​​in the gradient arrangement result of the four quality inspection data is exactly key1, value1, key2, value2, key3, value3, key4, value4. At this time, the service scores of key1, value1, key2, value2, key3, value3, key4, value4 are set to 2, 3, 4, 5, 6, 7, 8, 9, respectively, which are in an increasing relationship.

[0132] By pre-setting service scores for all key and value values ​​in the script text quality inspection dataset, it is easier to evaluate whether the current service push data meets the preset push requirements by scoring.

[0133] Continue to refer to Figure 5 , Figure 5 yes Figure 2 A flowchart of a specific embodiment of step 206 shown includes:

[0134] Step 501: Identify all service rating fields contained in the current service push data;

[0135] Step 502: Sum and average the service scores corresponding to all the service rating fields to obtain the average service score corresponding to the current service push data;

[0136] Step 503: Determine whether the average service score exceeds a preset score threshold;

[0137] Step 504: If the average service score does not exceed the score threshold, then the current service push data does not meet the push requirements.

[0138] Step 505: If the average service score exceeds the score threshold, then the current service push data meets the push requirements.

[0139] The scoring method is used to evaluate whether the current service push data meets the preset push requirements, which makes the push judgment method simpler and saves model computing resources.

[0140] Step 207: If the current service push data does not meet the push requirements, then steps 205 to 206 are executed iteratively to modify, optimize, and evaluate the latest service push data until the current service push data meets the push requirements, at which point the iterative execution stops.

[0141] Step 208: If the current service push data meets the push requirements, then the current service push data is sent to the target receiving end.

[0142] In this embodiment, after performing the step of sending the current service push data to the target receiving end, the method further includes: according to a preset modification and optimization log, counting the total number of times the current service push data has been modified and optimized, wherein the modification and optimization log is used to record the number of times each key data has been modified and optimized; determining whether the total number of times the current service push data has been modified and optimized exceeds a preset number threshold; if the total number of times the current service push data has been modified and optimized exceeds the number threshold, sending feedback that the number of modification and optimizations has exceeded the threshold to the target management end; if the total number of times the current service push data has been modified and optimized does not exceed the number threshold, sending incentive feedback to the target server.

[0143] Specifically, by statistically analyzing the total number of times the current service push data has been modified and optimized, a comprehensive quality assessment of the service push data input by financial business sales personnel can be conducted. This helps the target management end, such as the business quality management system of financial institutions, to manage the quality of service push data for financial business sales personnel. This facilitates timely training for financial business sales personnel who have undergone more modifications and optimizations, while providing timely rewards for those who have undergone fewer modifications and optimizations, enabling financial institutions to manage their financial business sales personnel in a more humane way.

[0144] The service data push method described in this application is applied to scenarios where financial business sales personnel promote products or introduce services to target customers. Before the actual push, the service data push method of this application can be used to modify and optimize the service push data. This not only allows for quality control of financial business sales personnel but also makes the service push data received by customers more diverse. Furthermore, the introduction of automatic optimization and data evaluation models enables automated modification and optimization of service sales data. The data evaluation model can also identify the final number of modification and optimization steps, facilitating timely business training and rewards for financial business sales personnel by the financial institution's management system, making management more humane.

[0145] This application involves acquiring service push data; performing semantic analysis to extract key data; iteratively modifying and optimizing the service push data based on the key data to obtain modified and optimized service push data; and evaluating the data until it meets preset push requirements before sending it to the target receiver. Applying this service data push method to scenarios where financial sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual delivery. This enables sales personnel to control the quality of their work and also makes the service push data received by customers more diverse.

[0146] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0147] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0148] In this embodiment, service push data is acquired; semantic analysis is performed to extract key data; the service push data is iteratively modified and optimized based on the key data to obtain modified and optimized service push data; through evaluation, the current service push data is sent to the target receiving end until it meets the preset push requirements. Applying this service data push method to scenarios where financial sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual push, enabling both business quality control for financial sales personnel and a more diverse range of service push data received by customers.

[0149] Further reference Figure 6 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of a service data push device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0150] like Figure 6 As shown, the service data push device 600 described in this embodiment includes: a service push data acquisition module 601, a key data extraction module 602, a violation judgment module 603, an evaluation and processing first branch module 604, an evaluation and processing second branch module 605, a push satisfaction judgment module 606, an iterative modification and optimization module 607, and a service push data sending module 608. Wherein:

[0151] The service push data acquisition module 601 is used to acquire service push data from the target server.

[0152] The key data extraction module 602 is used to perform semantic analysis on the service push data and extract key data;

[0153] The violation determination module 603 is used to determine whether the key data contains violation key data;

[0154] The evaluation processing first branch module 604 is used to input the service push data into a preset service evaluation model if the key data does not contain any non-compliant key data.

[0155] The second branch module 605 is used to modify and optimize the service push data if the key data contains non-compliant key data, obtain modified and optimized service push data, and input the service push data into the service evaluation model.

[0156] The push satisfaction judgment module 606 is used to evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model.

[0157] The iterative modification and optimization module 607 is used to iteratively modify, optimize, and evaluate the latest service push data if the current service push data does not meet the push requirements, until the current service push data meets the push requirements, and then stop the iterative execution.

[0158] The service push data sending module 608 is used to send the current service push data to the target receiving end if the current service push data meets the push requirements.

[0159] This application involves acquiring service push data; performing semantic analysis to extract key data; iteratively modifying and optimizing the service push data based on the key data to obtain modified and optimized service push data; and evaluating the data until it meets preset push requirements before sending it to the target receiver. Applying this service data push method to scenarios where financial sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual delivery. This enables sales personnel to control the quality of their work and also makes the service push data received by customers more diverse.

[0160] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0161] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0162] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed] for details. Figure 7 , Figure 7This is a basic structural block diagram of the computer device in this embodiment.

[0163] The computer device 7 includes a memory 7a, a processor 7b, and a network interface 7c that are interconnected via a system bus. It should be noted that only the computer device 7 with components 7a-7c is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0164] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.

[0165] The memory 7a includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 7a may be an internal storage unit of the computer device 7, such as the hard disk or memory of the computer device 7. In other embodiments, the memory 7a may also be an external storage device of the computer device 7, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 7. Of course, the memory 7a may include both the internal storage unit and its external storage device of the computer device 7. In this embodiment, the memory 7a is typically used to store the operating system and various application software installed on the computer device 7, such as computer-readable instructions for a service data push method. In addition, the memory 7a can also be used to temporarily store various types of data that have been output or will be output.

[0166] In some embodiments, the processor 7b may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 7b is typically used to control the overall operation of the computer device 7. In this embodiment, the processor 7b is used to execute computer-readable instructions stored in the memory 7a or to process data, for example, to execute computer-readable instructions for the service data push method.

[0167] The network interface 7c may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 7 and other electronic devices.

[0168] The computer device proposed in this embodiment belongs to the field of data processing technology and is applied in service information presentation scenarios. This application acquires service push data; performs semantic analysis to extract key data; iteratively modifies and optimizes the service push data based on the key data to obtain modified and optimized service push data; and through evaluation, until the current service push data meets preset push requirements, the current service push data is sent to the target receiving end. Applying this service data push method to scenarios where financial business sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual push, enabling both business quality control for financial business sales personnel and providing customers with more diverse service push data.

[0169] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by a processor to cause the processor to perform the steps of the service data push method described above.

[0170] The computer-readable storage medium proposed in this embodiment belongs to the field of data processing technology and is applied in service information presentation scenarios. This application acquires service push data; performs semantic analysis to extract key data; iteratively modifies and optimizes the service push data based on the key data to obtain modified and optimized service push data; and through evaluation, until the current service push data meets preset push requirements, the current service push data is sent to the target receiving end. Applying this service data push method to scenarios where financial business sales personnel promote products or introduce services to target customers allows for multiple modifications and optimizations of the service push data before actual push, enabling both business quality control for financial business sales personnel and providing customers with more diverse service push data.

[0171] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0172] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. A service data push method characterized by comprising: Includes the following steps: Step 201: Obtain service push data from the target server; Step 202: Perform semantic analysis on the service push data to extract key data; Step 203: Determine whether the key data contains non-compliant key data; Step 204: If the key data does not contain any non-compliant key data, then the service push data is input into the preset service evaluation model. Step 205: If the key data contains non-compliant key data, then the service push data is modified and optimized to obtain modified and optimized service push data, and the service push data is input into the service evaluation model. Before executing the step of modifying and optimizing the service push data to obtain modified and optimized service push data if the key data contains non-compliant key data, the method further includes: Obtain a preset speech text quality inspection dataset. Each quality inspection data in the speech text quality inspection dataset consists of a key value and a value value pair. The key value represents the detected hit word, and the value value represents the modified and optimized word. A gradient optimization strategy is generated based on the key and value values ​​of each quality inspection data point. The gradient optimization strategy is deployed as optimization knowledge into a preset automatic optimization model, which includes an LLM-based text automatic optimization model. Step 206: Evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model; Step 207: If the current service push data does not meet the push requirements, then steps 205 to 206 are executed iteratively to modify, optimize, and evaluate the latest service push data until the current service push data meets the push requirements, at which point the iterative execution stops. Step 208: If the current service push data meets the push requirements, then the current service push data is sent to the target receiving end.

2. The service data push method according to claim 1, characterized in that, Before performing the step of semantic analysis on the service push data to extract key data, the method further includes: A key data dictionary is constructed based on the experience of the prophets, which includes key data on violations, key data in colloquial language, and key data in standardized language. Deploy the key data dictionary onto the pre-trained semantic analysis model; The step of performing semantic analysis on the service push data to extract key data specifically includes: The service push data is input into the semantic analysis model, wherein the semantic analysis model is a semantic analysis model based on natural language processing technology; The semantic analysis model is used to perform semantic analysis on the service push data, and all key data contained in the service push data are filtered out based on the semantic analysis results. After performing the step of semantic analysis on the service push data to extract key data, the method further includes: Based on the aforementioned key data dictionary, all key data in the service push data are classified to obtain key data classification results.

3. The service data push method of claim 1, wherein, The step of generating a gradient optimization strategy based on the key and value of each quality inspection data point specifically includes: Different quality inspection data are sequentially obtained from the aforementioned script text quality inspection dataset and used as the current quality inspection data; Semantic information encoding is performed on the key and value values ​​in the current quality inspection data to obtain the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data. Based on the semantic information encoding results corresponding to the key and value values ​​of all quality inspection data, calculate the semantic encoding distance between the value of the current quality inspection data and the key value of other quality inspection data. Based on the semantic coding distance from smallest to largest, all quality inspection data are sorted in a gradient order to obtain the gradient ordering result. The gradient optimization strategy is generated based on the gradient arrangement result, wherein the gradient optimization strategy includes: if the modified and optimized words obtained after the current optimization still do not meet the push requirements, then the next quality inspection data corresponding to the modified and optimized words is selected from the gradient arrangement result; and the modified and optimized words of the next quality inspection data are obtained as modified and optimized words for further modification and optimization.

4. The service data push method of claim 3, wherein, The step of modifying and optimizing the service push data to obtain modified and optimized service push data if the key data contains non-compliant key data specifically includes: The service push data is input into the automatic optimization model; Based on the key data classification results, the non-compliant key data in the service push data was identified; Based on the gradient optimization strategy, all key data with violations are iteratively modified and optimized; The current iteration and optimization will cease once all non-compliant key data has been at least converted into conversational key data. Replace each of the non-compliant key data with its corresponding modified key data in the service push data to generate the modified and optimized service push data. The step of modifying and optimizing the service push data to obtain modified and optimized service push data further includes: Input the service push data that does not meet the push requirements into the automatic optimization model; Based on the key data classification results, the colloquial key data in the service push data was identified; Based on the gradient optimization strategy, all key colloquial data are iteratively modified and optimized; The current round of iterative transformation and optimization will cease once all colloquial key data has been at least converted into standardized key data. Replace each of the colloquial key data points with its corresponding modified key data points in the service push data to generate the modified and optimized service push data.

5. The service data push method of claim 1, wherein, Before executing the step of evaluating whether the current service push data meets the preset push requirements based on the preset data scoring strategy in the service evaluation model, the method further includes: Set all key values ​​and all value values ​​in the script text quality inspection dataset as service scoring fields and deploy them to the service evaluation model; Set a service score for each service rating field according to the gradient ranking results; The step of evaluating whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model specifically includes: Identify all service rating fields contained in the current service push data; The service scores corresponding to all the service rating fields are summed and averaged to obtain the average service score corresponding to the current service push data. Determine whether the average service score exceeds a preset score threshold; If the average service score does not exceed the score threshold, then the current service push data does not meet the push requirements; If the average service score exceeds the score threshold, then the current service push data meets the push requirements.

6. The service data push method according to claim 1, characterized in that, After performing the step of sending the current service push data to the target receiver, the method further includes: Based on the preset modification and optimization log, count the total number of times the current service push data has been modified and optimized; Determine whether the total number of times the current service push data has been modified and optimized exceeds the preset threshold. If the total number of times the current service push data has been modified and optimized exceeds the threshold, feedback that the number of modifications and optimizations has exceeded the threshold will be sent to the target management terminal. If the total number of times the current service push data has been modified and optimized does not exceed the threshold, then an incentive feedback is sent to the target server.

7. A service data push apparatus characterized by comprising: include: The service push data acquisition module is used to acquire service push data from the target server. The key data extraction module is used to perform semantic analysis on the service push data and extract key data; The violation determination module is used to determine whether the key data contains non-compliant key data; The evaluation processing first branch module is used to input the service push data into a preset service evaluation model if the key data does not contain any non-compliant key data. The second branch module for evaluation and processing is used to modify and optimize the service push data if the key data contains non-compliant key data, obtain modified and optimized service push data, and input the service push data into the service evaluation model. Before executing the step of modifying and optimizing the service push data if the key data contains non-compliant key data, the module further includes: Obtain a preset speech text quality inspection dataset. Each quality inspection data in the speech text quality inspection dataset consists of a key value and a value value pair. The key value represents the detected hit word, and the value value represents the modified and optimized word. A gradient optimization strategy is generated based on the key and value values ​​of each quality inspection data point. The gradient optimization strategy is deployed as optimization knowledge into a preset automatic optimization model, which includes an LLM-based text automatic optimization model. The push satisfaction judgment module is used to evaluate whether the current service push data meets the preset push requirements according to the preset data scoring strategy in the service evaluation model. The iterative modification and optimization module is used to iteratively modify, optimize, and evaluate the latest service push data if the current service push data does not meet the push requirements, until the current service push data meets the push requirements, at which point the iterative execution stops. The service push data sending module is used to send the current service push data to the target receiving end if the current service push data meets the push requirements.

8. A computer device, comprising: The system includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the service data push method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the service data push method as described in any one of claims 1 to 6.

Images