Intelligent video customer service access method and system

Abstract

The present invention discloses an intelligent video customer service access method and system, and relates to the technical field of video customer service access. The performance coefficient of the video customer service system at operation is generated by the load coefficient and stability coefficient of the video customer service system. If the performance coefficient does not exceed the performance threshold, the video customer service system is optimized after matching the corresponding optimization scheme; the improvement degree is generated by the customer service feedback data, and the corresponding processing strategy is adopted according to the relationship between the improvement degree and the improvement threshold. The feedback coefficient is generated by the feedback data within the feedback cycle. If the feedback coefficient is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the issuing state of the feedback instruction. If the feedback density exceeds the expectation, the video customer service system optimizes the knowledge graph to give the video customer service system a corresponding optimization scheme. Targeted optimization of the video customer service system is achieved, so that it can obtain better feedback effect when accessing customers.

Description

Technical Field

[0001] The present invention relates to the technical field of video customer service access, and in particular to an intelligent video customer service access method and system. Background Art

[0002] The video customer service system is a new communication platform developed by combining communications, the Internet and 5G technologies. It provides enterprises with a new way of customer service. Video customer service systems are usually divided into customer service systems that only provide services through video communications and customer service systems that provide multi-modal communication forms such as video, online, and voice. The system can be used in service scenarios that require remote face-to-face and instant communication, such as technical guidance, after-sales service, and 1-to-1 VIP membership services, covering multiple industries such as digital technology, e-commerce retail, automotive travel, and financial insurance.

[0003] Video customer service systems are widely used in insurance loss assessment, distance education, product shopping guides, etc. Through the video customer service system, consumers can enjoy face-to-face VIP services without leaving home, which improves customer experience while reducing corporate costs and improving customer service efficiency.

[0004] In the Chinese invention patent with application publication number CN116346786 A, an intelligent customer service method, system, device and storage medium are disclosed, which belongs to the technical field of customer service system, and solves the technical problem that the existing customer service system has no video function and the user experience is poor. The method is: obtain a user access request sent by the user by scanning a QR code; the access server allocates a temporary account according to the user access request, and registers the temporary account with sipregister. After the registration is successful, the customer service guidance video list, voice call, and video call are pushed to the user for interface presentation; if the user access request is to call the customer service guidance video, the database is queried to obtain the customer service guidance video, and it is played for the user to watch; if the user access request is to call a customer service representative, the idle customer service representative is queried to obtain, and a voice call or video call is established between the user and the customer service representative; when a hang-up request initiated by the user is obtained, the customer service guidance video playback, voice call or video call is ended.

[0005] Combined with the above application and the contents of the prior art:

[0006] When using the telephone customer service system to connect with customers and conduct product marketing to them, since they can usually only use language to introduce products to customers or solve their doubts, the customer's acceptance and understanding are relatively poor, and it is difficult to give a good feedback effect. Therefore, in this case, a video customer service system is usually configured; when the current feedback effect is poor and video customer service intervention is required, a better intervention point is selected to send or accept the customer's video customer service request;

[0007] However, considering that after the customer is connected, there may be frequent switching between telephone customer service and video customer service, which may easily lead to certain deficiencies in the stability of the video customer service system. Therefore, before the video customer service system goes online, it usually needs to undergo more tests, such as stress testing, performance testing, and functional testing. However, after the intelligent video customer service access system has undergone several tests and is online, there may still be situations where customer access effects are poor and it is difficult to give better feedback effects. It is difficult for the existing access system to perform targeted optimization for this and it is difficult to continuously improve the feedback effect.

[0008] To this end, the present invention provides an intelligent video customer service access method and system. Summary of the invention

[0009] 1. Technical issues to be resolved

[0010] In view of the deficiencies of the prior art, the present invention provides an intelligent video customer service access method and system, which optimizes the video customer service system by matching the corresponding optimization scheme to the system; generates an improvement degree from the customer service feedback data, adopts a corresponding processing strategy based on the relationship between the improvement degree and the improvement threshold, generates a feedback coefficient from the feedback data within the feedback cycle, and issues a feedback instruction to the outside if the feedback coefficient is lower than the feedback threshold; generates a feedback density from the issuance state of the feedback instruction, and if the feedback density exceeds expectations, the video customer service system optimizes the knowledge graph to provide a corresponding optimization scheme for the video customer service system. Targeted optimization of the video customer service system is achieved, so that it can obtain better feedback effects when accessing customers, thereby solving the technical problems raised in the background technology.

[0011] (II) Technical solution

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0013] An intelligent video customer service access method, comprising: obtaining feedback data when a customer service system is running within a feedback cycle, aggregating the obtained feedback data to generate a feedback data set, and generating a feedback coefficient from the feedback data set , if the feedback coefficient Below the feedback threshold, build a video customer service system;

[0014] Test the video customer service system in a load test scenario and generate a load factor from the load test data set , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a second-level warning will be issued;

[0015] After receiving the first or second level prompt, the load factor of the video customer service system and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization solution;

[0016] After running the video customer service system online, the improvement degree is generated from the customer service feedback data obtained , based on the improvement The relationship between the improvement threshold and the corresponding processing strategy is adopted; among them, the improvement degree is generated The method is as follows:

[0017]

[0018] Weight coefficient: , ; is the feedback coefficient in the ith sub-period, is the corresponding qualified standard value, is the number of sub-periods;

[0019] Generate feedback coefficients from feedback data within a feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the state of issuing the feedback instruction. , if the feedback density Exceeding expectations, the video customer service system optimized the knowledge graph and provided corresponding optimization solutions for the video customer service system.

[0020] Furthermore, a feedback data collection cycle is set so that the collection cycle includes a number of sub-cycles, and after the customer is connected to the customer service system, the feedback data obtained is aggregated to generate a feedback data set;

[0021] Generate feedback coefficients from feedback data in a feedback data set , where the query frequency , Communication duration and problem solving rate After linear normalization, the corresponding data values ​​are mapped to the interval Then follow the following formula:

[0022]

[0023] in, , is the total number of sub-periods, is the query frequency in the ith sub-period, It is the qualified reference value for the inquiry frequency; is the communication duration of the ith sub-cycle, It is a qualified reference value for communication duration; is the problem solving rate of the ith sub-cycle, is the qualified reference value of problem solving rate; weight coefficient: , , ,and .

[0024] Furthermore, the working status of the video customer service system is tested in a load test scenario and corresponding working status data is obtained, and the obtained load test data is aggregated to construct a load test data set; a load coefficient is generated from the load test data set. , where the response speed and throughput Under dimensionless conditions, according to the following formula:

[0025]

[0026] in, is the response speed on the i-th test node, is the throughput on the i-th test node, is the weight coefficient, .

[0027] Furthermore, in the stress test scenario, the working status of the video customer service system is tested and the corresponding working status data is obtained, and the obtained stress test data is aggregated to construct a stress test data set; the stability coefficient is generated from the stress test data set , where memory usage and network bandwidth usage Under dimensionless conditions, according to the following formula:

[0028]

[0029] in, is the memory usage on the i-th test node, is the network bandwidth usage on the i-th test node, is the weight coefficient, .

[0030] Furthermore, the stability factor and load factor Under dimensionless conditions, the performance coefficient is generated according to the following formula :

[0031]

[0032] Weight coefficient: , , is a constant correction factor; if the performance coefficient If the performance threshold is not exceeded, optimization instructions are issued to the outside world;

[0033] After receiving the optimization instruction, the test scenario data and the corresponding test data of the video customer service system are summarized and feature extraction is performed to obtain the corresponding operation characteristics; after obtaining the optimization schemes of several video customer service systems, a corresponding optimization scheme library is constructed; the optimization scheme library gives the corresponding optimization scheme for the video customer service system; after obtaining the output optimization scheme, the optimization scheme is executed to optimize the video customer service system.

[0034] Furthermore, the optimized video customer service system accesses the video customer service after verifying the user's identity, encrypts and records the video process, sets an observation period including several sub-periods, and after accessing the video customer service, re-acquires the corresponding feedback data in each sub-period to construct a customer service feedback data set in each sub-period.

[0035] Furthermore, the corresponding feedback coefficient is generated from the customer service feedback data set After setting the corresponding qualified standard value for each feedback, under dimensionless conditions, the feedback coefficient in each sub-period is calculated as follows Generate improvement :

[0036] If the improvement If the improvement rate is lower than the preset improvement threshold, an alarm command will be issued to the outside; after receiving the alarm command, the video customer service system will be adjusted; if the improvement rate exceeds the improvement threshold, the operation data and feedback data of the video customer service system will be collected and recorded, and feedback reports will be generated and issued regularly.

[0037] Furthermore, a feedback cycle including several sub-cycles is set, feedback reports within several sub-cycles are continuously obtained, and feedback coefficients within the corresponding cycle are generated from the feedback data in the feedback reports. , record the time node when the feedback instruction is issued within the feedback cycle, and use it as the feedback node, and the ratio of the feedback coefficient below the feedback threshold each time the feedback instruction is issued, and use it as the feedback ratio; the feedback density is generated from the feedback node and the feedback ratio , if the feedback density obtained is When the density threshold is exceeded, a debugging command is issued to the outside.

[0038] Furthermore, after receiving the debugging instruction, the user feedback data within the feedback cycle and the operation data of the video customer service system are obtained, and the corresponding optimization features are obtained after feature extraction; the video customer service system is used as the target word, and the video customer service system optimization knowledge graph is pre-built; among them, the feedback density The generation method is as follows:

[0039]

[0040] Where k is the number of times the feedback command is issued, is the interval from the i-th feedback instruction to the j-th feedback instruction, is the mean value of the interval, is the i-th feedback ratio, is the qualified reference value of the feedback ratio, is the variance of the feedback ratio, is the weight coefficient, .

[0041] An intelligent video customer service access system includes a feedback data collection unit, which obtains feedback data when a customer service system is running within a feedback cycle, aggregates the obtained feedback data to generate a feedback data set, and generates a feedback coefficient from the feedback data set , if the feedback coefficient Below the feedback threshold, build a video customer service system;

[0042] Performance testing unit, testing the video customer service system under load testing scenarios, generating load factors from load testing data sets , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a second-level warning will be issued;

[0043] The performance evaluation unit receives the first-level prompt or the second-level prompt, and the load factor of the video customer service system is and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization solution;

[0044] Improvement analysis unit, after running the video customer service system online, generates improvement degree from the customer service feedback data obtained , based on the improvement Relationship with the improvement threshold, adopt corresponding treatment strategies;

[0045] Feedback optimization unit, which generates feedback coefficients from feedback data within the feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the state of issuing the feedback instruction. , if the feedback density Exceeding expectations, the video customer service system optimized the knowledge graph and provided corresponding optimization solutions for the video customer service system.

[0046] (III) Beneficial effects

[0047] The present invention provides an intelligent video customer service access method and system, which has the following beneficial effects:

[0048] 1. Based on the feedback coefficient Evaluate the feedback status of the current customer service system. If the expected feedback effect cannot be obtained, connect to the video customer service system to improve the current customer service feedback effect.

[0049] 2. Determine through load testing whether the current video customer service system can simultaneously receive a large number of customers. If not, after receiving the first-level prompt, you can perform targeted processing on the video customer service system to improve its load capacity.

[0050] 3. Based on the stability coefficient Determine whether the performance can remain stable after a long period of operation. If it cannot remain stable, after receiving the second-level prompt, upgrade the performance of the corresponding equipment or optimize the video customer service system to improve the stability of the video customer service system after long-term operation.

[0051] 4. Based on the performance coefficient obtained Conduct a comprehensive evaluation of the current video customer service system to determine overall whether the current video customer service system can meet customer access needs. If it cannot meet current usage needs, then match the corresponding optimization plan for the video customer service system based on the obtained test data and corresponding qualified reference standards to achieve preliminary optimization of the video customer service system.

[0052] 5. Construct the corresponding feedback coefficient from the feedback data , according to the feedback coefficient The change in the generated improvement , combined with the improvement Determine whether the feedback given by customers has improved after connecting to the video customer service system. If so, it means that the current video customer service system can achieve the expected effect in marketing management for customers. If not, the video customer service system still needs to be adjusted to ensure the availability of the video customer service system.

[0053] 6. Continue to monitor the operation of the video customer service system and obtain corresponding feedback, and build feedback density based on feedback instructions ; Based on feedback density It can further evaluate the effectiveness of the video customer service system and, through multiple rounds of testing and evaluation, comprehensively understand the operating status of the video customer service system in order to facilitate timely optimization and adjustment.

[0054] 7. The video customer service system optimizes the knowledge graph to provide a corresponding optimization plan, further realizing the targeted optimization of the video customer service system, so that it can obtain better feedback effects when connecting with customers. After several iterations, it can continuously improve customer feedback effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of the intelligent video customer service access method of the present invention;

[0056] Figure 2 This is a schematic diagram of the structure of the intelligent video customer service access system of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] See also Figure 1 The present invention provides an intelligent video customer service access method, comprising:

[0059] Step 1: Obtain feedback data from the customer service system during the feedback cycle, aggregate the feedback data to generate a feedback data set, and generate a feedback coefficient from the feedback data set. , if the feedback coefficient Below the feedback threshold, build a video customer service system;

[0060] The step 1 includes the following contents:

[0061] Step 101: Set a feedback data collection cycle so that the collection cycle includes several sub-cycles. After the customer is connected to the customer service system, the inquiry frequency, communication duration, and problem solving rate of the user in each sub-cycle are counted, and the feedback data obtained are aggregated to generate a feedback data set.

[0062] Step 102: Generate feedback coefficients from feedback data in the feedback data set , where the query frequency , Communication duration and problem solving rate After linear normalization, the corresponding data values ​​are mapped to the interval Then follow the following formula:

[0063]

[0064] in, , is the total number of sub-periods, is the query frequency in the ith sub-period, It is the qualified reference value for the inquiry frequency; is the communication duration of the ith sub-cycle, It is a qualified reference value for communication duration; is the problem solving rate of the ith sub-cycle, It is the qualified reference value of problem solving rate;

[0065] Weight coefficient: , , ,and , the weight coefficient can be obtained by referring to the hierarchical analysis method;

[0066] Based on historical data and management expectations of customer service feedback effects, a feedback threshold is pre-set; if the feedback coefficient obtained is lower than the feedback threshold, it means that the current customer service system gives poor feedback after accessing the user, and the current customer feedback needs to be improved, otherwise it will have a negative impact on subsequent marketing management. At this time, an access instruction is issued to the outside; after receiving the access instruction, the customer service video call module is integrated into the current customer service system, and the video customer service system is integrated;

[0067] When using, combine the contents in steps 101 and 102:

[0068] When the telephone customer service system is in operation, the current customer service status is monitored and recorded, and the feedback coefficient is generated from the feedback data obtained. , according to the feedback coefficient Evaluate the feedback status of the current customer service system. If the expected feedback effect cannot be obtained, connecting to the video customer service system can improve the current customer service feedback effect.

[0069] Taking into account that after accessing the customer, there may be frequent switching between telephone customer service and video customer service, which may easily lead to certain deficiencies in the stability of the video customer service system. Therefore, before the video customer service system goes online, it usually needs to undergo more tests, such as stress testing, performance testing, and functional testing. However, after the intelligent video customer service access system has undergone several tests and is online, there may still be situations where customer access effects are poor and it is difficult to give better feedback effects. It is difficult for the existing access system to perform targeted optimization for this and it is difficult to continuously improve the feedback effect.

[0070] Step 2: Test the video customer service system under the load test scenario and generate the load factor from the load test data set , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a second-level warning will be issued;

[0071] The step 2 includes the following contents:

[0072] Step 201: design a load test scenario for the video customer service system, such as simulating a large number of users to access the video customer service system at the same time, test the working status of the video customer service system under the load test scenario and obtain corresponding working status data, including response time, number of requests processed by the system per unit time, number of users online at the same time, and video resolution, frame rate and delay, etc.; aggregate the obtained load test data to construct a load test data set;

[0073] Step 202: Generate a load factor from a load test data set , where the response speed and throughput Under dimensionless conditions, according to the following formula:

[0074]

[0075] in, is the response speed on the i-th test node, is the throughput on the i-th test node, is the weight coefficient, ;

[0076] Based on historical data and the expected load capacity of the video customer service system, the load threshold is set in advance; if the load factor If the load does not exceed the preset threshold, it means that the load capacity of the video customer service system is low. When the video customer service system needs to access many customers at the same time, it may be difficult to complete the expected work tasks. At this time, a first-level prompt is issued to the outside.

[0077] When in use, after completing the access to the video customer service system, perform a load test on the video customer service system to determine whether the current video customer service system can simultaneously receive a large number of customers. If not, after receiving a first-level prompt, perform targeted processing on the video customer service system to improve its load capacity;

[0078] Step 203: design a stress test scenario for the video customer service system, for example, increase the user's request frequency, simulate the access pressure during peak hours, and use the video customer service system in the above working state for a long time; in the stress test scenario, test the working state of the video customer service system and obtain corresponding working state data, including CPU usage, memory usage, network bandwidth usage, storage usage, etc., and aggregate the obtained stress test data to construct a stress test data set;

[0079] Step 204: Generate a stability coefficient from the stress test data set , the stability coefficient is used to evaluate the stability of the video customer service system under extreme conditions, where the memory usage and network bandwidth usage Under dimensionless conditions, according to the following formula:

[0080]

[0081] in, is the memory usage on the i-th test node, is the network bandwidth usage on the i-th test node, is the weight coefficient, , the weight coefficient value is consistent with the previous value;

[0082] Based on historical data and expectations for the stability management of the video customer service system, a stability threshold is set in advance; if the stability coefficient If the stability threshold is not exceeded, it means that the stability of the video customer service system is low under high-load operation conditions and needs to be optimized. In this case, a secondary prompt is issued to the outside.

[0083] When using, combine the contents in steps 201 to 204:

[0084] After completing the load test on the video customer service system, perform stress test and generate the corresponding stability coefficient based on the stress test data. , according to the stability factor Determine whether the performance can remain stable after a long period of operation. If it cannot remain stable, after receiving the second-level prompt, it is necessary to upgrade the performance of the corresponding equipment or optimize the video customer service system to improve the stability of the video customer service system after long-term operation.

[0085] Step 3: After receiving the first or second level prompt, the load factor of the video customer service system and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization solution;

[0086] The step three includes the following contents:

[0087] Step 301: After receiving the first level prompt or the second level prompt, the performance coefficient of the video customer service system during operation is generated from the load coefficient and the stability coefficient of the video customer service system, wherein the stability coefficient and load factor Under dimensionless conditions, the performance coefficient is generated according to the following formula :

[0088]

[0089] Weight coefficient: , , is a constant correction coefficient whose value falls within Internally, the weight coefficients are obtained by referring to the analytic hierarchy process;

[0090] Based on historical data and performance management expectations for the video customer service system, performance thresholds are set in advance; if the performance coefficient If the performance threshold is not exceeded, it means that the performance of the video customer service system may not be able to meet the maximum demand during operation, and it needs to be optimized and an optimization instruction needs to be issued to the outside.

[0091] Step 302: After receiving the optimization instruction, the test scene data and the corresponding test data of the video customer service system are aggregated and feature extraction is performed to obtain corresponding operation features;

[0092] After obtaining several optimization plans for the video customer service system by pre-making or collecting, a corresponding optimization plan library is constructed; according to the correspondence between the operation characteristics and the optimization plan, the optimization plan library provides a corresponding optimization plan for the video customer service system;

[0093] After obtaining the output optimization plan, execute the optimization plan to optimize the video customer service system;

[0094] When used, combine the contents in steps 301 and 302:

[0095] Obtaining the load factor of the video customer service system and stability factor Post-association performance coefficient , based on the performance coefficient obtained A comprehensive evaluation can be conducted on the current video customer service system to determine whether the current video customer service system can meet customer access needs. If it cannot meet current usage needs, the corresponding optimization plan can be matched for the video customer service system based on the obtained test data and corresponding qualified reference standards to achieve preliminary optimization of the video customer service system.

[0096] Step 4: After running the video customer service system online, generate improvement scores based on the customer service feedback data obtained , based on the improvement Relationship with the improvement threshold, adopt corresponding treatment strategies;

[0097] The step 4 includes the following contents:

[0098] Step 401: After optimizing the video customer service system, the optimized video customer service system accesses the video customer service after verifying the user's identity, encrypts and records the video process, and ensures that the user data and call records are synchronized with the existing database; an observation period including several sub-periods is set, and after accessing the video customer service, the corresponding feedback data is re-obtained in each sub-period, and a customer service feedback data set in each sub-period is constructed;

[0099] Step 402: Generate a corresponding feedback coefficient from the customer service feedback data set After setting the corresponding qualified standard value for each feedback, under dimensionless conditions, the feedback coefficient in each sub-period is calculated as follows Generate improvement :

[0100]

[0101] Weight coefficient: , ; is the feedback coefficient in the ith sub-period, is the corresponding qualified standard value, is the number of sub-periods, and the weight coefficient can be obtained by referring to the hierarchical analysis method;

[0102] Pre-set improvement thresholds based on historical data and usage management expectations for the video customer service system;

[0103] If the improvement If it is lower than the preset improvement threshold, it means that after optimization, the trial operation feedback of the video customer service system is still lower than expected, and further optimization may be required, and an alarm instruction may be issued to the outside. After receiving the alarm instruction, the video customer service system is adjusted to improve the application effect of the customer service system.

[0104] If the degree of improvement exceeds the improvement threshold, it means that the video customer service system can meet the expected use. The operation data and feedback data of the video customer service system are collected and recorded, and feedback reports are generated and issued regularly.

[0105] When using, combine the contents in steps 401 and 402:

[0106] After the video customer service system enters the use state, the feedback data of the connected customers is obtained again, and the corresponding feedback coefficient is constructed based on the feedback data. , according to the feedback coefficient The change in the generated improvement , combined with the improvement Determine whether the feedback given by customers has improved after connecting to the video customer service system. If so, it means that the current video customer service system can achieve the expected effect in marketing management for customers. If not, the video customer service system still needs to be adjusted to ensure the availability of the video customer service system.

[0107] Step 5: Generate feedback coefficients from feedback data within the feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the state of issuing the feedback instruction. , if the feedback density Exceeding expectations, the video customer service system optimizes the knowledge graph and provides corresponding optimization solutions for the video customer service system;

[0108] The step five includes the following contents:

[0109] Step 501: Set a feedback cycle including several sub-cycles, continuously obtain feedback reports within several sub-cycles, and generate feedback coefficients within the corresponding cycle from the feedback data in the feedback reports. , if the feedback coefficient If the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the time node when the feedback instruction is issued within the feedback cycle is recorded as the feedback node, and the ratio of the feedback coefficient lower than the feedback threshold each time the feedback instruction is issued is recorded as the feedback ratio;

[0110] Step 502: Generate feedback density from feedback nodes and feedback ratios , as follows:

[0111]

[0112] Where k is the number of times the feedback command is issued, is the interval from the i-th feedback instruction to the j-th feedback instruction, is the mean value of the interval, is the i-th feedback ratio, is the qualified reference value of the feedback ratio, is the variance of the feedback ratio, is the weight coefficient, ;

[0113] Based on historical data and feedback management of the video customer service system, the density threshold is pre-set; if the feedback density obtained is If the density threshold is exceeded, it means that after the video customer service system has been tested and optimized, there are still some deficiencies in actual operation, so debugging is required. At this time, a debugging command is issued to the outside.

[0114] When in use, after initially verifying the availability of the video surveillance system, after the video surveillance system has been running for a period of time, continue to monitor the operation of the video customer service system and obtain corresponding feedback, and build feedback density based on feedback instructions. ; Based on feedback density It can further evaluate the effectiveness of the video customer service system. Through multiple rounds of testing and evaluation, it can comprehensively understand the operating status of the video customer service system so that it can be optimized and adjusted in time when necessary.

[0115] Step 503: after receiving the debugging instruction, the user feedback data and the operation data of the video customer service system within the feedback cycle are obtained, and the corresponding optimization features are obtained after feature extraction; the video customer service system is used as the target word, and after data retrieval and collection and entity relationship building, a video customer service system optimization knowledge graph is pre-built; according to the correspondence between the optimization scheme and the optimization features, the video customer service system optimization knowledge graph gives the corresponding optimization scheme as a reference scheme, so that the video customer service system can be further optimized;

[0116] When using, combine the contents in steps 501 and 503:

[0117] When debugging is needed again, the video customer service system optimizes the knowledge graph based on the data within the feedback cycle to give a corresponding optimization plan. This plan can be used as a reference to further achieve targeted optimization of the video customer service system, so that it can obtain better feedback when connecting with customers. After several iterations, it can continuously improve customer feedback effects.

[0118] The analytic hierarchy process (AHP) is a decision-making method whose core idea is to decompose the elements related to the decision-making problem into levels such as goals, criteria, and plans, and conduct qualitative and quantitative analysis on this basis. This method is suitable for dealing with target systems with hierarchical and staggered evaluation indicators, especially when the target value is difficult to describe quantitatively.

[0119] The application of analytic hierarchy process is very extensive, not only in decision analysis, but also in the fields of syntactic structure analysis. In decision analysis, it decomposes the problem into different components, and aggregates and combines the factors at different levels according to the mutual influence and affiliation between the factors to form a multi-level analysis structure model, so that the problem is ultimately reduced to the determination of the relative importance of the lowest level (schemes and measures for decision-making, etc.) relative to the highest level (overall goal) or the ranking of the relative order of superiority and inferiority.

[0120] The method of building a knowledge graph for optimizing the video customer service system can refer to the following content:

[0121] To clarify the goal of building a knowledge graph, first of all, we need to clarify the role and goal of the knowledge graph in the video customer service system. The knowledge graph should be able to support the video customer service system to more accurately understand user questions, provide relevant information, and perform automatic reasoning, thereby improving service efficiency and quality.

[0122] Choose a construction method. There are three main methods for constructing knowledge graphs: bottom-up, top-down, and a hybrid of the two. In the video customer service system, you can choose the appropriate method according to the actual situation. Bottom-up method: Extract entities, attributes, and relationships from open linked data sources, and gradually build a knowledge graph. This method is suitable for situations where data sources are rich but the structure is complex. Top-down method: Build the model layer of the knowledge graph starting from the top-level concepts, and then refine the concepts and relationships. This method is suitable for situations where the domain knowledge is clear and the structure is clear. Hybrid method: Combine bottom-up and top-down methods to flexibly adjust the construction process according to actual needs.

[0123] Specific construction steps: Data collection: Collect data from various sources such as the Internet, database, text, etc. These data should cover all kinds of knowledge and information required by the video customer service system.

[0124] Knowledge extraction: Entity extraction: Use natural language processing technology to automatically or semi-automatically discover specific entities such as names of people, organizations, places, dates, and times. Relationship extraction: Discover semantic relationships between entities from text, such as the relationship between products and services, the relationship between users and orders, etc. Attribute extraction: Extract relevant attributes of entities, such as product prices, specifications, production dates, etc.

[0125] Knowledge fusion: Eliminate redundancy and ambiguity between different data sources, and integrate and normalize the extracted knowledge. This includes steps such as entity alignment and attribute value filling. Knowledge processing: Perform conceptual abstraction on the constructed data layer and construct the model layer of the knowledge graph. This includes steps such as ontology construction and quality assessment to ensure the accuracy and reliability of the knowledge graph.

[0126] Knowledge graph storage and query: Select appropriate storage methods (such as graph databases, relational databases, etc.) to store knowledge graph data, and design efficient query algorithms to support the real-time query requirements of the video customer service system. Knowledge graph expansion and update: Regularly update the knowledge graph to include the latest knowledge and information, and maintain the timeliness and integrity of the knowledge graph.

[0127] Multimodal data fusion: Consider integrating multimodal data such as video and audio into the knowledge graph to provide more comprehensive information support. In the video customer service system, this helps to better understand the user's behavior and emotions. Personalized recommendation: Based on the user's historical query records and preference information, implement personalized recommendation functions to improve user experience.

[0128] See also Figure 2 The present invention provides an intelligent video customer service access system, comprising:

[0129] The feedback data collection unit obtains the feedback data of the customer service system during the feedback cycle, aggregates the obtained feedback data to generate a feedback data set, and generates a feedback coefficient from the feedback data set. , if the feedback coefficient Below the feedback threshold, build a video customer service system;

[0130] Performance testing unit, testing the video customer service system under load testing scenarios, generating load factors from load testing data sets , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a second-level warning will be issued;

[0131] The performance evaluation unit receives the first-level prompt or the second-level prompt, and the load factor of the video customer service system is and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization solution;

[0132] Improvement analysis unit, after running the video customer service system online, generates improvement degree from the customer service feedback data obtained , based on the improvement Relationship with the improvement threshold, adopt corresponding treatment strategies;

[0133] Feedback optimization unit, which generates feedback coefficients from feedback data within the feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the state of issuing the feedback instruction. , if the feedback density Exceeding expectations, the video customer service system optimized the knowledge graph and provided corresponding optimization solutions for the video customer service system.

[0134] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0135] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0136] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0137] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0138] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0139] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0140] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0141] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent video customer service access method, characterized in that: include, During the feedback cycle, feedback data from the customer service system is obtained, and the obtained feedback data is aggregated to generate a feedback data set, and the feedback coefficient is generated from the feedback data set. , if the feedback coefficient Below the feedback threshold, build a video customer service system; Set the feedback data collection cycle so that the collection cycle includes several sub-cycles. After the customer is connected to the customer service system, the feedback data obtained is aggregated to generate a feedback data set; the feedback coefficient is generated from the feedback data in the feedback data set. , where the query frequency , Communication duration and problem solving rate After linear normalization, the corresponding data values ​​are mapped to the interval Then follow the following formula: ； in, , is the total number of sub-periods, is the query frequency in the ith sub-period, It is the qualified reference value for the inquiry frequency; is the communication duration of the ith sub-cycle, It is a qualified reference value for communication duration; is the problem solving rate of the ith sub-cycle, is the qualified reference value of problem solving rate; weight coefficient: , , ,and ; Test the video customer service system in a load test scenario and generate a load factor from the load test data set , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a secondary prompt is issued; in which, the working status of the video customer service system is tested in the load test scenario and the corresponding working status data is obtained, and the obtained load test data is aggregated to construct a load test data set; the load coefficient is generated from the load test data set , where the response speed and throughput Under dimensionless conditions, according to the following formula: ； in, is the response speed on the i-th test node, is the throughput on the i-th test node, is the weight coefficient, ; In the stress test scenario, test the working status of the video customer service system and obtain the corresponding working status data, aggregate the obtained stress test data to construct a stress test data set; generate a stability coefficient from the stress test data set , where memory usage and network bandwidth usage Under dimensionless conditions, according to the following formula: ； in, is the memory usage on the i-th test node, is the network bandwidth usage on the i-th test node, is the weight coefficient, ; After receiving the first or second level prompt, the load factor of the video customer service system and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization plan; among them, the stability coefficient and load factor Under dimensionless conditions, the performance coefficient is generated according to the following formula : ； Weight coefficient: , , is a constant correction factor; if the performance coefficient If the performance threshold is not exceeded, optimization instructions are issued to the outside world; After receiving the optimization instruction, the test scenario data and the corresponding test data of the video customer service system are aggregated and then feature extracted to obtain the corresponding operation features; after obtaining several optimization solutions for the video customer service system, a corresponding optimization solution library is constructed; the optimization solution library provides the corresponding optimization solution for the video customer service system; after obtaining the output optimization solution, the optimization solution is executed to optimize the video customer service system; After the video customer service system is running online, the improvement degree is generated by the customer service feedback data obtained , based on the improvement The relationship between the improvement threshold and the corresponding processing strategy is adopted; among them, the improvement degree is generated The method is as follows: ； Weight coefficient: , ; is the feedback coefficient in the ith sub-period, is the corresponding qualified standard value, is the number of sub-periods; Generate feedback coefficients from feedback data within a feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the issuance status of the feedback instruction , if the feedback density Exceeding expectations, the video customer service system optimizes the knowledge graph to provide a corresponding optimization solution for the video customer service system; among them, the feedback density The generation method is as follows: ； Where k is the number of times the feedback command is issued, is the interval from the i-th feedback instruction to the j-th feedback instruction, is the mean value of the interval, is the i-th feedback ratio, is the qualified reference value of the feedback ratio, is the variance of the feedback ratio, is the weight coefficient, .

2. According to claim 1, an intelligent video customer service access method is characterized in that: The optimized video customer service system accesses the video customer service after verifying the user's identity, encrypts and records the video process, sets an observation period consisting of several sub-periods, and after accessing the video customer service, re-acquires the corresponding feedback data in each sub-period to construct a customer service feedback data set in each sub-period.

3. The intelligent video customer service access method according to claim 2, characterized in that: Generate the corresponding feedback coefficient from the customer service feedback data set After setting the corresponding qualified standard value for each feedback, under dimensionless conditions, the feedback coefficient in each sub-period is calculated as follows Generate improvement : If the improvement If the improvement rate is lower than the preset improvement threshold, an alarm command will be issued to the outside; after receiving the alarm command, the video customer service system will be adjusted; if ... if the improvement rate is lower than the preset improvement threshold, an alarm command will be issued to the outside; if the improvement rate is lower than the preset When the improvement threshold is exceeded, the operation data and feedback data of the video customer service system are collected and recorded, and feedback reports are generated and issued regularly.

4. The intelligent video customer service access method according to claim 3, characterized in that: Set a feedback cycle containing several sub-cycles, continuously obtain feedback reports within several sub-cycles, and generate feedback coefficients within the corresponding cycle from the feedback data in the feedback reports , record the time node when the feedback instruction is issued within the feedback cycle, which is used as the feedback node, and the proportion of the feedback coefficient lower than the feedback threshold each time the feedback instruction is issued, which is used as the feedback proportion; Generate feedback density from feedback nodes and feedback ratio , if the feedback density obtained is When the density threshold is exceeded, a debugging command is issued to the outside.

5. The intelligent video customer service access method according to claim 4, characterized in that: After receiving the debugging instruction, the user feedback data within the feedback cycle and the operation data of the video customer service system are obtained, and the corresponding optimization features are obtained after feature extraction; Take video customer service system as the target word and build the video customer service system optimization knowledge graph in advance.

6. An intelligent video customer service access system, applying the method described in any one of claims 1 to 5, characterized in that: include, The feedback data collection unit obtains the feedback data of the customer service system during the feedback cycle, aggregates the obtained feedback data to generate a feedback data set, and generates a feedback coefficient from the feedback data set. , if the feedback coefficient Below the feedback threshold, build a video customer service system; Performance testing unit, testing the video customer service system under load testing scenarios, generating load factors from load testing data sets , if the load factor If the result does not exceed expectations, a first-level prompt will be issued; the video customer service system will be tested under a stress test scenario, and a stability coefficient will be generated from the stress test data set. , if the stability coefficient If it does not exceed expectations, a second-level warning will be issued; The performance evaluation unit receives the first-level prompt or the second-level prompt, and the load factor of the video customer service system is and stability coefficient Generate the performance coefficient of the video customer service system at runtime , if the performance coefficient Without exceeding the performance threshold, the video customer service system is optimized after matching the corresponding optimization solution; Improvement analysis unit, after running the video customer service system online, generates improvement degree from the customer service feedback data obtained , based on the improvement Relationship with the improvement threshold, adopt corresponding treatment strategies; Feedback optimization unit, which generates feedback coefficients from feedback data within the feedback cycle , if the feedback coefficient When the value is lower than the feedback threshold, a feedback instruction is issued to the outside; the feedback density is generated by the issuance status of the feedback instruction , if the feedback density Exceeding expectations, the video customer service system optimized the knowledge graph and provided corresponding optimization solutions for the video customer service system.

Images