Intelligent telephone traffic management system based on multi-source data analysis

Abstract

The invention belongs to the technical field of telephone traffic management, and particularly relates to an intelligent telephone traffic management system based on multi-source data analysis, which comprises a data fusion and convergence module, a telephone traffic feature deep analysis module, an intelligent telephone traffic distribution decision module, a telephone traffic dynamic scheduling optimization module, a telephone traffic trend prediction and early warning module, a central data warehouse and a telephone traffic management end. According to the invention, the fused and converged data is deeply analyzed to mine the feature information of the telephone traffic, and the telephone traffic distribution scheme is intelligently generated according to the telephone traffic features and the customer service personnel information and is dynamically adjusted according to the actual situation. The telephone traffic trend prediction and early warning module predicts the future change trend of the telephone traffic and generates early warning information in time, so that the telephone traffic can be quickly and accurately allocated to appropriate customer service personnel, and resource preparation can be made in advance to improve the resource utilization efficiency; and the telephone traffic distribution scheme can be dynamically adjusted to adapt to different telephone traffic scenes and changes, so that intelligent management of the telephone traffic is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of call management, and in particular to an intelligent call management system based on multi-source data analysis. Background Art

[0002] Traffic management refers to the process of effectively organizing, planning, controlling, and optimizing the volume of communications traffic. It aims to ensure efficient and stable operation of the communications system while improving service quality and customer satisfaction. Traditional traffic management systems rely primarily on manual experience for traffic allocation and scheduling.

[0003] However, current call management systems lack the ability to comprehensively analyze and utilize multi-source data. This not only leads to irrational call distribution, uneven workloads for customer service personnel, long customer wait times, and difficulty ensuring service quality, but also makes it difficult to allocate resources in advance, affecting the efficiency and effectiveness of the entire call processing process. Furthermore, it is impossible to conduct reasonable analysis and timely warnings on call response results and customer service personnel's management and control performance, hindering the realization of intelligent call management.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent call management system based on multi-source data analysis, which solves the problem that the existing technology lacks comprehensive analysis and utilization of multi-source data, which not only leads to unreasonable call distribution, but also makes it difficult to allocate resources in advance, and is unable to reasonably analyze the call response effect and the call customer service personnel management performance and timely issue warnings, which is not conducive to the realization of intelligent call management.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The intelligent call management system, based on multi-source data analysis, includes a data fusion and aggregation module, a call feature in-depth analysis module, an intelligent call distribution and decision-making module, a call dynamic scheduling and optimization module, a call trend prediction and warning module, a central data warehouse, and a call management terminal. The data fusion and aggregation module collects call-related data from multiple sources, integrates and aggregates this scattered data to form a unified data source, and sends it to the central data warehouse and the call feature in-depth analysis module. The call feature in-depth analysis module conducts in-depth analysis of the integrated data, extracts call feature information, and sends it to the central data warehouse and the intelligent call distribution and decision-making module.

[0008] The intelligent call distribution decision module intelligently generates a call distribution plan based on call characteristics and customer service personnel information, and sends the distribution results to the customer service terminal through the message queue, and feeds back the distribution information to the call dynamic scheduling optimization module; during the call processing process, the call dynamic scheduling optimization module monitors the call status and the work status of the customer service personnel in real time, and dynamically adjusts the call distribution plan according to the actual situation. When the call distribution needs to be adjusted, the distribution plan is updated and resent to the customer service terminal; the call trend prediction and warning module predicts the future change trend of the call volume, and generates warning information when it predicts a call peak or abnormal situation and sends it to the call management end.

[0009] Furthermore, the data fusion and aggregation module establishes connections with different data sources through multiple interfaces, including call records in the customer service system, customer information databases, chat records of online customer service platforms, and customer feedback on social media. For each data source, the module adopts a specific data collection strategy and uses data cleaning technology to remove duplicate, erroneous and incomplete data, and uses data conversion algorithms to convert data in different formats into a unified format, and stores the processed data in the central data warehouse to complete the fusion and aggregation of multi-source data.

[0010] Furthermore, the call feature in-depth analysis module uses data mining algorithms including cluster analysis and association rule mining to conduct multi-dimensional analysis of call data. Specifically, in terms of time dimension, it analyzes the trend of call volume changes in different time periods to identify peak and trough periods; in terms of problem type dimension, it classifies and counts customer consultation issues to identify hot issues; in terms of processing time dimension, it analyzes the average processing time and maximum processing time for different types of calls.

[0011] Furthermore, the call dynamic scheduling optimization module obtains real-time status information of calls, including the waiting time and processing progress of calls, as well as the work status information of customer service personnel, including the idle time of customer service personnel and the number of calls being processed; it uses real-time data analysis algorithms including sliding window analysis and anomaly detection algorithms to analyze real-time data; if it is detected that the call waiting time is too long or the workload of customer service personnel is unbalanced, it will automatically trigger the scheduling optimization mechanism, readjust the call distribution plan, and reallocate part of the calls to other suitable customer service personnel.

[0012] Furthermore, the traffic trend prediction and warning module obtains historical traffic data from the central data warehouse, combines it with the current traffic feature information provided by the traffic feature in-depth analysis module, and uses time series analysis or machine learning prediction algorithms to predict future changes in traffic volume; the prediction results include traffic volume forecast values ​​for different time periods and the predicted time of traffic peaks.

[0013] Furthermore, the traffic management end is communicated with the traffic response improvement evaluation module, which is used to set a detection period, evaluate and analyze the management improvement effect of traffic management during the detection period, generate an improvement evaluation qualified signal or an improvement evaluation abnormal signal through analysis, and send the improvement evaluation qualified signal or the improvement evaluation abnormal signal to the traffic management end.

[0014] Furthermore, the specific analysis process of the call management improvement assessment module includes:

[0015] Obtain the satisfaction level of all call processing during the detection period, mark the call processing process that does not meet the satisfaction requirement as a characteristic process, mark the proportion of the number of characteristic processes during the detection period as an unsatisfactory detection value, compare the unsatisfactory detection value with a preset unsatisfactory detection threshold, and generate an improvement assessment abnormality signal if the unsatisfactory detection value exceeds the preset unsatisfactory detection threshold;

[0016] If the unsatisfactory detection value does not exceed the preset unsatisfactory detection threshold, the waiting time of all traffic in the detection period is obtained, and the corresponding waiting time is compared with the corresponding preset waiting time threshold. If the waiting time exceeds the corresponding preset waiting time threshold, the corresponding traffic is marked as non-timely processed traffic; the number of non-timely processed traffic in the detection period is obtained and the ratio is calculated with the total number of traffic in the detection period to obtain the traffic processing delay detection value, and the ratio of the corresponding waiting time to the corresponding preset waiting time threshold is marked as the waiting time value, and all waiting time values ​​in the detection period are averaged to obtain the waiting time analysis value;

[0017] The call center delay detection value and the waiting time analysis value are numerically compared with the preset call center delay detection threshold and the preset waiting time analysis threshold respectively. If the call center delay detection value or the waiting time analysis value exceeds the corresponding preset threshold, an improvement evaluation abnormality signal is generated; if the call center delay detection value and the waiting time analysis value do not exceed the corresponding preset threshold, an improvement evaluation qualified signal is generated.

[0018] Furthermore, the call management improvement evaluation module is communicated with the customer service control and monitoring module. The call management improvement evaluation module sends the improvement evaluation qualified signal to the customer service control and monitoring module. When the customer service control and monitoring module receives the improvement evaluation qualified signal, it analyzes the control performance of all customer services during the detection period, generates a customer service control qualified signal or a customer service control abnormality signal through analysis, and sends the customer service control qualified signal or the customer service control abnormality signal to the call management end.

[0019] Furthermore, the specific analysis process of the customer service management and monitoring module includes:

[0020] The real-time status of the corresponding customer service personnel is obtained. If the corresponding customer service personnel is not handling the call, it is determined that the customer service personnel is in a non-answering state. Several analysis periods are set within the detection period. The length of time that the corresponding customer service personnel is in the non-answering state within the corresponding analysis period is marked as the total non-answering time value, and the variance of the total non-answering time value of all customer service personnel on the job within the corresponding analysis period is calculated to obtain the traffic distribution imbalance coefficient. The traffic distribution imbalance coefficient is numerically compared with a preset traffic distribution imbalance coefficient threshold. If the traffic distribution imbalance coefficient exceeds the preset traffic distribution imbalance coefficient threshold, the corresponding analysis period is marked as an abnormal distribution period.

[0021] If the call distribution imbalance coefficient does not exceed the preset call distribution imbalance coefficient threshold, the total non-response time of all customer service personnel at the work position in the corresponding analysis period will be averaged to obtain the non-response average time value, and the non-response average time value will be numerically compared with the preset non-response average time value range. If the non-response average time value is not within the preset non-response average time value range, the corresponding analysis period will be marked as an abnormal allocation period; the number of abnormal allocation periods in the detection period is obtained and the ratio is calculated with the number of analysis periods to obtain the allocation decision coefficient, and the allocation decision coefficient is numerically compared with the preset allocation decision coefficient threshold. If the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, a customer service management abnormality signal is generated.

[0022] Furthermore, if the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, the number of complaints received by the corresponding customer service personnel during the detection period is obtained and the ratio is calculated with the number of calls handled during the detection period to obtain the complained detection value, and the average handling time of the corresponding customer service personnel for the corresponding type of calls during the detection period is calculated with the corresponding preset average handling time threshold to obtain the average handling time value, and the average handling time value of all types of calls handled by the corresponding customer service personnel during the detection period is averaged to obtain the handling time characteristic value;

[0023] The number of times the corresponding customer service personnel used uncivilized language during call handling during the detection period is obtained, and the ratio of the number of times the uncivilized language appeared to the total time the corresponding customer service personnel performed call handling during the detection period is calculated to obtain the language abnormality value; the customer service tracking value is calculated by weighted summing the alleged detection value, the time characteristic value and the language abnormality value, and the customer service tracking value is compared with the preset customer service tracking threshold. If the customer service tracking value exceeds the preset customer service tracking threshold, the corresponding customer service personnel is marked as a person in need of training;

[0024] The number of personnel who need training during the detection period is obtained and the ratio is calculated with the total number of customer service personnel to obtain the training requirement percentage value, and the customer service tracking values ​​of all customer service personnel during the detection period are averaged to obtain the customer service monitoring value. The training requirement percentage value and the customer service monitoring value are numerically compared with the preset training requirement percentage threshold value and the preset customer service monitoring threshold value respectively. If the training requirement percentage value or the customer service monitoring value exceeds the corresponding preset threshold value, a customer service management abnormality signal is generated; if both the training requirement percentage value and the customer service monitoring value do not exceed the corresponding preset threshold value, a customer service management qualified signal is generated.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1. The present invention intelligently generates a call allocation plan based on call characteristics and customer service personnel information and dynamically adjusts it according to actual conditions. It also predicts future trends in call volume and generates timely warning information. This facilitates the rapid and accurate allocation of calls to appropriate customer service personnel, enabling advance resource preparation to improve resource utilization efficiency. Furthermore, the call allocation plan can be dynamically adjusted to adapt to different call scenarios and changes, achieving intelligent call management.

[0027] 2. In the present invention, the management improvement effect of call management during the detection period is evaluated and analyzed through the call management improvement evaluation module, and when an improvement evaluation abnormal signal is generated, a cause investigation and analysis is conducted and corresponding optimization measures are taken. When an improvement evaluation qualified signal is generated, the management and control performance of all customer services during the detection period is analyzed, and when a customer service management abnormal signal is generated, the subsequent customer service personnel arrangement plan is adjusted to achieve effective management and control of customer service personnel and ensure the subsequent call processing effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings;

[0029] Figure 1 This is a system block diagram of Embodiment 1 of the present invention;

[0030] Figure 2 This is a system block diagram of Embodiment 2 and Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0032] Example 1: Figure 1 As shown, the intelligent traffic management system based on multi-source data analysis proposed in the present invention includes a data fusion and aggregation module, a traffic feature in-depth analysis module, an intelligent traffic allocation and decision module, a traffic dynamic scheduling and optimization module, a traffic trend prediction and warning module, a central data warehouse, and a traffic management terminal. The data fusion and aggregation module collects traffic-related data from multiple different sources, fuses and aggregates this scattered data to form a unified data source, and sends it to the central data warehouse and the traffic feature in-depth analysis module. By fusing multi-source data, it is possible to comprehensively obtain various traffic-related information, avoiding the limitations of a single data source, providing a richer and more accurate data foundation for subsequent intelligent analysis, and helping to improve the decision-making accuracy of the system.

[0033] It should be noted that the data fusion and aggregation module establishes connections with different data sources through multiple interfaces. The data sources include call records in the customer service system, customer information databases, chat records of online customer service platforms, customer feedback on social media, etc.; for each data source, the module adopts specific data collection strategies, such as timed polling, event triggering, etc. to obtain data, and uses data cleaning technology to remove duplicate, erroneous and incomplete data, and then uses data conversion algorithms to convert data in different formats into a unified format, such as JSON or XML format, and finally stores the processed data in the central data warehouse to complete the fusion and aggregation of multi-source data.

[0034] The in-depth analysis module for call characteristics conducts in-depth analysis of the aggregated data, mining characteristic information of call traffic (such as the temporal distribution of call volume, hot issues consulted by customers, and the processing time of different types of calls). This information is then sent to the central data warehouse and the intelligent call distribution decision module. This in-depth analysis of call characteristics helps the system better understand the patterns and characteristics of call traffic, providing a basis for subsequent call distribution and scheduling, and enabling the system to more rationally allocate resources based on call characteristics.

[0035] Furthermore, the in-depth analysis module for call characteristics uses data mining algorithms to conduct multi-dimensional analysis of call data, such as cluster analysis and association rule mining. In terms of time, it analyzes call volume trends over different time periods to identify peak and trough periods. In terms of question type, it categorizes and compiles statistics on customer inquiries to identify hot issues. In terms of processing time, it analyzes the average and maximum processing times for different call types. Through these analyses, key call characteristics are extracted.

[0036] The intelligent call allocation decision module intelligently generates call allocation plans based on call characteristics and customer service agent information, sends the allocation results to the customer service terminal via a message queue, and feeds the allocation information back to the call dynamic scheduling optimization module, ensuring that calls are efficiently and accurately allocated to the appropriate customer service agents. This improves the efficiency and quality of call processing, reduces customer waiting time, and rationally distributes the workload of customer service agents, preventing some agents from being overworked while others are idle.

[0037] The specific operation process of the intelligent call distribution decision module is as follows: obtaining call feature information from the central data warehouse, and at the same time obtaining data such as the customer service staff's skill level, current workload, historical service quality, etc.; using intelligent decision-making algorithms, such as rule-based reasoning algorithms, machine learning algorithms, etc., comprehensively considering factors such as the urgency of the call, the type of problem, the professional ability of the customer service staff and the current work status, the most suitable customer service staff is selected for each call for distribution, and the distribution results are sent to the corresponding customer service terminal through the message queue.

[0038] During call processing, the dynamic traffic scheduling and optimization module monitors the call status and customer service staff's work status in real time, and dynamically adjusts the call distribution plan based on the actual situation. When the call distribution needs to be adjusted, the distribution plan is updated and resent to the customer service terminal. Dynamic scheduling optimization can promptly respond to emergencies during call processing, ensure smooth call processing, improve customer satisfaction, and further optimize the workload of customer service staff.

[0039] It should be noted that the call dynamic scheduling optimization module obtains real-time status information of calls, such as the waiting time and processing progress of calls, as well as the work status information of customer service personnel, including the idle time of customer service personnel, the number of calls being processed, etc.; it uses real-time data analysis algorithms to analyze real-time data, such as sliding window analysis and anomaly detection algorithms; if it is detected that the call waiting time is too long or the workload of customer service personnel is unbalanced, the scheduling optimization mechanism is automatically triggered, the call distribution plan is readjusted, and part of the calls are reallocated to other suitable customer service personnel.

[0040] The traffic trend forecasting and warning module predicts future trends in traffic volume and generates and sends warning information to the traffic management terminal when traffic peaks or abnormal situations are predicted. Traffic trend forecasting and warning can help managers make resource allocations in advance, such as adding customer service staff and adjusting system configurations, to cope with possible traffic peaks or abnormal situations, and ensure the stable operation of the call processing system.

[0041] It should be noted that the traffic trend prediction and warning module obtains historical traffic data from the central data warehouse, combines the current traffic feature information provided by the traffic feature in-depth analysis module, and uses time series analysis or machine learning prediction algorithms to predict future changes in traffic volume. The prediction results include traffic volume forecast values ​​for different time periods and the predicted time of traffic peaks. At the same time, a warning threshold is set. When it is predicted that the traffic volume may exceed the threshold or experience abnormal fluctuations, the module sends warning information to management personnel via email, text messages, etc.

[0042] The present invention facilitates the rapid and accurate allocation of calls to appropriate customer service personnel, reduces customer waiting time and improves call processing speed through the collaborative work of multiple modules such as multi-source data fusion and aggregation, in-depth analysis of call characteristics, intelligent call distribution decision-making, dynamic call scheduling optimization, and call trend prediction and early warning. It can also prepare resources in advance, avoid waste and shortage of resources, improve resource utilization efficiency, and dynamically adjust call distribution plans to adapt to different call scenarios and changes, thereby realizing intelligent management of call traffic.

[0043] Example 2: Figure 2 As shown, the difference between this embodiment and the first embodiment is that the traffic management terminal is in communication with the traffic response improvement evaluation module. The traffic management improvement evaluation module is used to set a detection period, preferably, the detection period is fifteen days; evaluate and analyze the management improvement effect of the traffic management during the detection period, and generate an improvement evaluation qualified signal or an improvement evaluation abnormal signal through the analysis;

[0044] The improvement assessment qualified signal or improvement assessment abnormal signal is sent to the call management terminal. When the call management terminal receives the improvement assessment abnormal signal, it issues a corresponding warning to remind the back-end personnel to conduct cause investigation and analysis and take corresponding optimization measures, which is conducive to ensuring the timeliness and satisfaction of subsequent call processing. The specific analysis process of the call management improvement assessment module is as follows:

[0045] Obtain the satisfaction level of all call processing during the detection period, mark the call processing process that does not meet the satisfaction requirement as a characteristic process, mark the proportion of the number of characteristic processes during the detection period as a non-satisfactory detection value, and compare the non-satisfactory detection value with a preset non-satisfactory detection threshold. If the non-satisfactory detection value exceeds the preset non-satisfactory detection threshold, indicating that the reputation of the call processing during the detection period is not good, then an improvement evaluation abnormality signal is generated;

[0046] If the unsatisfactory detection value does not exceed the preset unsatisfactory detection threshold, indicating that the reputation of the call handling during the detection period is relatively good, then the waiting time of all calls during the detection period is obtained, and the corresponding waiting time is compared with the corresponding preset waiting time threshold. If the waiting time exceeds the corresponding preset waiting time threshold, indicating that the handling of the corresponding call is not timely, the corresponding call is marked as untimely handling call;

[0047] The number of unprocessed calls during the detection period is obtained and the ratio thereof is calculated with the total number of calls during the detection period to obtain a call processing delay detection value, and the ratio of the corresponding waiting time to the corresponding preset waiting time threshold is marked as a waiting time value, and all waiting time values ​​during the detection period are averaged to obtain a waiting time analysis value;

[0048] The call center delay detection value and the waiting time analysis value are numerically compared with the preset call center delay detection threshold and the preset waiting time analysis threshold respectively. If the call center delay detection value or the waiting time analysis value exceeds the corresponding preset threshold, it indicates that the timeliness of call processing during the detection period is poor, and an improvement evaluation abnormality signal is generated; if the call center delay detection value and the waiting time analysis value do not exceed the corresponding preset threshold, it indicates that the timeliness of call processing during the detection period is good, and an improvement evaluation qualified signal is generated.

[0049] Example 3: Figure 2 As shown, the difference between this embodiment and the first and second embodiments is that the call management improvement evaluation module is communicatively connected to the customer service management and monitoring module. The call management improvement evaluation module sends an improvement evaluation qualified signal to the customer service management and monitoring module. Upon receiving the improvement evaluation qualified signal, the customer service management and monitoring module analyzes the management and control performance of all customer service personnel during the detection period and generates a customer service management qualified signal or a customer service management abnormality signal through the analysis.

[0050] The customer service control qualified signal or customer service control abnormal signal is sent to the call management terminal. When the call management terminal receives the customer service control abnormal signal, it issues a corresponding warning to remind the back-end personnel to adjust the subsequent customer service personnel arrangement plan, such as strengthening the training and control of the corresponding customer service personnel, and reasonably adjusting the number of customer service personnel, etc., to achieve effective control of customer service personnel and further ensure the subsequent call processing effect; the specific analysis process of the customer service control monitoring module is as follows:

[0051] The real-time status of the corresponding customer service personnel is obtained. If the corresponding customer service personnel is not handling the call, it is determined that the customer service personnel is in a non-answering state. Several analysis periods are set within the detection period. The length of time that the corresponding customer service personnel is in the non-answering state during the corresponding analysis period is marked as the total non-answering time value, and the variance of the total non-answering time value of all customer service personnel on the job position during the corresponding analysis period is calculated to obtain the traffic distribution imbalance coefficient. The traffic distribution imbalance coefficient is numerically compared with a preset traffic distribution imbalance coefficient threshold. If the traffic distribution imbalance coefficient exceeds the preset traffic distribution imbalance coefficient threshold, it indicates that the working hours of various customer service personnel in the corresponding analysis period fluctuate greatly and the task distribution is uneven. In this case, the corresponding analysis period is marked as an abnormal distribution period.

[0052] If the traffic distribution imbalance coefficient does not exceed the preset traffic distribution imbalance coefficient threshold, indicating that the working hours of various customer service personnel in the corresponding analysis period have little fluctuation, then the total non-answer time of all customer service personnel at the work position in the corresponding analysis period is averaged to obtain the non-answer average time value, and the non-answer average time value is numerically compared with the preset non-answer average time value range. If the non-answer average time value is not within the preset non-answer average time value range, indicating that there are too many or too few customer service personnel in the corresponding analysis period, then the corresponding analysis period is marked as an abnormal allocation period;

[0053] The number of abnormal time periods allocated during the detection period is obtained and the ratio is calculated with the number of analysis periods to obtain the allocation decision coefficient, and the allocation decision coefficient is numerically compared with the preset allocation decision coefficient threshold. If the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, it indicates that there is an abnormality in the task arrangement or quantity arrangement for customer service personnel during the detection period, and a customer service management abnormality signal is generated.

[0054] Furthermore, if the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, indicating that there is no abnormality in the task arrangement or quantity arrangement for the customer service personnel during the detection period, the number of complaints received by the corresponding customer service personnel during the detection period is obtained and the ratio is calculated with the number of calls handled during the detection period to obtain the complained detection value, and the average handling time of the corresponding customer service personnel for the corresponding type of calls during the detection period is calculated with the corresponding preset average handling time threshold to obtain the average handling time value, and the average handling time value of the corresponding customer service personnel for all types of calls handled during the detection period is averaged to obtain the handling time characteristic value;

[0055] The number of times the corresponding customer service personnel used uncivilized language (mainly insulting voice and text content, etc.) during call handling during the detection period was obtained, and the number of times uncivilized language appeared was calculated by comparing the total time the corresponding customer service personnel spent handling calls during the detection period to obtain the language abnormality value;

[0056] The customer service tracking value is calculated by taking a weighted sum of the disputed detection value, the time-of-dealing characteristic value, and the wording anomaly value. That is, the disputed detection value, the time-of-dealing characteristic value, and the wording anomaly value are each assigned a corresponding preset weight coefficient, and the disputed detection value, the time-of-dealing characteristic value, and the wording anomaly value are multiplied by the corresponding preset weight coefficient, and the sum of the three sets of product results is marked as the customer service tracking value. It should be noted that the larger the value of the customer service tracking value, the worse the work performance of the corresponding customer service personnel during the detection period, and the more timely strengthening of the training and control of the corresponding customer service personnel is needed.

[0057] Compare the customer service tracking value with the preset customer service tracking threshold. If the customer service tracking value exceeds the preset customer service tracking threshold, it indicates that the work performance of the corresponding customer service staff during the detection period is poor and timely strengthening of training and management of the corresponding customer service staff is required. In this case, the corresponding customer service staff will be marked as a person in need of training.

[0058] The number of personnel who need training during the detection period is obtained and the ratio is calculated with the total number of customer service personnel to obtain the training requirement percentage value, and the customer service tracking values ​​of all customer service personnel during the detection period are averaged to obtain the customer service monitoring value. The training requirement percentage value and the customer service monitoring value are numerically compared with the preset training requirement percentage threshold value and the preset customer service monitoring threshold value respectively. If the training requirement percentage value or the customer service monitoring value exceeds the corresponding preset threshold value, it indicates that the management and control performance for all customer service personnel during the detection period is poor, and a customer service management abnormality signal is generated; if both the training requirement percentage value and the customer service monitoring value do not exceed the corresponding preset threshold value, it indicates that the management and control performance for all customer service personnel during the detection period is good, and a customer service control qualified signal is generated.

[0059] The working principle of the present invention is as follows: when in use, the data fusion and aggregation module collects call-related data from multiple different sources and fuses and aggregates them; the call feature deep analysis module conducts in-depth analysis on the fused and aggregated data to mine the feature information of the call; the intelligent call allocation decision module intelligently generates a call allocation plan based on the call features and customer service personnel information; the call dynamic scheduling optimization module dynamically adjusts the call allocation plan according to the actual situation during the call processing process; the call trend prediction and early warning module predicts the future change trend of the call volume and generates early warning information in a timely manner, which is conducive to quickly and accurately allocating calls to appropriate customer service personnel, reducing customer waiting time and improving call processing speed, and can prepare resources in advance to improve resource utilization efficiency, and can dynamically adjust the call allocation plan to adapt to different call scenarios and changes, thereby realizing intelligent management of call traffic.

[0060] The thresholds, preset values, preset ranges, etc. in the technical solution of the present invention are set for result comparison and analysis in order to determine whether they are good or bad. As for their size, they are set for entry and storage based on a combination of large-scale model analysis of sample data and manual experience, and can also be appropriately adjusted based on seasonal or common-sense influencing conditions. As for the settings of preset weight coefficients, influencing factors, etc., specific numerical values ​​are assigned based on the influence of each parameter on the result, ultimately reflecting the influence on the result. They are also set for entry and storage based on a combination of large-scale model analysis of sample data and manual experience, and can also be appropriately adjusted based on seasonal or common-sense influencing conditions.

[0061] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. The preferred embodiments do not describe all details in detail, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention and enable those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent traffic management system based on multi-source data analysis, characterized by: It includes a data fusion and aggregation module, a traffic feature in-depth analysis module, an intelligent traffic allocation and decision-making module, a traffic dynamic scheduling and optimization module, a traffic trend prediction and warning module, a central data warehouse, and a traffic management terminal. The data fusion and aggregation module collects traffic-related data from multiple sources and integrates these scattered data into a unified data source. The traffic feature analysis module conducts in-depth analysis on the integrated data to mine the characteristic information of traffic; The intelligent call distribution decision module intelligently generates a call distribution plan based on call characteristics and customer service personnel information and sends the distribution results to the customer service terminal through the message queue; During the call processing process, the call dynamic scheduling optimization module monitors the call status and the work status of customer service personnel in real time, and dynamically adjusts the call allocation plan according to the actual situation; The traffic trend prediction and warning module predicts the future trend of traffic volume, and generates warning information when traffic peaks or abnormal situations are predicted and sends it to the traffic management end.

2. The intelligent traffic management system based on multi-source data analysis according to claim 1 is characterized in that: The data fusion and aggregation module establishes connections with different data sources through multiple interfaces. For each data source, the module adopts a specific data collection strategy and uses data cleaning technology to remove duplicate, erroneous and incomplete data. It also uses data conversion algorithms to convert data in different formats into a unified format and stores the processed data in the central data warehouse.

3. The intelligent traffic management system based on multi-source data analysis according to claim 1, characterized in that: The in-depth analysis module of call characteristics uses data mining algorithms to conduct multi-dimensional analysis of call data. Specifically, in terms of time, it analyzes the trend of call volume changes in different time periods to identify peak and trough periods of call volume; in terms of problem type, it classifies and counts customer consultation issues to identify hot issues; in terms of processing time, it analyzes the average processing time and maximum processing time for different types of calls.

4. The intelligent traffic management system based on multi-source data analysis according to claim 1, characterized in that: The call dynamic scheduling optimization module obtains real-time status information of calls and the work status information of customer service personnel, and uses real-time data analysis algorithms to analyze real-time data; if it detects that the call waiting time is too long or the workload of customer service personnel is unbalanced, it automatically triggers the scheduling optimization mechanism and readjusts the call distribution plan.

5. The intelligent traffic management system based on multi-source data analysis according to claim 1 is characterized in that: The traffic trend prediction and warning module obtains historical traffic data from the central data warehouse, combines it with the current traffic feature information provided by the traffic feature in-depth analysis module, and uses time series analysis or machine learning prediction algorithms to predict future changes in traffic volume.

6. The intelligent traffic management system based on multi-source data analysis according to claim 1, characterized in that: The traffic management end is communicated with the traffic response improvement evaluation module, which is used to set a detection period, evaluate and analyze the management improvement effect of traffic management during the detection period, and send an improvement evaluation qualified signal or an improvement evaluation abnormal signal to the traffic management end.

7. The intelligent traffic management system based on multi-source data analysis according to claim 6, characterized in that: The specific analysis process of the call management improvement assessment module includes: Obtain the satisfaction level of all call processing during the detection period, mark the call processing process that does not meet the satisfaction requirement as a characteristic process, mark the proportion of the number of characteristic processes during the detection period as an unsatisfactory detection value, compare the unsatisfactory detection value with a preset unsatisfactory detection threshold, and generate an improvement assessment abnormality signal if the unsatisfactory detection value exceeds the preset unsatisfactory detection threshold; If the unsatisfactory detection value does not exceed the preset unsatisfactory detection threshold, the number of calls that were not processed in a timely manner during the detection period is obtained and the ratio is calculated with the total number of calls during the detection period to obtain the call center delay detection value, and the ratio of the corresponding waiting time to the corresponding preset waiting time threshold is marked as the waiting time value, and the average of all waiting time values ​​in the detection period is calculated to obtain the waiting time analysis value; if the call center delay detection value or the waiting time analysis value exceeds the corresponding preset threshold, an improvement evaluation abnormality signal is generated; if both the call center delay detection value and the waiting time analysis value do not exceed the corresponding preset threshold, an improvement evaluation qualified signal is generated.

8. The intelligent traffic management system based on multi-source data analysis according to claim 7 is characterized in that: The call management improvement evaluation module is communicated with the customer service control and monitoring module. When the customer service control and monitoring module receives the improvement evaluation qualified signal, it will analyze the control performance of all customer services during the detection period, and send the customer service control qualified signal or the customer service control abnormality signal to the call management end.

9. The intelligent traffic management system based on multi-source data analysis according to claim 8, characterized in that: The specific analysis process of the customer service management and monitoring module includes: obtaining the number of abnormal time periods allocated during the detection period and calculating the ratio between it and the number of analysis time periods to obtain the allocation decision coefficient. If the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, a customer service management abnormality signal is generated.

10. The intelligent traffic management system based on multi-source data analysis according to claim 9, characterized in that: If the allocation decision coefficient exceeds the preset allocation decision coefficient threshold, the training requirement percentage value and the customer service monitoring value will be numerically compared with the preset training requirement percentage threshold and the preset customer service monitoring threshold respectively. If the training requirement percentage value or the customer service monitoring value exceeds the corresponding preset threshold, a customer service control abnormality signal is generated; if both the training requirement percentage value and the customer service monitoring value do not exceed the corresponding preset threshold, a customer service control qualified signal is generated.

Images