Multi-channel customer service data isolation processing method, system and device and medium

Abstract

The invention relates to the technical field of data processing, in particular to a multi-channel customer service data isolation processing method, system and device and a medium, and the method comprises the steps: obtaining consultation information, and carrying out the analysis and recognition of the consultation information, and obtaining channel type information; acquiring total channel information, matching sub-channels corresponding to the total channel information according to the channel category information, and dividing the consultation information into the sub-channels; according to the method, the condition comparison parameters are obtained, the processing condition information of the sub-channels is obtained in real time, the processing condition information is compared with the condition comparison parameters, comparison result information used for reflecting the processing condition of each channel is obtained, the working difficulty brought by a multi-channel customer service system to customer service staff is reduced, and the processing efficiency of multi-channel information is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a data isolation processing method, system, device and medium for multi-channel customer service. Background Art

[0002] In the development of modern customer service systems, the introduction of multi-channel customer service systems has become a key measure to improve user experience and customer service efficiency. The core concept of this system is to allow users to consult and provide feedback through multiple channels, such as telephone, email, online chat and social media. The use of such a diverse communication method can greatly facilitate users and provide more contact points. Users can choose a convenient way to contact according to their needs.

[0003] At present, customer service personnel usually obtain pending consultation content on a system. In order to save costs and make full use of limited time, the system usually chooses whether to assign tasks according to the total number of tasks currently remaining for customer service personnel, and customer service personnel will be configured with pending consultation content from multiple channels. In this working environment, there are the following problems: Since customer service personnel need to handle requests from multiple channels at the same time, information interlacing and interference will inevitably occur. This situation makes it difficult for customer service personnel to focus on a specific channel, thus affecting the work efficiency of customer service personnel; secondly, when requests are gathered together through multiple channels, customer service personnel need to spend extra time to identify and classify these requests. This mixed request flow not only increases the complexity of the work, but also reduces the overall processing efficiency, because each request needs to go through additional screening and allocation steps; in addition, due to the decrease in the efficiency of processing requests, the user's waiting time also increases, which directly affects the user's satisfaction. In summary, although the multi-channel customer service system provides users with more convenience and choices, it will reduce work efficiency when processing requests from different channels. The above problems need to be solved. Summary of the invention

[0004] In order to reduce the work difficulty brought to customer service personnel by the multi-channel customer service system and improve the processing efficiency of multi-channel information, this application provides a data isolation processing method, system, device and medium for multi-channel customer service, which adopts the following technical solutions:

[0005] In a first aspect, the present application provides a data isolation processing method for multi-channel customer service, comprising:

[0006] Obtain consulting information, analyze and identify the consulting information, and obtain channel category information;

[0007] Obtain total channel information, match sub-channels corresponding to the total channel information according to channel category information, and divide consulting information into sub-channels;

[0008] Obtain situation comparison parameters and obtain processing situation information of sub-channels in real time, compare the processing situation information with the situation comparison parameters, and obtain comparison result information reflecting the processing situation of each channel.

[0009] Preferably, the specific steps of dividing the consulting information into sub-channels are:

[0010] Threads are constructed for the sub-channels, and the consulting information is divided into one of the threads of the corresponding sub-channel.

[0011] Preferably, the specific steps of constructing threads for sub-channels are:

[0012] Obtain historical request data of the same sub-channel type, build a model based on the historical request data, and obtain a load prediction model;

[0013] Obtaining forecast request data for the current period according to the load forecast model;

[0014] Obtaining request thread corresponding information, matching the preset number of threads in the request thread corresponding information according to the predicted request data, and obtaining the preset thread information;

[0015] Build and adjust the number of threads in the sub-channel according to the preset thread information.

[0016] Preferably, the specific steps of constructing threads for sub-channels are:

[0017] Acquire the current request quantity value in real time, analyze the current request quantity value and preset thread information, and obtain a first analysis and calculation value;

[0018] In the case where the first analysis calculation value exceeds a preset analysis comparison threshold, the thread adjustment quantity in the corresponding information of the request thread is matched according to the first analysis calculation value to obtain the thread adjustment information.

[0019] Preferably, the channel category information includes telephone information, email information, online chat information and social media information.

[0020] Preferably, it also includes:

[0021] When the comparison result information is completed, satisfaction survey information for conducting a survey on users is obtained according to the channel category information matching.

[0022] Preferably, it also includes:

[0023] When the comparison result information is incomplete, obtain the pending processing time information corresponding to the comparison result information, compare the pending processing time information with the preset time threshold, and obtain reminder information for reminder processing when the pending processing time information is greater than the time threshold.

[0024] In a second aspect, the present application provides a data isolation processing system for multi-channel customer service, comprising:

[0025] The channel identification module is used to obtain consulting information, analyze and identify the consulting information, and obtain channel category information;

[0026] The classification module is used to obtain the total channel information, match the sub-channels corresponding to the total channel information according to the channel category information, and divide the consulting information into sub-channels;

[0027] The monitoring module is used to obtain the situation comparison parameters and the processing situation information of the sub-channels in real time, compare the processing situation information with the situation comparison parameters, and obtain the comparison result information reflecting the processing situation of each channel.

[0028] In a third aspect, the present application provides a data isolation processing device for multi-channel customer service, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the data isolation processing method for multi-channel customer service as described above.

[0029] In a fourth aspect, the present application provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the data isolation processing method for multi-channel customer service as described above when running.

[0030] In summary, compared with the prior art, the technical solution provided by this application has at least the following beneficial effects:

[0031] This application analyzes and identifies the obtained consulting information, determines which channel the consulting information is from, obtains channel category information, matches the sub-channel corresponding to the total channel information according to the channel category information, and divides the consulting information into sub-channels, so that the consulting requests to be processed by customer service personnel can be clearly divided, and obtains situation comparison parameters and real-time processing situation information of the sub-channel, compares the processing situation information and situation comparison parameters, obtains comparison result information reflecting the processing situation of each channel, and monitors the processing situation in real time, which can enable customer service personnel to allocate time more reasonably to handle consulting requests, reduce the work difficulty brought to customer service personnel by the multi-channel customer service system, and improve the processing efficiency of multi-channel information. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a flow chart of a data isolation processing method for multi-channel customer service described in an embodiment of the present application.

[0033] Figure 2 It is a first flow chart of the specific steps of constructing threads for sub-channels described in an embodiment of the present application.

[0034] Figure 3 It is a second flow chart of the specific steps of thread construction for sub-channels described in the embodiment of the present application.

[0035] Figure 4 It is a schematic diagram of the process after the comparison described in the embodiment of the present application is completed.

[0036] Figure 5 It is a module diagram of a data isolation processing system for multi-channel customer service described in an embodiment of the present application.

[0037] Figure 6 It is a module diagram of the satisfaction information feedback module described in the embodiment of the present application.

[0038] Figure 7 It is a module schematic diagram of a data isolation processing device for multi-channel customer service described in an embodiment of the present application.

[0039] Description of reference numerals:

[0040] 1. Channel identification module; 2. Classification module; 3. Monitoring module; 4. Satisfaction information feedback module. DETAILED DESCRIPTION

[0041] The following combination Figure 1-Figure 5 The present application is described in further detail. The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments and are not intended to be limiting.

[0042] Reference Figure 1 , a data isolation processing method for multi-channel customer service involved in this application specifically includes:

[0043] Obtain consulting information, analyze and identify the consulting information, and obtain channel category information;

[0044] Obtain total channel information, match sub-channels corresponding to the total channel information according to channel category information, and divide consulting information into sub-channels;

[0045] Obtain situation comparison parameters and obtain processing situation information of sub-channels in real time, compare the processing situation information with the situation comparison parameters, and obtain comparison result information reflecting the processing situation of each channel.

[0046] Specifically, the present application analyzes and identifies the acquired consulting information, determines which channel the consulting information is a request from, and obtains channel category information. The channel category information is specifically the analysis result obtained by analyzing and identifying one of the consulting information, indicating which channel category the consulting information is specifically. The total channel information is a sub-channel transmission method that is pre-divided into several categories. The sub-channel corresponding to the total channel information is matched according to the channel category information, and the consulting information is divided into sub-channels, so that the consulting requests to be processed by the customer service personnel can be clearly divided. And the present application obtains the situation comparison parameters and the processing situation information of the sub-channel in real time, compares the processing situation information with the situation comparison parameters, obtains the comparison result information used to reflect the processing situation of each channel, and monitors the processing situation in real time, which can enable customer service personnel to allocate time more reasonably to handle consulting requests, reduce the work difficulty brought to customer service personnel by the multi-channel customer service system, and improve the processing efficiency of multi-channel information.

[0047] As one implementation manner, the channel category information includes telephone information, email information, online chat information, and social media information.

[0048] Specifically, the embodiment of the present application analyzes and identifies the consultation information after obtaining it, thereby identifying the consultation information as one of telephone information, email information, online chat information and social media information.

[0049] In the embodiment of the present application, the analysis and identification adopts the natural language processing (NLP) algorithm, which identifies the channel to which the request belongs by analyzing the content and source of the request. Each data processing process only analyzes and identifies one of the consultation information.

[0050] The data processing process of channel identification is to obtain the original data containing the request content, such as emails, online chat records, etc., clean and standardize the text, such as removing punctuation, lowercase, and removing stop words, and use NLP technology, such as the bag-of-words model or TF-IDF, that is, term frequency-inverse document frequency, to convert the text into a feature vector. Apply a trained NLP model, such as a text classifier, to analyze text features and identify the source of the request, such as whether it is an email, to obtain channel category information.

[0051] The data processing process of request classification is to use NLP algorithms to determine the channel of the request, such as email and online chat. Extract features from the request, such as subject and content keywords, train classification algorithms, such as random forest and support vector machine, use labeled data sets to classify requests into different processing queues, classify new requests through the trained model, and assign requests to the appropriate queue according to the features.

[0052] In one embodiment, a request is identified as "online chat" through channel identification, and then the content of the request is analyzed through a classification algorithm, such as "order problem", and it is classified into the "customer service" queue, while another telephone request of "technical problem" is classified into the "technical support" queue.

[0053] One embodiment is to collect data from different channels, such as email content, online chat records, social media posts, etc. The displayed content obtained in the email is "Subject: Account issue\nBody: I ​​am having trouble accessing my account", and the obtained displayed content is cleaned, such as removing punctuation, converting to lowercase, and removing stop words, to obtain cleaned data "subject account issue body trouble accessingaccount".

[0054] The bag-of-words model is used to convert text into feature vectors. The weights of words are then calculated using TF-IDF to measure the importance of words. Feature vectors are constructed to represent email and online chat texts as vectors. Support vector machines (SVMs) or naive Bayes classifiers are selected and trained using cleaned and labeled datasets. Each record is labeled with its channel, such as email, online chat, etc. Finally, the trained model is used to classify new requests. The preprocessed new request text is input into the model, and the model outputs channel category information, such as "email."

[0055] Another embodiment is to collect data from different channels, such as email content, online chat records, social media posts, etc. The content displayed in the email is "User: I can't log in to my account.", and the obtained display content is cleaned, such as removing punctuation, converting to lowercase, and removing stop words, to obtain cleaned data "user cant log in account".

[0056] The text is converted into a feature vector through the bag-of-words model, and the word weight is calculated through TF-IDF to measure the importance of the word, and the feature vector is constructed to represent the email and online chat text as a vector. Select a support vector machine (SVM) or a naive Bayes classifier, and use the cleaned and labeled data set for training. Each record is labeled with its channel, such as email, online chat, etc. Finally, the trained model is used to classify new requests. The preprocessed new request text is input into the model, and the model outputs channel category information, such as "online chat".

[0057] One of the embodiments is that after the channel identification is completed and the channel category information of the embodiment of the present application is online chat, the data is first preprocessed, such as the online chat request: "User can't log in" is preprocessed to: "user cant log in", and the request text is converted into a feature vector.

[0058] Then, we select the classification algorithm, such as random forest, support vector machine SVM or deep learning model LSTM, and train the prepared labeled training data set, such as "technical support" and "accounting problem". For example, if the data shows "I can't log in", it will be classified as "technical support" and "My bill is incorrect" will be classified as "accounting problem". Finally, we classify the new request and assign it to the correct processing queue. For example, if we classify the online chat request "User can't log in", the model will assign it to the "technical support" queue.

[0059] As one implementation method, the specific steps of dividing consulting information into sub-channels are as follows:

[0060] Threads are constructed for the sub-channels, and the consulting information is divided into one of the threads of the corresponding sub-channel.

[0061] Specifically, in the embodiment of the present application, the consulting information analyzed and identified each time and the channel category information obtained are a single target, that is, the embodiment of the present application obtains one of the consulting information for analysis and identification to identify which channel it belongs to. After identification, the total channel information is obtained, which is a number of pre-divided channel categories and includes the division path of the channel category, so that the corresponding sub-channel in the total channel information can be matched according to the channel category information. In this way, the consulting information can be divided into sub-channels.

[0062] Specifically, information about all channels is obtained from a database, API, or other data source. This information usually includes channel ID, channel name, channel category, etc. First, a query operation is performed to obtain relevant channel data. Assuming that the data is stored in a database, the embodiment of the present application uses SQL query to extract data. Then obtain a sub-channel list, that is, the total channel information, which contains the ID and category information of each sub-channel.

[0063] Map each channel to the corresponding sub-channel according to the channel category information. For example, if the category of a channel is "Live Chat", the information will be assigned to the "Live Chat Sub-Channel".

[0064] Then obtain the consultation information from the user input or system, which includes the consultation content, timestamp, user ID, etc. According to the source or category of the consultation information, it is assigned to the appropriate sub-channel. For example, all consultations from "online chat" should be assigned to the sub-channel of online chat.

[0065] In order to separate each consulting information, the constructed sub-channel creates an independent thread for each sub-channel, and each thread is responsible for processing the consulting information of its corresponding sub-channel. A multi-threaded framework or library is used to create and manage threads. For example, the threading library is used in Python.

[0066] Distribute the consultation information to the corresponding threads. Each thread receives the consultation information of the corresponding sub-channel and processes it, thereby improving processing efficiency and system response speed.

[0067] One of the embodiments is that this embodiment needs to process consultation requests from different sources, such as online chat and telephone. In order to efficiently process these requests, the system will use independent processing threads. First, identify the channel to which the information belongs. For example, if two messages are identified as "online chat" and "telephone" respectively, match the "online chat" channel with the online chat sub-channel, and match the "telephone" channel with the telephone sub-channel. If the received consultation information is identified as "online chat", it will be assigned to the online chat sub-channel. Create independent threads for each online chat sub-channel and telephone sub-channel so that they can process their own requests separately. Each thread receives and processes the consultation information assigned to it, ensuring that different types of requests are processed in independent threads and reducing mutual interference between information. This enables the system to efficiently process consultation requests from different sources while maintaining the independence between each processing thread.

[0068] As one implementation method, the specific steps for thread construction of the sub-channel are as follows:

[0069] Obtain historical request data of the same sub-channel type, build a model based on the historical request data, and obtain a load prediction model;

[0070] Obtaining forecast request data for the current period according to the load forecast model;

[0071] Obtaining request thread corresponding information, matching the preset number of threads in the request thread corresponding information according to the predicted request data, and obtaining the preset thread information;

[0072] Build and adjust the number of threads in the sub-channel according to the preset thread information.

[0073] Specifically, when the system is running, the embodiment of the present application obtains predicted request data for the current time period or the future time period through a load prediction model, so that each sub-channel is pre-set with a suitable number of threads, so that when the consultation information is obtained and its category is identified and analyzed and needs to be divided, there will be a sufficient number of preset threads in the sub-channel to store a certain amount of consultation information.

[0074] In order to ensure that requests from each channel are assigned independent processing threads without interfering with each other. A multi-threaded processing framework is used to assign independent processing threads to each sub-channel, and each sub-channel includes at least one independent thread. To build a multi-threaded processing framework, a thread pool is first created to manage different types of thread resources. The thread pool can pre-create a certain number of threads and allocate threads from the pool when needed, reducing the overhead of frequent thread creation and destruction. And according to the requested channel type, the corresponding thread is scheduled to process specific types of requests. The scheduling mechanism ensures that requests from different channels do not interfere with each other during processing. It enables the effective isolation of requests from different channels, thereby improving overall processing efficiency and user satisfaction. The multi-threaded processing framework ensures the independence of requests from each channel, reduces interference, and improves the system's response speed and service quality.

[0075] After receiving and classifying requests from different channels, the system will assign the requests to the corresponding threads according to the classification results. Each thread of the request category will independently process its corresponding request, and the thread will focus on processing the request assigned to it and perform necessary operations or replies. For example, the online chat thread processes the user's instant messages, while the phone processing thread processes the voice interaction on the phone, so that customer service personnel can obtain independent information feedback of several channel types, so that they can obtain data more clearly, thereby improving work efficiency.

[0076] One embodiment is that the thread pool configuration creates two thread pools, one specifically processes online chat requests and the other processes telephone requests. For example, the online chat thread pool contains 10 threads and the telephone request thread pool contains 5 threads.

[0077] When the system receives a request, it identifies the source of the request. If the request is for online chat, it is assigned to a thread in the online chat thread pool. If the request is a phone call request, it is assigned to a thread in the phone processing thread pool. The online chat thread takes over the chat conversation and responds in real time; the phone processing thread takes over the phone call and performs voice interaction and recording.

[0078] In one embodiment, a multi-channel customer service system is operated, which supports customer support through website chat, email, and social media. In order for the system to efficiently handle all requests, an appropriate number of threads need to be configured for each sub-channel, such as website chat, email, and social media. The number of requests for each sub-channel may vary at different time periods, so load prediction and thread adjustment are required.

[0079] First, we extract the request data for the past 6 months, including the request volume for each sub-channel. For example, website chat usually has higher request volume between 9 am and 11 am, while email has higher request volume between 2 pm and 4 pm.

[0080] Use historical request data to build a load forecasting model. For example, a time series forecasting model, ARIMA or LSTM, can be used to predict the number of requests per hour per day for the next week, which helps to understand how many requests are expected in different time periods. In the embodiment of the present application, the model predicts that the number of requests for website chat from 9:00 to 11:00 a.m. next Monday is 2,000 requests, and also predicts the number of requests for other time periods, for example, the number of email requests from 2:00 to 4:00 p.m. is 1,500.

[0081] The maximum number of requests that each request thread can handle is set, such as one thread can handle 50 requests. Based on the predicted data, the number of threads required is calculated. If it is estimated that the website chat needs to handle 2,000 requests in a certain period of time, the number of threads that need to be pre-built is: 2,000 / 50 = 40 threads, and the preset number of threads is obtained.

[0082] According to the thread processing capacity and the predicted request volume, find and match the appropriate thread configuration. For example, if it is detected that there are 40 threads in the system of the embodiment of the present application that can be used for website chat, these threads are allocated to the sub-channel to cope with the predicted high request volume.

[0083] Configure the system based on the number of threads obtained. For example, the number of website chat threads previously set is 20, but the forecast shows that 40 threads are needed, so the thread pool needs to be adjusted to increase it to 40 threads. This allows the system to efficiently handle all requests during periods of high request volume.

[0084] Finally, when the system receives website chat requests, it processes these requests according to the real-time load and the preset thread configuration. If the number of requests in a certain period exceeds the prediction, this embodiment can dynamically adjust according to the real-time data so that each sub-channel has enough threads to handle all requests. The processing capacity of the customer service system is optimized through data prediction and thread configuration, so that customer service can be provided more efficiently and smoothly, thereby improving the experience.

[0085] One embodiment is that in order to reasonably pre-build the number of threads for each sub-channel to cope with the requested allocation task, this embodiment adopts load prediction technology. Load prediction technology involves analyzing and predicting the request load of the system in a specific period of time, so as to pre-allocate an appropriate number of processing threads for each sub-channel.

[0086] The historical data prediction method of time series analysis is used to predict future request loads, including autoregressive moving average ARMA, autoregressive integrated moving average ARIMA, seasonal ARIMA, etc. The load prediction in this embodiment is specifically to use the autoregressive integrated moving average ARIMA model to predict the request load. First, data collection is performed to collect historical request data, and the data includes timestamps and request volumes, such as collecting request volume data per hour for the past year. After removing outliers and processing missing values, the timestamp is set as an index, and the data is arranged in chronological order. At this time, check whether the time series data is stable. If it is not stable, differential processing is required. Select the ARIMA model in a stable state and determine the parameters of ARIMA, namely p, d, and q. Use the autocorrelation function ACF and the partial autocorrelation function PACF graph to determine the order of the AR and MA terms, and use the selected parameters to train the ARIMA model, and finally use the trained model to predict future loads.

[0087] Another embodiment is to use machine learning regression analysis to predict request load, and common regression models include linear regression, decision tree regression, support vector regression SVR, etc. The data processing steps include data collection, feature processing, modeling, prediction, and evaluation optimization.

[0088] Another embodiment is to use a neural network model, such as a long short-term memory network (LSTM), a convolutional neural network (CNN), etc. to perform complex load forecasting. The data processing steps include data collection, feature processing, modeling, forecasting, and evaluation optimization.

[0089] As one implementation method, the specific steps for thread construction of the sub-channel are as follows:

[0090] Acquire the current request quantity value in real time, analyze the current request quantity value and preset thread information, and obtain a first analysis and calculation value;

[0091] In the case where the first analysis calculation value exceeds a preset analysis comparison threshold, the thread adjustment quantity in the corresponding information of the request thread is matched according to the first analysis calculation value to obtain the thread adjustment information.

[0092] Specifically, after constructing and adjusting the number of threads in the sub-channel according to the preset thread information, the application needs to obtain the current request number value in real time, and monitor, adjust and optimize the thread management according to the current request number value. Monitor the running status and performance indicators of the thread, such as the response time and processing speed of the thread, and dynamically adjust the thread allocation according to the system load. For example, during peak hours, the number of threads can be increased or the request allocation strategy can be optimized.

[0093] One embodiment is to continuously monitor and obtain the real-time request volume of each sub-channel, such as the current real-time request volume of website chat is 500. The current request volume value and the preset thread information are analyzed, and the real-time request volume is compared with the preset thread information and the maximum request volume that each thread can handle. If each thread can handle 50 requests, then 10 threads are required for the current 500 requests, that is, 500 / 50=10.

[0094] This embodiment calculates the ratio of the current request volume to the preset thread capacity or other analysis indicators. For example, if the current thread configuration is 20 threads, the gap between the current request processing capacity and the actual demand is calculated. The first analysis value calculated is 0.8, that is, the current processing capacity can meet 80% of the request demand.

[0095] This embodiment sets an analysis comparison threshold, such as 0.7. If the first analysis calculation value exceeds the threshold, it means that the current load of the system is close to or exceeds the expected processing capacity. Assume that the first analysis value is 0.8, which is greater than 0.7, so the thread configuration needs to be adjusted.

[0096] Then match the thread adjustment number in the corresponding information of the request thread, and find and determine the number of threads that need to be added according to the first analysis calculation value. For example, the first analysis value is 0.8, which means that the current thread configuration may not be sufficient to meet the actual request requirements, so additional threads are needed to improve processing capabilities. According to the preset adjustment strategy, 5 threads need to be added.

[0097] Finally, the thread adjustment information is obtained, and specific thread adjustment information is generated according to the matching results. If you need to increase from 20 threads to 25 threads, the system will automatically adjust the thread pool configuration and allocate the additional 5 threads to the website chat sub-channel so that the processing capacity meets the current request requirements.

[0098] Through real-time monitoring and dynamic adjustment mechanisms, the system can automatically increase the number of threads during periods of high request volume to ensure that each sub-channel can efficiently handle all requests, thereby improving user experience and optimizing system performance.

[0099] One embodiment is to continuously monitor and obtain the real-time request volume of each sub-channel and obtain the current call request volume, for example, there are currently 300 calls waiting in line to be answered.

[0100] The embodiment of the present application analyzes the current request quantity value and the preset thread information, and compares the real-time request quantity with the preset thread information, such as the number of calls that each telephone line can handle. Each telephone line can handle 50 calls, so 300 calls currently require 6 lines, that is, 300 / 50=6.

[0101] The embodiment of the present application calculates the ratio of the current request volume to the preset thread capacity or other analysis indicators. For example, if 10 telephone lines are currently configured, the gap between the current processing capacity and the actual demand is calculated. Assume that the first analysis value calculated is 0.6, that is, the current processing capacity can meet 60% of the request demand.

[0102] Set another analysis comparison threshold, such as 0.5. The first analysis calculation value exceeds the threshold, indicating that the current load of the system is close to or exceeds the expected processing capacity. The first analysis value is 0.6, which is greater than 0.5, so the thread configuration needs to be adjusted.

[0103] Then match the thread adjustment number in the corresponding information of the request thread, and find and determine the number of threads that need to be added according to the first analysis calculation value. For example, if the first analysis value is 0.6, it means that the current telephone line configuration is insufficient to meet the actual request requirements, and additional lines are needed to improve processing capabilities. According to the preset adjustment strategy, assume that 4 telephone lines need to be added. Based on the matching results, generate specific thread adjustment information. Assume that it is necessary to increase from 10 telephone lines to 14 telephone lines. The system will automatically adjust the telephone line configuration and allocate the additional 4 lines to telephone support to ensure that the processing capacity meets the current request requirements.

[0104] Through real-time monitoring and dynamic adjustment mechanisms, the number of telephone lines can be automatically increased during periods of high request volume to ensure that all calls can be answered efficiently, thereby improving user experience and optimizing system performance.

[0105] One embodiment is to continuously monitor and obtain the real-time request volume of each sub-channel, and monitor and obtain the number of currently unprocessed emails. For example, there are currently 800 emails to be processed.

[0106] The embodiment of the present application analyzes the current request quantity value and the preset thread information, and compares the real-time request quantity with the preset thread information, such as the number of emails that each customer service representative can handle per day. For example, each customer service representative can handle 100 emails per day. Then, currently 800 emails require 8 lines, that is, 800 / 100=8.

[0107] The embodiment of the present application calculates the ratio of the current request volume to the preset thread capacity or other analysis indicators. For example, if 12 lines are currently configured, the gap between the current processing capacity and the actual demand is calculated. The first analysis value calculated is 0.67, that is, the current processing capacity can meet 67% of the request demand.

[0108] Set an analysis comparison threshold, such as 0.6. If the first analysis calculation value exceeds the threshold, it means that the current load of the system is close to or exceeds the expected processing capacity. Assume that the first analysis value is 0.67, which is greater than 0.6, so the thread configuration needs to be adjusted.

[0109] Based on the first analysis calculation value, find and determine the number of threads that need to be added. If the first analysis value is 0.67, it means that the current thread configuration is insufficient to meet the actual email demand, and additional customer service representatives are needed to increase processing power. According to the preset adjustment strategy, assume that 4 threads need to be added. Based on the matching results, generate specific thread adjustment information. It is necessary to increase from 12 threads to 16 threads. The system will automatically adjust the customer service representative configuration and allocate the additional 4 threads to email support to ensure that the processing capacity meets the current request needs.

[0110] Through this real-time monitoring and dynamic adjustment mechanism, the email support system can automatically increase the number of threads during periods of high request volume to ensure that all emails can be processed efficiently, thereby improving user experience and optimizing system performance.

[0111] As one implementation mode, it also includes:

[0112] When the comparison result information is completed, satisfaction survey information for conducting a survey on users is obtained according to the channel category information matching.

[0113] When the comparison result information is incomplete, obtain the pending processing time information corresponding to the comparison result information, compare the pending processing time information with the preset time threshold, and obtain reminder information for reminder processing when the pending processing time information is greater than the time threshold.

[0114] Specifically, the embodiment of the present application uses a monitoring system, such as a Prometheus monitoring system, to monitor the processing of online chat requests using Prometheus, promptly discover processing delay problems, and play a role in real-time monitoring of request processing in various channels, thereby promptly discovering and solving problems.

[0115] In one embodiment, the embodiment of the present application obtains situation comparison parameters, and the system obtains the processing situation information of the sub-channel in real time and compares this information with the set comparison parameters. This step is used to evaluate the processing efficiency and quality of each channel.

[0116] In the embodiment of the present application, the system generates the comparison result information of the processing status according to the comparison situation. If the comparison result shows that the processing has been completed, the system will match the corresponding channel category information, prepare and send the satisfaction questionnaire to the user.

[0117] Further, the present application embodiment handles the unfinished situation. If the comparison result shows that the processing is not completed, the system will obtain the pending processing time information and compare it with the preset time threshold. If the pending processing time exceeds the threshold, the system will generate and send a reminder message to prompt timely processing.

[0118] After the request is processed, the system will send a satisfaction survey to the user to collect user feedback on the processing effect. This helps to evaluate the service quality and make improvements. This improves the timeliness of processing and enables the effective collection of user feedback.

[0119] Reference Figure 5 , an embodiment of the present application provides a data isolation and processing system for multi-channel customer service, the system comprising a channel identification module 1, a classification module 2 and a monitoring module 3.

[0120] The channel identification module 1 is used to obtain consulting information, analyze and identify the consulting information, and obtain channel category information;

[0121] Classification module 2, used to obtain total channel information, match sub-channels corresponding to the total channel information according to channel category information, and classify consulting information into sub-channels;

[0122] The monitoring module 3 is used to obtain the situation comparison parameters and the processing situation information of the sub-channels in real time, and compare the processing situation information with the situation comparison parameters to obtain the comparison result information reflecting the processing situation of each channel.

[0123] Specifically, the channel identification module 1 of the present application analyzes and identifies the obtained consulting information, determines which channel the consulting information is requested from, and obtains the channel category information. The classification module 2 matches the sub-channel corresponding to the total channel information according to the channel category information, and divides the consulting information into sub-channels, so that the consulting requests to be processed by the customer service personnel can be clearly divided. The monitoring module 3 obtains the situation comparison parameters and the processing situation information of the sub-channel in real time, compares the processing situation information with the situation comparison parameters, and obtains the comparison result information reflecting the processing situation of each channel. The real-time monitoring of the processing situation can enable the customer service personnel to allocate time more reasonably to handle the consulting requests, reduce the work difficulty brought to the customer service personnel by the multi-channel customer service system, and improve the processing efficiency of multi-channel information.

[0124] As one implementation method, the specific steps of the classification module 2 for classifying the consulting information into sub-channels are as follows:

[0125] Threads are constructed for the sub-channels, and the consulting information is divided into one of the threads of the corresponding sub-channel.

[0126] As one implementation method, the specific steps for the classification module 2 to construct threads for sub-channels are:

[0127] Obtain historical request data of the same sub-channel type, build a model based on the historical request data, and obtain a load prediction model;

[0128] Obtaining forecast request data for the current period according to the load forecast model;

[0129] Obtaining request thread corresponding information, matching the preset number of threads in the request thread corresponding information according to the predicted request data, and obtaining the preset thread information;

[0130] Build and adjust the number of threads in the sub-channel according to the preset thread information.

[0131] As one implementation method, the specific steps for the classification module 2 to construct threads for sub-channels are:

[0132] Acquire the current request quantity value in real time, analyze the current request quantity value and preset thread information, and obtain a first analysis and calculation value;

[0133] In the case where the first analysis calculation value exceeds a preset analysis comparison threshold, the thread adjustment quantity in the corresponding information of the request thread is matched according to the first analysis calculation value to obtain the thread adjustment information.

[0134] As one implementation manner, the channel category information of the channel identification module 1 includes telephone information, email information, online chat information and social media information.

[0135] As one implementation mode, it also includes:

[0136] The satisfaction information feedback module 4 is used to obtain satisfaction survey information for conducting a survey on users according to the channel category information matching when the comparison result information is completed.

[0137] As one implementation mode, it also includes:

[0138] The satisfaction information feedback module 4 is used to obtain the pending processing time information corresponding to the comparison result information when the comparison result information is incomplete, compare the pending processing time information with the preset time threshold, and obtain reminder information for reminder processing when the pending processing time information is greater than the time threshold.

[0139] An embodiment of the present application provides a data isolation processing device for multi-channel customer service, including a memory and a processor, the memory storing a computer program, and the processor being configured to run the computer program to execute the data isolation processing method for multi-channel customer service as described above.

[0140] An embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the data isolation processing method for multi-channel customer service as described above when running.

[0141] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device and product can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0142] In several embodiments provided in this application, it should be understood that the disclosed methods, systems, devices and program products may be implemented in other ways.

[0143] 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. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0144] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A data isolation processing method for multi-channel customer service, characterized in that: include: Obtain consulting information, analyze and identify the consulting information, and obtain channel category information; Obtain total channel information, match sub-channels corresponding to total channel information according to channel category information, and divide consulting information into sub-channels; Obtain situation comparison parameters and obtain processing situation information of sub-channels in real time, compare the processing situation information with the situation comparison parameters, and obtain comparison result information reflecting the processing situation of each channel.

2. The data isolation processing method for multi-channel customer service according to claim 1 is characterized in that: The specific steps of dividing consulting information into sub-channels are: Threads are constructed for the sub-channels, and the consulting information is divided into one of the threads of the corresponding sub-channel.

3. The data isolation processing method for multi-channel customer service according to claim 2 is characterized in that: The specific steps of constructing the thread for the sub-channel are: Obtain historical request data of the same sub-channel type, build a model based on the historical request data, and obtain a load prediction model; Obtaining forecast request data for the current period according to the load forecast model; Obtaining request thread corresponding information, matching the preset number of threads in the request thread corresponding information according to the predicted request data, and obtaining the preset thread information; Build and adjust the number of threads in the sub-channel according to the preset thread information.

4. The data isolation processing method for multi-channel customer service according to claim 3 is characterized in that: The specific steps of constructing the thread for the sub-channel are: Acquire the current request quantity value in real time, analyze the current request quantity value and preset thread information, and obtain a first analysis and calculation value; In the case where the first analysis calculation value exceeds a preset analysis comparison threshold, the thread adjustment quantity in the corresponding information of the request thread is matched according to the first analysis calculation value to obtain the thread adjustment information.

5. The data isolation processing method for multi-channel customer service according to claim 1 is characterized in that: The channel category information includes telephone information, email information, online chat information and social media information.

6. The data isolation processing method for multi-channel customer service according to claim 1 is characterized in that: Also includes: When the comparison result information is completed, satisfaction survey information for conducting a survey on users is obtained according to the channel category information matching.

7. The data isolation processing method for multi-channel customer service according to claim 1 is characterized in that: Also includes: When the comparison result information is incomplete, obtain the pending processing time information corresponding to the comparison result information, compare the pending processing time information with the preset time threshold, and obtain reminder information for reminder processing when the pending processing time information is greater than the time threshold.

8. A data isolation processing system for multi-channel customer service, characterized in that: include: The channel identification module is used to obtain consulting information, analyze and identify the consulting information, and obtain channel category information; The classification module is used to obtain the total channel information, match the sub-channels corresponding to the total channel information according to the channel category information, and divide the consulting information into sub-channels; The monitoring module is used to obtain the situation comparison parameters and the processing situation information of the sub-channels in real time, compare the processing situation information with the situation comparison parameters, and obtain the comparison result information reflecting the processing situation of each channel.

9. A data isolation processing device for multi-channel customer service, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the data isolation processing method for multi-channel customer service according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the data isolation processing method for multi-channel customer service described in any one of claims 1-7 when running.

Images